chapter 1: introduction

6 downloads 1627 Views 2MB Size Report
technologies that are available for presenting information on mobile phones; different ... Technology ()IHISM) Project – a departmental project aimed at developing and ...... Users who need information such as health information, license application .... illiterate in rural Bangalore) would be interested in for the job search for ...
UNIVERSITY OF BOTSWANA

FACULTY OF SCIENCE MSc. Computer Information Systems TITLE: Information Representation for the Semi-Literate and Illiterate Users Name: Resego Morakanyane ID No: 200001000 Supervisor: Dr. Audrey Masizana A Dissertation Submitted to the Department of Computer Science in Partial Fulfilment of Master’s Degree in Computer Information Systems

June 2010 1

Abstract The use of mobile technology in communicating has become a necessity in today’s digital world. On the other hand, access to one’s health information is equally trivial in today’s sicknesses. Different researches are therefore being carried out to see how the different communities can use mobile technology to access health information. This study considers the different technologies by which data and information can be transmitted from the requester (mainly communities) to the responder (the servers) when accessing health information through mobile phone technologies. Special interest is on the technologies that are available for presenting information on mobile phones; different technologies on that can be used to represent information for the users and finally selecting the most relevant technology that can be used to represent information for the users, especially the semi-literate and illiterate users in Botswana. The selected technology is then used to model a system that enables the users to access health information through mobile phones with as little difficulty as possible. This dissertation is a continuation of the previous studies on telemedicine which was aimed at looking at factors to consider when employing telemedicine technologies in Botswana. The study recommended that telemedicine was a viable option to Botswana’s health problems and it identified different technologies that can be employed to implement telemedicine. It is also influenced by the University of Botswana Computer Science Department’s Integrated Healthcare Information Service through Mobile Technology ()IHISM) Project – a departmental project aimed at developing and testing internet based healthcare information services.

Morakanyane R.S. MSc.

Page 2

Table of Contents CHAPTER 1: INTRODUCTION .............................................................................................................. 6 1.1 Overview of the Study ...................................................................................................................... 6 1.2 The Digital Inclusion Problem .......................................................................................................... 8 1.3 Research Objectives .......................................................................................................................... 9 1.4 Research Questions ......................................................................................................................... 10 1.5 Expected Outcomes ........................................................................................................................ 10 1.6 Motivation of the Study .................................................................................................................. 11 CHAPTER 2: LITERATURE REVIEW .................................................................................................. 12 2.1 Introduction ..................................................................................................................................... 12 2.2 About the Semi-Literate and Illiterate ............................................................................................ 12 2.2.1 Human Computer Perceptions in Semi-Literacy ..................................................................... 14 2.2.2 Psychological Behaviour in Semi-Literacy ............................................................................. 15 2.3 About Mobile Telephony ................................................................................................................ 16 2.3.1 Mobile Telephony in Botswana ............................................................................................... 17 2.3.2 Digital Inclusion Benefits of Mobile Telephony in Botswana ................................................ 18 2.4 Information Representation for Semi-Literate and Illiterate Users in Mobile Phones ................... 20 2.4.1 Use of Plain Full Text Messages ............................................................................................. 21 2.4.2 Grouping by Keywords ............................................................................................................ 22 2.4.3 Graphics and Picture Messaging .............................................................................................. 22 2.4.4 Voice and Speech Technology................................................................................................. 23 2.4.5 Use of Video Technology ........................................................................................................ 24 2.4.6 Use of Menu Based Technologies ........................................................................................... 25 2.5 Related Work .................................................................................................................................. 26 2.5.1 Text-Free User Interfaces for the Semi-Literate and Illiterate Users ................................. 26 2.5.2 OpenPhone Telephony System for HIV/AIDS Care-givers ............................................... 27 2.5.3 Ipoletse HIV/AIDS Call Center ............................................................................................... 27 2.5.4 Orange Prepaid Services Center .............................................................................................. 28 2.6 Conclusion ...................................................................................................................................... 28 CHAPTER 3: REQUIREMENTS GATHERING AND ANALYSIS...................................................... 30 3.1 Introduction ..................................................................................................................................... 30 3.2 Objectives of Data Collection ......................................................................................................... 30 3.3 Results and Analysis ....................................................................................................................... 32 3.3.1 Current Situation of Target Users ............................................................................................ 32 3.3.2 Accessibility and Use of Mobile Phone ................................................................................... 35 3.3.3 Perceptions and Psychological Acceptance ............................................................................. 37 3.4 General Experiences Gained During Data Collection ................................................................ 38 3.5 Conclusions and Recommendations ........................................................................................... 38 CHAPTER 4: THE DIGITAL INCLUSION SOLUTION ....................................................................... 40 4.1 Introduction ..................................................................................................................................... 40 4.2 Interactive Voice Response System ................................................................................................ 40 4.2.1 IVR System Software Requirements ....................................................................................... 41 4.2.2 IVR System Hardware Requirements ...................................................................................... 42 4.2.3 IVR System Architecture Requirements .................................................................................. 46 4.3 IVR System Functional Requirements............................................................................................ 47 4.3.1 User Interactivity ..................................................................................................................... 48 4.3.2 System Interactivity ................................................................................................................. 48 Morakanyane R.S. MSc.

Page 3

4.4 IVR System Non-Functional Requirement ..................................................................................... 49 4.4.1 Recording of Voice Clips......................................................................................................... 49 4.4.2 Nomenclature of Voice Clips................................................................................................... 49 4.4.3 Depth Level of Menu Options ................................................................................................. 50 4.4.5 Security Requirements ............................................................................................................. 50 4.5 Operational Requirements .............................................................................................................. 50 4.5.1 Call Handling ........................................................................................................................... 50 4.5.2 Length of an Instructing Voice Clip ........................................................................................ 51 4.5.3 Waiting Time Before Response ............................................................................................... 51 4.5.4 Error Handling ......................................................................................................................... 51 4.5.5 Consistency Control: ................................................................................................................ 51 4.6 Summary ......................................................................................................................................... 52 CHAPTER 5: DESIGN AND IMPLEMENTATION .............................................................................. 53 5.1 Introduction ..................................................................................................................................... 53 5.2 Design Considerations .................................................................................................................... 53 5.3 System’s Functional Design ........................................................................................................... 53 5.3.1 Data Design .............................................................................................................................. 53 5.3.2 Algorithm Design......................................................................................................................... 55 5.3.3 Software Design ....................................................................................................................... 60 5.3.4 Library Specification ............................................................................................................... 63 5.3.5 Storage of Voice Clips ............................................................................................................. 63 5.3.6 Example ................................................................................................................................... 64 5.4 System’s Non-Functional Design ................................................................................................... 65 5.4.1 Content Design......................................................................................................................... 65 5.4.2 Menu Levels............................................................................................................................. 66 5.4.3 Audibility ................................................................................................................................. 66 5.4.4 Waiting Time before Response ................................................................................................ 66 5.5 Implementation of the Ideal Digital Inclusion Solution ................................................................. 67 5.5.1 Configuration and Installation Process .................................................................................... 67 5.5.2 Testing...................................................................................................................................... 67 5.5.3 User Training ........................................................................................................................... 68 5.5.4 System Location....................................................................................................................... 68 5.6 Implementing the Prototype ............................................................................................................ 68 5.7 Challenges Encountered in Developing the Prototype ................................................................... 70 CHAPTER 6: CONCLUSION AND RECOMMENDATIONS .............................................................. 72 6.1 Conclusion ...................................................................................................................................... 72 6.2 Recommendations ........................................................................................................................... 73 REFERENCES ......................................................................................................................................... 74 APPENDICES .......................................................................................................................................... 77

Morakanyane R.S. MSc.

Page 4

List of Tables Table 1: Summary of the Data Collection Sample Environment and Population……….32 Table 2: Results on Usability and Interactivity with Mobile Phones by the Participants..36 Table 3: Results on Perceptions and Psychological Acceptance………………………...37 Table 4: Examples of IVR Software …………………………………………………….41 Table 5: User Task Analysis Summary…………………………………………………..48 Table 6: System Task Analysis Summary………………………………………………..49 Table 7: System Algorithm Design………………………………………………………55 Table 8: Sample Clips to be used in the System…………………………………………63 Table 9: A typical data input activity of the system……………………………………...64

List of Figures Figure 1: A picture of a typical family in a rural settlement……………………………...17 Figure 2: Cellular Network Coverage in Botswana……………………………………....18 Figure 3: Graphical Illustration of Information Representation on Mobile Phones……...26 Figure 4: Results on Current Situation of the Participants………………………………..34 Figure 5: Conceptual View of Components of an IVR System……………..…………....43 Figure 6: A Typical Workflow Diagram in the IVR System……………………………..45 Figure 7: Typical Architecture Components of the IVR System…………………………46 Figure 8: Conceptual View of the Interactive Voice Responsive System………………...54 Figure 9: Graphical Illustration of the Voicent IVR Studio for the Proposed Solution…..61 Figure 10: System’s Functions……………………………………………………………63 Figure 11: Configuring Text-To-Speech (TTS) Commands……………………………...69 Figure 12: Check Design Output………………………………………………………….70

Morakanyane R.S. MSc.

Page 5

CHAPTER 1: INTRODUCTION 1.1 Overview of the Study Botswana as a developing country is one of those countries hard hit by the HIV/AIDS pandemic with an estimated 330 000 to 380 000 of its adult population (aged 15-49) living with HIV/AIDS which is about 35.5 to 39.1% (WHO, 2005). The most affected population is in the rural areas and the semi literate and the illiterate communities remain the victims (CSO, 2005). In his address to the UN General Assembly in 2001, Botswana’s then President Festus Mogae cried that ‘We are threatened with extinction. People are dying in chillingly high numbers. It is a crisis of the first magnitude.’ He advocated these deaths to the lack of adequate health information especially on HIV/AIDS which is the number one cause of death in Botswana (Mogae, 2001). In an attempt to fight this major challenge, the Government of Botswana, through its different ministries, has created some programs which attempts to reduce the impact made by this scourge. Some of these attempts include: the rolling out of the anti-retroviral (ARV) therapy; setting up of departments such as NACA, which are dedicated to different issues relating to HIV/AIDS as well as availing HIV/AIDS information by all means possible. Nevertheless, Botswana is still faced with yet other constraints which overwhelm efforts of fighting the disease. These constraints include the following:

Shortage of Health Personnel Botswana is currently facing a serious shortage of health personnel which is a very risky situation in this HIV/AIDS pandemic. Botswana Ministry of Health’s human resource levels indicate that in 2004 approximately 30 physicians and 262 nurses were available per 100,000 population which is a very high ratio as compared to the situations in the other countries (vs. 56 / 471 for South Africa and 229 / 897 for Canada) (BOTEC, 2004).

Low Literacy Level The Botswana Literacy Survey Report of 2003 shows that by 2003 the literacy rate of Batswana over the age of 15 was 81% (UNESCO, 2004). The report went on further to show that the literacy rate is lower in the rural areas as compared to the urban areas (see Appendix 1). From the table, Kweneng West, Ghanzi, Kgalagadi South and Ngamiland West all of which are predominantly rural, have literacy rates of 58, 60, 65 and 66 % respectively.

Morakanyane R.S. MSc.

Page 6

Uneven Population Distribution and Geographical Infrastructure The Central Statistics Office Population Survey Report of 2001 states that Botswana is also facing a challenge of uneven population and development distribution with most population, developments and higher levels of literacy found in urban areas while rural areas experience less population, less developments and lower literacy levels (semi literate and illiterate) who are also most economically challenged (CSO, 2002).

Inadequate Methods of Presenting and Disseminating Information Although there is no previous research on this issue, one also observes that the main methods of disseminating information in Botswana, which include radio, newspapers, pamphlets, television and campaigns are not necessary adequate as information disseminated through this media is not necessarily customized for the needs of the semi-literate and illiterate in the rural areas.

In an attempt to address these different issues, different researches and projects have therefore been proposed and carried out as ways of contributing towards improving these situations. In attempts to address issues of shortage of health professionals the Botswana Technology Centre undertook a Telemedicine Feasibility Study to see if telemedicine could be a solution to Botswana’s shortage of health professionals. The Ministry of Health on the other hand is currently implementing the Integrated Patient Management System (IPMS) which is aimed at improving services at different health facilities while the Botswana-Baylor Children’s Clinical Centre of Excellence is planning an OpenTelephony Pilot Project which will provide the home based pediatric care givers with different health information that they may require during the process of care giving. Furthermore, the Ministry of Health has established a telemedicine link between the country’s largest hospital – Princess Marina and Ghanzi Primary Hospital; one of the primary hospitals in the semi urban areas (BOTEC, 2004).With regard to low literacy level, besides availing formal education to all parts of the country, the government has attempted programs like adult education and distance learning. This is in an attempt to avail the opportunity to those who could not have access to formal education.

Even though the above mentioned initiatives help in terms of alleviating problems of shortage of health personnel, they have not managed to totally address issues of information dissemination and representation for the semi literate and illiterate populations to access and use effectively. There is still a

Morakanyane R.S. MSc.

Page 7

need for provision of information to semi-literate and illiterate users especially in rural areas through mechanisms better customised for their needs.

1.2 The Digital Inclusion Problem It has been established that there are some members of the community, the semi-literate and illiterate, who have difficulties in accessing or understanding health information that the health professionals would otherwise want them to receive (CSO, 2005). This could be due to the far distances at which these populations live or due to inability to comprehend the information in the form it is being presented in. These populations are commonly found in rural areas. With Botswana being hard hit by the HIV/AIDS scourge, these people are vulnerable to contracting the disease as they have limited means of accessing and comprehending information being passed. Mostly information on HIV/AIDS is disseminated through pamphlets, books and other materials, most of which are availed on request. The semi-literate and illiterate members of society therefore would not get these very important messages on prevention, treatment, care and support that various stakeholders in the fight against the disease normally render.

The war on HIV/AIDS will not be won if some members of society are excluded not because they do not want to be involved but by the fact that they could not access the readily available information since they cannot comprehend the contents of the medium of communication used. It is against this background that the difficulties which the semi-literate and illiterate members of society have in accessing information are to be addressed. There is need to explore more friendly modes of disseminating information to this group and representing it in a way that they would comprehend it. This research will therefore explore the use of mobile technology as one mode of disseminating information.

It is also observed that Botswana has a high mobile penetration of about 85 people for every 100, including rural areas and semi-literate and illiterate populations (BTA Report, 2009). With this, the research aims at using this opportunity to explore how mobile technology can be utilized to improve access to health information by representing the information in ways that the semi-literate and illiterate populations can be able to comprehend and use effectively.

Morakanyane R.S. MSc.

Page 8

In a nutshell, uneven population distribution, shortage of health workers, ineffective methods of representing HIV/AIDS information for the semi-literate and illiterate communities result in these communities not receiving HIV/AIDS that the health providers would otherwise want them to receive. From these factors, this research aims at taking advantage of the high mobile penetration in Botswana to define the digital inclusion problem. The research aims at addressing these issues by looking at the technical feasibility of using mobile phones to disseminate and improve accessibility to HIV/AIDS information as well as look at human perceptions of the availed technology with regard to disseminating and improving accessibility of information.

1.3 Research Objectives This study will investigate the different ways on how information can be represented in mobile phones and finally selecting one way that can be used to represent HIV/AIDS information for the semi-literate and illiterate users. The problem therefore, henceforth referred to as the Digital Inclusion Problem is to propose a system that will best serve semi-literate and illiterate users to access HIV/AIDS information using mobile phones. To achieve this, following factors will be taken into consideration:

Technical Factors Feasibility: Establishing whether HIV/AIDS information can be represented on mobile phones as well as establishing if the high mobile penetration can enable the semi-literate and illiterate users to access HIV/AIDS information more easily. Accessibility: Establishing if the current national telecommunications infrastructure would support the solution to be proposed to be easily accessible. Affordability: Establishing if the target users would find it affordable to use the solution to be proposed as well as establish if they would find the solution user friendly. Interactivity: Establish the system that would allow the user to select options of what they want information on Human Factors Level of Literacy of Target Users: Having identified the HIV/AIDS information needs of the semiliterate and illiterate people, the research will attempt to represent these information needs in a way that the users will be able to access efficiently and comprehend, regardless of their level of literacy.

Morakanyane R.S. MSc.

Page 9

Psychological Behaviour of Target Users: Here the research will concern itself more with the nature of thought, perceptions and emotions that the semi-literate and illiterate population will have with regard to how they will react to using technology to access HIV/AIDS information. Human Computer Perception Issues: Having identified the level of literacy of target users as well as their psychological behaviour towards accessing HIV/AIDS information through mobile phones, the research will look at how the information can be represented for the target users (semi-literate and illiterate) on mobile phones. It will also look at efficient and effective ways of how the users will access this information. Both representing the information and accessing the information will be carried out using a mobile phone.

1.4 Research Questions With the semi-literate and illiterate populations of Botswana having difficulties in accessing and comprehending HIV/AIDS information being presented to them while the country has a high mobile penetration, the study will therefore attempt to provide answers to the following questions: -

What are the different ways in which information can be represented for the semi-literate and illiterate users on mobile phones?

-

What are the human computer perception issues that need to be considered when developing the digital inclusion solution for semi-literate and illiterate users on mobile phones?

-

What different psychological behavior issues must be considered when developing the digital inclusion solution for semi-literate and illiterate users on mobile phones?

1.5 Expected Outcomes The expected outcomes in this research include well researched ways in which information can be represented on mobile phones such that all users, regardless of the level of literacy, can efficiently access and comprehend HIV/AIDS information with as little difficulties as possible. The solution to be proposed is expected to include the information needs of the semi-literate and the illiterate members of the society. It is also to provide user friendly interfaces that will enable these populations to interactively access the information at no or minimal costs. Specifically this solution will target users in the rural areas as they are normally the economically challenged; also they are the forgotten population when it comes to information dissemination and provision of human social services. Morakanyane R.S. MSc.

Page 10

1.6 Motivation of the Study In recognition of the aforesaid, the University of Botswana Computer Science Department has since proposed the Integrated Healthcare Information Service through Mobile Technology (IHISM) Project. This project is meant to contribute towards communities, especially rural, being able to access health information through mobile technology. The motivation behind carrying out this research is therefore as follows:

1. As a member of the University of Botswana Computer Science Department, one felt the need to participate in this departmental project which has a great potential of being rolled out as a national project.

2. Content developers tend to ignore the expected end user issues when they develop information systems. Considering human computer interaction issues such as the level of literacy of the expected users and how ready are the users to accept and use these complex systems is a very important aspect in development. This research will therefore attempt to propose a solution which will ensure that all users, regardless of their level of literacy, are able to efficiently and optimally access HIV/AIDS information using their mobile phones.

Morakanyane R.S. MSc.

Page 11

CHAPTER 2: LITERATURE REVIEW

2.1 Introduction Having defined the digital inclusion problem, it is therefore important to take an in-depth look at the attributes defining this digital inclusion problem. This chapter therefore aims at looking into issues of the semi-literate and illiterate: what defines the semi-literate and illiterate; their perceptions and psychological behaviour towards use of technology. The chapter would also look into issues of mobile telephony in Botswana so as to justify why a mobile phone based solution for the digital inclusion problem. The chapter would then look into issues of representing information for the semi-literate and illiterate people in mobile phones. The chapter would then conclude by analyzing already existing digital inclusion solutions with regard to the defined digital problem.

2.2 About the Semi-Literate and Illiterate The definition of literacy differs depending on the different contexts in which the word is used (Van Niekerk, 1999). Collins dictionary gives a general definition of literacy as the ability to read, write, communicate, and comprehend. This, of course, raises the issue of the level of the complexity required for reading and or writing. UNESCO suggests that a literate person is the one who can read and write a short statement on their everyday life (Chlebowska, 1990). Nevertheless, a lot of people considered to be literate by UNESCO are not able to read a newspaper or an information pamphlet. Attempts to define literacy often set the standard in relation to the number of years of formal schooling. Of course there are quite a few problems with such a definition. Firstly, literacy is a highly perishable skill unless continuously practiced, and people may have obtained their formal educational qualifications many years previously. Secondly, there are rarely any consistent standards relating educational level and literacy ability. There may be large differences between educational authorities, examination boards, schools, teacher ability to name but a few.

The semi-literate and the illiterate therefore are people who have never gone to school to have formal education or people who have been to school at some stage of life, but have lost touch with reading and or writing. These could also be people who have learnt selective terms, at their own initiative by peer learning. In their study to learn more about the semi literate and illiterate, Joshi and Rasal (2002) realized that most semi-literate and illiterate populations live in rural areas. Other researches also shows Morakanyane R.S. MSc.

Page 12

that close to 4.8 Billion semi-literate and illiterate people live in developing countries and many of these live in rural areas (UNDP, 2005).

In computers, literacy then refers to the knowledge and ability to use information and communication technologies efficiently. It can also refer to the comfort level someone has with using computer programs and other applications that are associated with computers. As of 2005, having basic computer skills is a significant asset in the developed countries (Computer Literacy, 2007). The semi-literate therefore refers to people who have very minimal information and communications technology skills while the illiterate refer to populations who have absolutely no information and communications technology skills.

This research will therefore adopt the meaning of literacy as the ability of an individual to use information and communications technologies efficiently. The literacy level will however be strongly influenced by the level of education an individual has received, with those who have received lower levels of formal education being viewed as more of illiterate than literate. It is also important to note that the terms semi-literacy and target users may, in some instances, be used to refer to both cases of partial literacy and total illiteracy henceforth.

It has been observed that rural semi-literate and illiterate populations in Botswana have many information needs which are not effectively addressed. Such information needs include access to government and social services, learning about empowerment initiatives that exist in other areas, communicating with distant family members, friends, customers or suppliers, easy access to crucial information, such as disease out-breaks; spot announcements, among others. This is mainly because these populations are in rural areas and it is difficult for information to reach them and where it reaches them, comprehending the information is not necessarily easy.

These communities also face other constraints such as lower purchasing power, poor connections to main lines of communication and infrastructure (power and fixed line network) as well as lower levels of general education. As a result these challenges hinder these populations by making them the victims of digital divide and lagging behind when it comes to information accessibility.

Morakanyane R.S. MSc.

Page 13

Because of lack of exposure to technology, the semi-literate and illiterate people may suffer anxiety, become impatient and withdrawn when technology is being presented to them. Baecker and Buxton (1987) describe how novice users, who are also somewhat semi-literate or illiterate, feel frustrated, insecure and even frightened when they have to deal with systems whose behaviour is incomprehensible, mysterious and intimidating. At the same time, these populations significantly contribute to the high levels of mobile phone penetration as they have been observed to be among mobile phone users. The technology solutions to be proposed therefore, must take into consideration and strive to provide appropriate and relevant solutions that will benefit these people. It is for these reasons that Human Computer Perceptions have to be a key factor in finding solutions to represent information for these populations.

2.2.1 Human Computer Perceptions in Semi-Literacy Human Computer Perceptions (HCPs) concern themselves with how human-beings, in this case semiliterate and illiterate, would look into interacting with the solution for the digital inclusion problem. These perceptions, which are strongly influenced by Human Computer Interaction (HCI) concepts, would consider interaction between users (semi-literate and illiterate) and systems (solution to the defined digital divide problem). Interaction between users and ICTs occurs at the user interface level, which can be software, hardware or both. For example, in a mobile phone, the mobile phone screen, keypads and the mobile phone operating system are all user interfaces. The basic goal of HCI then is to ensure that users find it easy to maneuver through the different menus of a mobile phone. Ensure that user use the phone with ease as well as improving interaction between the user and system (phone) by making the system more receptive to the user’s needs.

Human Computer Perceptions in the Digital Inclusion Problem The digital inclusion problem is more concerned with providing technology based solutions to improve information access to people in rural areas. In this case, it is concerned with providing a mobile phone based solution used by the semi-literate and illiterate rural communities to access information. With the semi-literate and illiterate being people with very minimal or no information and communications technology skills, HCPs issues are therefore concerned with ensuring that the mobile phone based system to be used by the semi-literate and illiterate users allow users to maneuver through the system with ease as well as allow users to select options of what they want information on in the system. Morakanyane R.S. MSc.

Page 14

For instance, since when using a Bank Automatic Teller Machine (ATM) a semi-literate person may have difficulties in maneuvering through the complicated process of getting money. HCPs professionals may therefore develop user interfaces that should enable the semi-literate or illiterate user to use the ATM with ease. In their study to look at the different attitudes semi-literate and illiterate users have towards ways in which information has been represented in ATMs Thatcher, et.al. (2002) identified speech-based and icon-based interfaces as the best possible interfaces that can be used to present information to the semi-literate and illiterate on ATMs.

In their study on e-Advocacy for the Global South, Joyce (2007) indicated how mobile phones are a great opportunity for e-advocates who want to reach a mass audience. They went on further to describe different opportunities where different functions of mobile phones can be used to bridge the digital divide. However, Joyce (2007) went on to state how SMS is not a silver bullet for the digital divide, emphasizing more on how they could be a communication barrier to the semi-literate and illiterate. In the same way, this research will consider human computer perceptions to determine which interfaces would be more relevant when representing HIV/AIDS information for the semi-literate and illiterate users on mobile phones.

2.2.2 Psychological Behaviour in Semi-Literacy Psychological behaviour in semi-literacy is more concerned with nature of thought and emotions that the semi-literate and illiterate populations have with life in general. In this particular digital inclusion problem, psychological behaviour will be concerned with how the target users will react to using mobile phones to access critical information such as HIV/AIDS information. By nature, the semi-literate and illiterate populations are observed to be passive when it comes to issues of technology. They appear to be withdrawn especially when complex technical procedures are applied. At the same time, these populations significantly contribute to the high levels of mobile phone penetration as they have been observed to be among mobile phone users (Parikh, 2006).

Psychological Behaviour in the Digital Inclusion Problem The technology solution to be proposed for this problem must recognize that the target users are facing some challenges. Since the solution will be based on mobile technology, psychological behaviour issues that may need to be considered include: Acceptance- how users will not be intimidated by their inability Morakanyane R.S. MSc.

Page 15

to understand the use of technology; Affordability: if the users be able to afford the costs that come with the system being proposed; Integrity: if the users will be able to trust the information being obtained from the system. In their study on Text-Free User Interfaces of Semi-literate and Illiterate Users, Medhi (2008) further described the importance of intense interaction with the target users during requirements gathering. This is so that target users own up the system right from the beginning to reduce chances of intimidation in the future.

2.3 About Mobile Telephony Wikipedia defines a mobile phone also called cellphone or cellular phone as a long-range, electronic device used for mobile voice or data communication over a network of specialized base stations known as cell sites (Mobile Telephony, 2009). In addition to the standard voice function of the mobile phone, mobile phones also support additional services and accessories such as Short Message Service (SMS) for text messaging and Multimedia Messaging Service (MMS) for sending and receiving graphical images, rich text and videos. Use of mobile phones dates as far back as in the early 1900s when U.S. Patent 887,357 (2009) for a wireless telephone was issued in to Nathan B. Stubblefield of Murray, Kentucky. In 1910 Lars Magnus Ericsson applied the idea of mobile phone by installing of telephone in his car. Using a pair of long electric wires he could connect to national telephone network travelling across the country. From these times different generations of mobile phones were invented by different researchers.

In 1947 hexagonal cells for mobile phone base stations were invented by Bell Labs engineers at AT&T and were developed further during the 1960s. In the Second World War years radiophones were used. Fully automatic cellular networks were first introduced in the early to mid-1980s (the 1G generation). The first fully automatic mobile phone system was the 1981 Nordic Mobile Telephone (NMT) system (Mamboservis, 2008). In Botswana use of mobile phones dates as far back as 1998. Due to their low establishment costs and rapid deployment, mobile phone networks have since spread rapidly throughout the country, outstripping the growth of fixed telephony (BTA Report, 2009).

Morakanyane R.S. MSc.

Page 16

2.3.1 Mobile Telephony in Botswana

Figure 1: A picture of a typical family in a rural settlement

Figure 1 above was taken during one of the data gathering field visitations, after one of the family interviews. The picture clearly shows a family in a rural setup. From the conversations held with the family, it was noted that most of the family members could be classified under the semi-literate or illiterate populations. A closer look at the picture shows an amazing factor. One of the members has a mobile phone. The next question could then be what would the mode of communication with family members in distant places be and the obvious answer was the mobile phone.

In his article titled Telecommunications in Botswana (Tamocha, 2007) verifies that mobile phone technology has taken off Botswana and Africa like no other technology ever did in the history of the continent. He states that Neil Ford of the Pan-African Business Magazine indicated that Africa has the fastest growing mobile phone market in the world with around 50 000 new subscribers everyday. He goes further to state that the Botswana Telecommunications Authority - a body which licenses telecommunications service providers, puts the figure of cellphone users in Botswana at 85 people for every 100 (BTA Report, 2009). Morakanyane R.S. MSc.

Page 17

Source: Orange (2009)

Source: Mascom (2009)

Figure 2: Cellular Network Coverage in Botswana

The two figures above indicate the density of cellphone coverage in Botswana. It is clear that the use of mobile phones is common throughout the country, including rural areas where most of the semi-literate and illiterate populations are. The two mobile operators and the newly licensed BTC aim at closest to 100% network coverage by the year 2016 as a contribution towards the Botswana National Vision (BTA Report, 2009). It is for these reasons that mobile phones will be used to bridge the digital divide through exploring how they can improve access to HIV/AIDS information for the semi-literate and illiterate populations. Mobile technology is after all the leading mode of telecommunications in Botswana and has reached a high level of acceptance by the users (Tamocha, 2007).

2.3.2 Digital Inclusion Benefits of Mobile Telephony in Botswana Parikh (2006) emphasizes how mobile phones are the right devices for rural accessibility. He advocates for mobile phones as ‘rural computers’. The hardware is a great fit for rural conditions since it is batterypowered and networked at a low cost. He goes on to observe that numeric keypads and the utility of voice communications are familiar to billions of users and therefore many users in rural places, including the semi-literate and the illiterate are generally comfortable with the use of mobile phones. Most importantly, mobile telephony can be used in accessing vital information such as latest news and out-breaking health information. Populations staying in settlements geographically disadvantaged from Morakanyane R.S. MSc.

Page 18

the main lines of communication can, through the use of mobile telephony, get in contact with the connected areas and information sources to access information on different subjects. This will hence save the rural settlement dwellers the costs of traveling to the information sources and will help them access information promptly.

Although mobile telephony is just over 10 years in Botswana, mobile penetration has grown exponentially. The benefits that it has brought have also surpassed by far the number of years it has been around. These benefits are significant to both general users and service providers (BTA Report, 2009). A typical example of a situation where mobile telephony was seen to be bridging the digital divide was the case of a nurse working in one of the remotest areas of Botswana.

The clinic in which the nurse worked at did not have a fixed line but rather used a radio phone to communicate with other health facilities. It happened that one evening when the nurse was on night duty, she had a woman on labour having some complications. The nurse tried to contact the other health facility for help through the radio phone but the line had too many disturbances and she could not reach the other end. She then used her cellular phone and she managed to contact a doctor who at the time was not even at the medical center. With this she managed to get the help she needed and she saved two lives- that of a mother and the baby that was being born. Thanks to mobile telephony.

However, it is important to note that although there are many uses of mobile phones in Botswana, there still are some information needs which are left out. It still remains difficult to access general social services information. Users who need information such as health information, license application information, business registration information and other information are still required to visit the information resource center to make their enquiries. There are no measures in place yet to help them access the information they need from where they are. On the other hand, where information is made available, most of the time it is in a format that is probably not easy for the semi-literate and illiterate to access, comprehend and utilize. With Botswana’s high mobile phone penetration, including in the rural areas where most of the semi-literate and illiterate populations are, it then calls for initiatives to see how this situation can be manipulated for the better. It is from this where the concept of exploring ways in which HIV/AIDS information can be represented for the semi-literate and illiterate users in mobile phone was conceived. Morakanyane R.S. MSc.

Page 19

2.4 Information Representation for Semi-Literate and Illiterate Users in Mobile Phones With the semi-literate and illiterate populations having different information needs that need to be urgently addressed, effectively representing information for them remains critical. The importance of representing health information, especially HIV/AIDS information, in a form that these populations can easily understand cannot be overemphasized. Well represented information will therefore allow these users to:

-

access more precise meaning of what is being communicated

-

access information which is in a form that is easy to understand

-

retrieve information in a fast and effective manner

Before discussing the different mechanisms of representing information for the semi-literate and illiterate, it is important to highlight the different data transmission technologies in mobile phones. These include the following: Short Message Service Technology (SMS): a mobile phone service that permits the sending of short messages between mobile devices. These could be plain full text messages or keyword grouped messages. Multimedia Messaging Service Technology (MMS): a standard for a telephony messaging systems that allows for sending of messages that have multimedia objects such as images, audio, video and rich text. Voice Technology: a technology which involves the use of voice to access functions of mobile phones. Different voice technologies include the normal voice to voice phone call; voice message, voice to text messages, interactive voice response technologies, among others.

It is therefore important to familiarize with these data transmission technologies as the solution to be proposed for representing HIV/AIDS information for the semi-literate and illiterate populations will be based around the use of some of these technologies. The following section therefore discusses different ways in which information can be represented in mobile phones.

Morakanyane R.S. MSc.

Page 20

2.4.1 Use of Plain Full Text Messages This is communication involving the sending and receiving of full plain text messages on mobile phones. This way of communicating has been developed from the SMS technology and it can be viewed as normal exchange of messages with continuous sentences. Within the context of the digital inclusion problem defined in this research, the users can be able to send full text SMS to the system to request for HIV/AIDS information as well as receive full text SMS from the system as responses. The system can also be used together with the faxes and emails such that when the users’ requests are larger than the normal SMS file (160 Characters), the response is sent to a fax number or email address that the user may have been required to provide.

Strengths: 1. Relatively cheap way of communicating which is affordable to even the economically disadvantaged communities 2. Can concatenate up to 3 messages and send them as one delivery to increase the level of detail of information being requested 3. Allows for message broadcasting Constraints: 1. Normally SMS are limited to 160 characters per message making it difficult to communicate all information requested to a larger degree of detail. 2. Most keypads do not support local languages which may be the mode of communication for the semi-literate and illiterate population therefore when typing in a request in a local language the phone may not be able to comprehend the language being used. 3. Fax machines are not very common to communities in the rural places. They can also be complex for the semi-literate and illiterate people whom the solution to be proposed is expected to target. Therefore representing data as full text messages with idea of sending the responses to fax machines will not be very beneficial to the rural users. 4. SMS communication requests users to type their request and read responses. Besides this being the major hindering factor to the target users (semi-literate and illiterate), it also involves a lot of cognitive load. Therefore the recipients may overlook some of the information they may be required to read due to inability to read.

Morakanyane R.S. MSc.

Page 21

2.4.2 Grouping by Keywords Grouping by keywords suggest that information and ideas can be grouped by a general keyword to represent ideas that the system may want to communicate to the users. This way of representing data can be developed from the SMS technology as well as utterance in voice technology. A typical example of grouping by keyword can be where a user sends the phrase ‘fighting the stigma’ to the system and the system picking the keyword sent to search the database. The system may then extract information to do with ‘fighting the stigma’ and send it to the number that sent the request. In this example, issues to do with the HIV/AIDS stigma such as HIV/AIDS in the Work Place, Living Positively, etc. can be grouped under one key word – ‘fighting the stigma’.

Using voice technology grouping by keywords is referred to as Utterance. This means the vocalization or speaking of a word or words that represent a single meaning to the system. Utterances can be used in the same example where the user can send the vocal statement ‘fighting the stigma’ and the system returning the response either through text or voice.

Strengths 1. Utterance is ideal for target users - who cannot read and type 2. Well organized and classified information hence it becomes easy to retrieve it from the database 3. Requires less cognitive load than full text messages as the users has to remember only the keywords used in the system Constraints 1. Typed in grouping by keyword is not ideal for target users – who cannot read and type 2. May require intense training so that the users know exactly the keywords used in the system for them to be able to use it and to avoid miscommunications 3. Keywords may tend to generalize information a lot making it difficult to obtain precise meaning and detail of what one is looking for

2.4.3 Graphics and Picture Messaging This method of data representation, which has been developed from the MMS technology, involves grouping information or ideas and representing them by graphical images. Graphics that can be used include icons, pictures, images, high resolution text, symbols to mention but a few. The common use of Morakanyane R.S. MSc.

Page 22

icons in mobile phones for instance, is the outline of an envelop to represent an unread message whereas a shaded envelop represents a read messages. Within the context of the digital inclusion solution to be proposed: in situations where a user needs to take 2 green pills at 1330 hours, the system may generate a picture message of a glass of water and two green pills and send it to the user at half past one as a reminder that it is time to take the two green pills.

Strengths 1. This method can be used where the users do not have the ability to read. 2. If the users are well trained in the meanings represented by the graphics then graphics require less cognitive load Constraints 1. Use of graphics to represent information is an application of the MMS technology therefore it has the constrains faced by the MMS technology such as incompatible content adaptation, difficulties in sending to distribution lists, inefficiency in sending bulk messaging and incompatible handset configuration. 2. The method may require a lot of user training such that users know exactly what each image means so as to avoid misinterpretations of information. 3. Image transmissions require special networks and may be carried out at high costs.

2.4.4 Voice and Speech Technology This method of representing information is developed from the voice technology which basically involves the use of human voice as a way of communicating. Typical examples of voice and speech technology include the system sending voice reminders at the times which the user has to take medication; a user calling in and talking with the system to enquire about any information they may be looking for; continuous conversation between the system and the user where the system identifies the user by their voice; where the user can call the system and follow the voice instructions they are getting through accessing the menu being presented by pressing the key pads; where the user can call the system and follow the voice instructions they are getting through accessing the menu being presented by voice.

Morakanyane R.S. MSc.

Page 23

Strengths 1. Uses less cognitive load 2. Can be used where users cannot read or write 3. Allows for interactivity as the users is able to select options of what they may be looking for and seek repeat of instructions provided Constraints 1. Spontaneous speech is bound to contain a lot of disfluencies and is much more difficult to recognize than speech read from script. 2. Recognition is generally more difficult when vocabularies are large or have many similar-sounding words. 3. Coming with a system that identifies individual voices is a complex and cumbersome procedure which is almost not possible to come up with as human voices are of different pitches and frequencies and one person can have different voice pitches and frequencies according to the different times of the day, climate conditions and may other factors.

2.4.5 Use of Video Technology This involves representing data using a combination of images and sound over time therefore this method is developed from both MMS and voice technologies. Typical video applications include the system sending video clips of general health information to the users.

Strengths 1. Video may tend to be more effective as it provides both images and sound Constraints 1. Require complex technologies to transmit 2. May be slow where there is limited bandwidth 3. Occupy large spaces and large files take time to transmit 4. They can be less interactive as the user will just be watching and having no major input

Morakanyane R.S. MSc.

Page 24

2.4.6 Use of Menu Based Technologies Menu based representation is a representation in which users traverse through some form of preprogrammed menu in order to access the system. A typical application of a menu based system include the common mobile communications providers’ prepaid service which the customer dials and access a menu to access different services such as checking their balance, playing a voice message, recharging their airtime and so on. Another example of a menu based representation is the common graphical or text based menu of a mobile phone which allows the user to access the different functions and change settings of their mobile phones.

In the context of the digital inclusion solution to be proposed, menu based representation can involve a preprogrammed text based or graphical menu linking to the system which users can download into their mobile phones and use it to access the system from any point. Another alternative can be a preprogrammed speech based menu which can be present to the users when they dial the number that activates the system.

Strengths 1. Easy to represent and to access 2. Basic menus will not be complicated to the users hence the users will experience less difficulties in accessing the system Constraints 1. Options tend to limit what information can be represented 2. Users may find it difficult to use the system if there are too many menu options in the system It is therefore important to realize that a combination of the technologies discussed above does not restrict itself to a single solution but rather different solutions. Different solutions that can be obtained from different combinations of these technologies include among others: plain text SMSes; menu-based speech recognition; text-to-speech; speech synthesis; menu-based SMS grouping by keyword; picture messaging; video messaging. Figure 3 below is a graphical illustration of the different ways in which information can be represented in mobile phones.

Morakanyane R.S. MSc.

Page 25

Figure 3: Graphical Illustration of Information Representation on Mobile Phones

2.5 Related Work It is important to emphasize that this study is not the first attempt towards the digital inclusion problem. Different other related studies have been carried along the similar issues. Below is a brief analysis of two of the already existing solutions so as to determine their strengths and shortfalls which shall contribute towards the basis of attempting a new digital inclusion solution.

2.5.1 Text-Free User Interfaces for the Semi-Literate and Illiterate Users In this study Medhi (2008) describe work towards the goal of a user interface designed such that even the novice illiterate users do not require any intervention to use the system. The research adopted the ethnographic design process to understand the kind of application the target users (semi-literate and illiterate in rural Bangalore) would be interested in for the job search for domestic labourers’ application as well as for a generic map that could be used for navigating a city. To this, the researchers observed that participants preferred semi-abstract cartoons and more photo-realistic graphics much better than complex abstract graphics. The researchers then developed the solution for the problem using specially designed applications which they tested on the users. Although the researchers have not yet achieved a Morakanyane R.S. MSc.

Page 26

fully assistance-less user interface for these novice users, they strongly believe that the achieved work took them a step closer to the solution. Although this solution is not a digital inclusion solution for providing HIV/AIDS information, experience gained from looking at the process of how this solution was developed could be used as part of design guidelines for developing relevant interfaces for semiliterate and illiterate people.

2.5.2 OpenPhone Telephony System for HIV/AIDS Care-givers Meraka Institute in the CSIR has developed an OpenPhone Telephony System for Botswana Baylor Children’s Clinical Centre of Excellence (Kuun, 2007). This system aims at providing a platform for HIV/AIDS care-givers to access the system and enquire on any care-giving challenges they may encounter as they take care of the HIV/AIDS infected children. To access the system care-givers have to make fixed line telephone calls to the system where the system would play pre-recorded voice clips directing the caller what to do in the form of an interactive voice response system. The team has this far managed to develop and test the system on the real users. Although the system seem like a possible solution, it still has limitations in that callers are restricted to using fixed line telephones, which are not necessary readily available to all parts of the country. The content information accessed from the system is also restricted to HIV/AIDS care-givers for children.

2.5.3 Ipoletse HIV/AIDS Call Center This is a live call center in which members of the public in Botswana can call to enquire on any information about HIV AIDS (MRI, 2002). In this centre callers make a call to a toll free fixed line number which is available 24 hours and an agent sitting at the other end receives the calls and provides the required information. This call center serves as a general information center and does not cater for special needs such as representing the information for the different beneficiaries such as the literate, semi-literate and the illiterate. It also does not concern itself about exploring the different technologies that can be used to represent information for the different beneficiaries. Due to this, the system is limited to users who have access to fixed telephone lines living out the semi-literate and illiterate users who do not have access to fixed lines in rural areas.

Morakanyane R.S. MSc.

Page 27

2.5.4 Orange Prepaid Services Center This is a menu based IVR system which Orange subscribers can call to access information on services provided by the company. Such services include checking air time balance, recharging airtime, Orange information service to mention but a few. To access this system the user dials the number 155 on their Orange mobile phone and follow through the menu as directed by the system (Orange, 2008). The system does not target any particular customer but targets all customers regardless of location, literacy level, type of mobile phone or otherwise. Although this system is totally not a digital inclusion solution, the idea behind the system can be explored more to see how it can be used to benefit semi-literate and illiterate communities in rural areas.

It remains important to note that to this end, there is no research that has been carried out in Botswana with regards to how mobile technology can be used to bridge the digital divide; how the semi-literate and illiterate respond to the use of technology in different situations as well as how mobile phones can be used to represent HIV/AIDS information to the society. This particular research therefore hopes to use lessons learnt from related works to investigate how mobile phones can be used to represent HIV/AIDS information for the semi-literate and illiterate users in rural areas.

2.6 Conclusion An in-depth understanding of the issues surrounding the digital inclusion has been attempted. A clear understanding of what semi-literacy and illiteracy means has been set as well as understanding characteristics that define these people. Issues on why a mobile phone based solution is required for the digital inclusion problem in Botswana were also looked at. The chapter went on to look at the different technologies of representing information on mobile phones. The chapter finally looked at already existing digital inclusion solutions with regard to the defined digital problem so as to analyze their strengths and shortfalls. In order to get the most relevant solution of representing information for the semi-literate and illiterate users out of the identified technologies, the different identified options need to be explored further. It is therefore important to carry out a field survey on the target users in attempt to establish the kind of solution they would expect or they would be willing to have for the identified digital inclusion problem. In pursuing this solution, the following issues remain key:

Morakanyane R.S. MSc.

Page 28

-

Human computer perceptions and psychological behaviour issues

-

Accessibility and use of mobile phones

-

Users’ literacy level – both educational and technical literacy

-

Access to health facilities

The next chapter would therefore be dedicated to field survey and collection of data from target users with regard to the solution they expect in the digital inclusion problem.

Morakanyane R.S. MSc.

Page 29

CHAPTER 3: REQUIREMENTS GATHERING AND ANALYSIS

3.1 Introduction Human Computer Perceptions and Psychological Behaviour issues; Accessibility and use of mobile phones; Users’ literacy level – both educational and technical literacy as well as Access to health facilities have been identified as the key issues to consider when determining the relevant solution for the digital inclusion problem defined in this research. The research will therefore explore these issues through a data collection process. Findings and observations established in this process will then be used to draw conclusions and determine which of the identified data representation methods is more relevant to representation HIV/AIDS information to semi-literate and illiterate users on mobile phones.

3.2 Objectives of Data Collection The general objective of the data collection process was to collect data that will help determine how best to represent HIV/AIDS for semi-literate and illiterate users on mobile phones. This process entailed assessing the following: 

Current situation of target users o Participants profile o The environment in which the participants live in o Education and technology literacy level of participants o Access to health facilities



Use of mobile phone by target users o If target users have mobile phones and if so, how they use their mobile phones on day to day basis o Users’ ability to establish mobile phone calls o Users’ ability to communicate through text messages o Users’ ability to communicate through picture messages o Users’ ability to communicate through voice messages



Perception and psychological acceptance of a mobile phone based solution by target users

Morakanyane R.S. MSc.

Page 30

o Different psychological behaviour issues and human computer perceptions that the target users may have with regard to as well as accessing HIV/AIDS information on mobile phones At the end of the data collection process, the research hopes to propose a well researched way of representing HIV/AIDS information on mobile phones which will be most appropriate to the semiliterate and illiterate members of the society.

Data Collection Instruments Three types of data collection tools were to collect data from participants. These included questionnaires, interviews and observations. The questionnaires were mainly be used to collect participants’ profiles; the interviews were carried out to collect data to do with participants’ perceptions and psychological behaviour while observations were used to determine patterns of how participants react to different uses of mobile phones being presented to them. Other support data collection media that were used include video and still picture cameras. The data collection instrument that was used is attached as Appendix 2 at the end of this report.

General Conduct of the Process The data collection process was conducted as informally as possible, with the participants being coaxed into talking more rather than less. Although the instrument was prepared beforehand, the process was made easy such that it appeared to be normal conversations to an external observer. To ensure that the participants were free and at easy, some participants were interviewed as small groups. This was also a way of reducing possible intimidation due to the low levels of literacy.

The types of questions that were asked in the data collection process were both open ended, close ended and probing questions. The reason why these types of questions were used was so that the discussions remained free and open. For questions that wanted precise answers the participants were asked close ended questions. Where the interviewer wanted the participant to give more detailed information, they asked open ended and probing questions. All participants were asked some questions to establish their level of literacy as guided by the participant’s profile section of the data collection instrument. Priority was given to those with lower levels of literacy as they are a better representation of the target users.

Morakanyane R.S. MSc.

Page 31

3.3 Results and Analysis 3.3.1 Current Situation of Target Users The study aims at proposing a mobile phone based solution for the semi-literate and illiterate members of the society. Previous chapters indicate how these populations (common in rural areas) have inadequate access to information, are intimidated by technology and are mostly the economically challenged. The data collection process was therefore carried out in some rural areas around Botswana which are dominated by these populations.

Sample Environment and Population The data collection process was restricted to two regions in Botswana. These areas are Ngamiland West and Kweneng West regions. These regions were selected based on them being the lowest regions with literacy level, according to Botswana’s Literacy Survey Report, (CSO, 2003). Data was collected in one village or a few villages in each of the regions. The selection of the villages was influenced by the availability of mobile phone networks. Table 1 below is a summary of results obtained from the sample environment and the population:

Table 1: Summary of the Data Collection Sample Environment and Population Region Main Village Kweneng West Letlhakeng Ngamiland West Shorobe Ngamiland West Komana Total NB: 8 hours = 1 day

No. of Participants 67 (40f: 27m) 26 (11f: 15m) 34 (15f: 19m) 127 (66f: 61m)

No. of Hours Spent 32 16 16 64

A total number of 127 data collection instruments were administered in both regions. This number was influenced by the number of people who volunteered to participate in the exercise. A total time not exceeding 35 hours was spent in each region and this number of hours was determined by how easily and relaxed the participants felt about taking part in the exercise.

To gain access into these

communities, Village Development Committees and other village administration bodies were involved. This was so that the communities felt comfortable by working with people they already know and trust.

Morakanyane R.S. MSc.

Page 32

Data Collection Process It is important to note that data collection was also strongly influenced by the contextual and ethnographic design principles (Hayes, 1999). This process was used to ensure that there was intense interaction with the target users and that the real needs, traits and responses of the users were established hence determining what kind of application the target users would be interested in.

In this process, participants were first made to feel comfortable with the researcher while at the same time the researchers made the participants to accept and trust them. For the first few minutes of each interview, the researchers helped the participants to overcome fear of technology and intimidation, particularly those who did not have mobile phones, by letting them to handle mobile phones and acclimatize themselves with how they work.

Each participant or group of participants was given the opportunity to explore if they would be comfortable with accessing information represented as text, as voice or as graphics. The mobile phone owned by the participant played an important role especially in determining if the representation could be graphical. Although economic status of each participant was not determined, the status also played an important role in determining whether the participants would be willing to spend money on the accessing the system after the system has been fully implemented. Besides observing the participants as they were taking the different exercises of the data collection process, the participants were also interviewed on perceptions and psychological acceptance of using the system to access HIV/AIDS information. Figure 4 below is a graphical illustration of these results

Morakanyane R.S. MSc.

Page 33

Figure 4: Results on Current Situation of the Participants

Of the 127 participants that took part in the data collection exercise, 61 were males and 66 were females. These participants were from a wide range of ages so as to get a broader representation of participants of different ages. 72% of the total participants were below the age of 46 while 6% of the participants did not know their ages.

The target participants in this data collection process were the semi-literate and illiterate people who would in turn be the main beneficiaries of the digital inclusion solution being proposed. To determine the literacy level of participants, the participants were assessed on their educational literacy and technical literacy. Educational literacy involved assessing the highest level of education received by participants and participant’s ability to identify counting numbers while technology literacy was determined by assessing if the participant had any ICT device, especially mobile phone. If they had one, researcher determined which type of phone it was as well as the functions of the phone the participant uses more often. The educational literacy assessment was based on the assumption that participants who had had lower levels of formal education were less literate that those who had received more levels of formal education. Ability of participants to identify counting numbers was assessed based on the assumption that those who are able to identify counting numbers stand a better chance of being able to use basic features of a mobile phone regardless of level of formal education received or whether they own a phone Morakanyane R.S. MSc.

Page 34

or not. In the participants that took part in the survey, 21% had not received any form of formal education while 65% had not received any secondary education level. 87% participants were able to identify counting numbers. By these results it was clear that majority of participants were indeed within the target group (semi-literate and illiterate) and that majority of the participants were able to use the most basic feature of mobile phones – dialling on the keypad.

The technical literacy assessment was based on the assumption that generally participants who own a more complex mobile phone, which is also expensive, are more technically literate than those with more basic phones. After all, participants with more complex phones stand a better chance of knowing how to use other complex mobile phone functions than participants with more basic phones. In this case, the complexity of phones was measured in terms of the phone having Multimedia Messaging Services (MMS). On assessing technical literacy, it was established that 70% of the participants had their own mobile phones. 95% of these had mobile phones which could not support MMS.

With 70% of participants having their own mobile phones, it was clear that the penetration supports the motion that mobile penetration is high in Botswana. To get opinion of mobile penetration from the participants, the participants were also asked if they thought the number of people who had mobile phones in their community was more than the number of those who did not have mobile phones. In response to this, 98% thought the number of people who owned mobile phones was more than those who did not own a mobile phone; hence the participants thought the mobile phone based solution to be proposed would generally have users.

3.3.2 Accessibility and Use of Mobile Phone Of the 87% participants who were able to identify counting numbers, 86% could establish a simple call rendering that if the solution to be proposed has a dial in function most of the users should be able to use the function. However the more important factor was to determine whether the participants could actually use the phones as intended, that is, to send and retrieve HIV information to the system to be proposed. Table 2 below is a graphical illustration of the results obtained from testing the participants on usability and interactivity of their mobile phones.

Morakanyane R.S. MSc.

Page 35

Table 2: Results on Usability and Interactivity with Mobile Phones by the Participants Usability and Interactivity No. 1 2 3 4 5 6 7 8 9 10 11 12

Variable Measured Establish a call Create SMS Send SMS Retrieve SMS Read SMS Reply SMS Grouping by Keyword Send Picture Message Retrieve Picture Message Create Voice Message Retrieve Voice Message Use IVR

Responses Ranked in Percentages (%) Below Above Poor Average Average Average Good 5 5 4 17 69 32 9 16 18 17 31 6 7 13 35 17 5 7 13 35 2 22 34 22 2 11 7 18 20 13 28 31 13 12 2 20 17 11 8 2 5 11 11 9 2 11 41 28 13 1 21 14 23 12 6 5 9 24 29 20

In terms of creating and retrieving SMS messages, more people could send and retrieve messages and yet less could create it which is the main function for FAQ input. As long as they could not create the SMS, other capabilities alone are not adequate enough to qualify usage and thus SMS technology would remain a challenge for the majority of the users. Majority of participants could not use a grouping by keyword function as they failed to either read and or interpret responses obtained from sending keyword messages. This means if the solution to be proposed is to have the grouping by keyword function, more users would find it difficult to use. With picture messages, majority of participants did not have mobile phones that supported picture messaging to start with. In addition, participants failed to store or send a picture message when they were provided with a phone that supports picture messaging. Mobile phones with technologies that use graphics, images and icons require special infrastructure may be too complex to be used by the semiliterate and illiterate people.

In observing the use of voice technologies, majority of participants could not confidently create nor retrieve a voice message. However, a significantly large majority (72%) were able to interact with a prerecorded voice messages which are accessed through menu based voice prompts. Menu based representation is a representation in which users traverse through some form of pre-programmed menu in order to access the system. A typical application of a menu based system include the common mobile communications providers’ prepaid service which the customer dials and access a menu to access Morakanyane R.S. MSc.

Page 36

different services such as checking their balance, playing a voice message, recharging their airtime. Therefore an IVR based solution may be relevant for the target users.

3.3.3 Perceptions and Psychological Acceptance

No 1 2 3

Table 3: Results on Perceptions and Psychological Acceptance Variable Measured Responses Ranked in Percentages (%) Willing to Pay Willing: 60%; Not Willing: 23%; Undecided: 17% Privacy and Trust Trusting: 87%; not Trusting: 1%; Undecided: 12% How Often the User Will Access the System Frequently: 4%; Often: 62%; Rarely: 34%

In terms of perception, participants seemed to embrace technology (87% Trust) and agree that having a mobile phone based solution would expedite the process of seeking for medical help. They further assert they would use the service often enough (62%) and would be willing to pay for the service as opposed to the long distance travel to health centres.

In conclusion, participants do welcome the idea of a mobile phone based solution. They own mobile phones, can establish simple mobile phone calls and are able to identify counting numbers on mobile phone. This is a positive indication that participants would be able to use their mobile phones to dial into the digital inclusion solution. However, beside issue of expense, most participants’ phones did not support picture messaging. The option does not seem to be viable as it would benefit only a few users. Plain text messaging is also eliminated as it requires users to be able to create messages, something the majority of the participants could not do. Grouping by keyword is eliminated as it uses keywords and outputs that may be complex for the semi-literate and illiterate people. This leaves only one option which seems to be viable: a voice based solution. However the majority of participants could not create or retrieve voice messages unless they were guided. Most of the participants found it easy to manoeuvre through the IVR system.

Morakanyane R.S. MSc.

Page 37

3.4 General Experiences Gained During Data Collection Differing Setswana Dialects Botswana has differing tone dialects in its differing regions. This eliminates the possibility of a speech recognition solution as different people in the different regions would not express the same concept in the same tone.

Proactive Participants Although it was hoped that it would be more difficult for the participants to participate freely and openly in the data collection exercise, participants tended to be more proactive and more willing to participate further into the research as they viewed the research as a possible solution to some of their challenges.

3.5 Conclusions and Recommendations 

Majority of participants owned mobile phones therefore a mobile phone based solution can be thought to have readily available users. However, the type of solution can not be MMS based as participants did not have mobile phones that support MMS technology. In addition, most participants failed to store or send a picture message when they were provided with a phone that supports MMS. Mobile phones with MMS technologies have therefore proofed to be complex to the semi-literate and illiterate people.



Majority of participants were able to identify counting numbers as well as establish a simple mobile phone call rendering that if the solution to be proposed has a dial in function most of the users should be able to use the function.



Although more participants could send and retrieve SMSes, fewer participants were able to create SMSes. As long as they could not create the SMS, other capabilities alone are not adequate enough to qualify SMS usage and thus SMS technology can not be a viable data representation method for the target users.



Majority of participants could not use a grouping by keyword function as they failed to either read and or interpret responses obtained from sending keyword messages. This means the digital inclusion solution can not have the grouping by keyword function as more users would find it difficult to use.

Morakanyane R.S. MSc.

Page 38



Majority of participants could not confidently create nor retrieve a stored voice message. However, a significantly large majority were able to maneuver through the Interactive Voice Response system with pre-recorded voice messages which are accessed through menu based key tones.



Participants also seemed to embrace technology and agree that having a mobile phone based solution would expedite the process of seeking for medical help. They further assert they would use the service often enough and would be willing to pay for the service as opposed to the long distance travel to health centres.



A menu based Interactive Voice Response solution appears to be a more relevant digital inclusion solution to the target users. The solution may be provided at a cost though it remains important to keep the cost as minimal as possible.

The following chapter will therefore look into proposing the digital inclusion solution in more details. The chapter will discuss conceptual view and architecture of the solution. It will also look into the system requirements in relation to the target users.

Morakanyane R.S. MSc.

Page 39

CHAPTER 4: THE DIGITAL INCLUSION SOLUTION 4.1 Introduction Having identified a menu based interactive voice response system as a more relevant solution to the digital inclusion problem, it is therefore important to take a closer look on what IVR systems are. The chapter will therefore define IVR Systems and how they work. It will also look at IVR system requirements (software, hardware and architecture) as well as functional, non-functional and operational requirements.

4.2 Interactive Voice Response System A menu based IVR system has been selected as the most appropriate solution to the digital inclusion problem. Call Centre Technology (2009) define IVR systems as phone system applications that prompts callers with recorded messages and options and processes voice input and/or touch-phone keypad selections from these menus. IVR scripts then respond to these inputs by providing appropriate information in the form of pre-recorded voice answers or live operator connections. IVR systems therefore provide menu options to the users and the user selects options which will provide them with relevant information to what they are looking for. In the digital inclusion solution, users will be able to make calls using their mobile phones into the IVR system. On receiving the call, the system will play introductory clip to the user and presents different options from the defined menu. The system will then get the option that the user has selected, search it through the knowledge base and play the relevant clip to the user.

Due to this nature of the IVR system, the system therefore requires user-centered approach to design. In this approach, the target users and what they are trying to achieve in the system are the driving force behind the development of the system (Preece et.al. 2002). This design process therefore focused on the target users as well as the tasks they perform. Cognitive and behavioural characteristics of the target users were observed during the requirements gathering process. The target users were also presented with different prototypes of possible solutions. They were allowed to explore these possible solutions while the researcher observed closely the different reactions they made. As a result, this helped the researcher to determine which of the presented options was better accepted by the users as the most appropriate solution to the defined digital inclusion problem. Morakanyane R.S. MSc.

Page 40

From the observations made during the requirements gathering process, the IVR based digital inclusion solution should therefore have special requirements and design principles that would enable the developer to meet the requirement of the target user as well as enable the target users to navigate the system with as less difficulties as possible. The following requirements are defined for the IVR system:

4.2.1 IVR System Software Requirements An IVR application functions in a client/server environment. The IVR program runs on the IVR phone system or on a different server that is processing the application programs. Some of the basic features of IVR programming system environment include the following: Visual Basic, C, C++, Active-X, and .NET Development Tools: offer development environment and libraries that can be used to develop IVR applications. VoiceXML Data Access Tools: enable the orientation of voice-based dialogs for telephone callers that feature the playing of speech prompts using pre-recorded and text-to-speech information, accepting spoken commands via speech recognition and touch tone inputs, and the recording of caller audio information. Phone Key Input: plug-in allow the users to input text using the 0-9, # and * keys of a phone. In this case, users will select the different options from 0-9, #, * on their phones to access information from the system. Text to Speech (TTS) Software extracts variable information from the knowledgebase or as input by the user and converts it to audible speech. Practical applications include the system reading out an option that the user selected on their keypad or reading out stored text messages.

Different vendors, both commercial and open-source have since developed different types of IVR Software. Some examples of the IVR software include the following:

Table 4: Examples of IVR Software No 1. 2. 3. 4.

Software Name Voicent IVR Studio Database Systems Corp. IVR Asterix DialogPalette Microsoft Speech Server

Morakanyane R.S. MSc.

Availability Commercial Commercial Open source Commercial Page 41

4.2.2 IVR System Hardware Requirements IVR systems also require relevant hardware devices in order to enable easy access of the information. Typical hardware devices used in IVR systems are GSM Mobile phone and processors. A normal GSM mobile phone such a NOKIA, SIEMENS, SONY, etc. phones are all appropriate to be used with an IVR system. Users dial these phones to access the IVR system. IVR systems also require efficient processors which are able to retrieve information clips from the database in time. These processors should be able to carry out digital filtering and signal processing in audio recognition. The complete IVR solution therefore comprises of a mobile phone, GSM Network, an IVR system which has logic as well as a database that contains information the user will be asking about. The diagram below shows the different components that make up the conceptual architecture of the IVR system:

Morakanyane R.S. MSc.

Page 42

MOBILE PHONE

GSM NETWORK

IVR

LOGIC

KNOWLEDGE BASE

Information Clips

Source: Author Figure 5: Conceptual View of Components of an IVR System GSM Mobile Phone: This is used by the user to connect to the system through the GSM network. GSM Network: This is the network which the mobile phone operates through. It links the user and the system. Interactive Voice Recognition System: The IVR consist of an E1 transceiver card connecting to the GSM network and the logic. It receives DTMF tones generated by the mobile phone. The IVR also contains the menu options that the user will be accessing Morakanyane R.S. MSc.

Page 43

Logic: The logic receives information from the GSM card about what to do. It will then play static prerecorded menu clips as well as collect information clips from the database and play them. The logic also manages the GSM card. Knowledge Base: The database stores information that the users will be accessing. It should be updated regularly with new information when it is available. Information clips from the database is accessed using SQL commands. Information Clips: The information to be accessed through the system will be classified in different categories. These categories are a representation of information that the community has proved to enquire about on day to day basis.

An IVR system should also have synchronized workflow logic. The flow logic drives the sequence of events being presented to the users for easy manipulation of the system. The figure below shows typical workflow logic of an IVR system.

Morakanyane R.S. MSc.

Page 44

START

USER DIALS

PLAY INTRODUCTION

ENGLISH

SETSWANA

SELECT LANGUAGE

PRESENT MENU

REPLAY MENU INVALID OPTION

SELECT MENU OPTION VALID OPTION

PLAY CLIP

QUIT

STOP

Figure 6: A Typical Workflow Diagram in the IVR System

Morakanyane R.S. MSc.

Page 45

4.2.3 IVR System Architecture Requirements An IVR System also has an architecture made of different components that support the operations of the IVR system. Typical architecture components of the IVR system are as illustrated in the diagram below:

TELECOMMS PROVIDER E.g. BTC

Land line Modem connected to CPU USER

Telephone API

Program files Code e.g. C++ E.g. VXML

Audio Files E.g. .wav

GSM

IVR

LOGIC

Source: Author Figure 7: Typical Architecture Components of the IVR System

Morakanyane R.S. MSc.

Page 46

Knowledge Database E.g. MySQL

Normal Telecommunications Network: The telecommunications network acts like the GSM network. Users dial into the network using the normal direct land line or a mobile phone. Modem Connected to PC: The modem acts like the IVR. It receives signal from the telecommunications network and decode them into signals that the telephone API can understand. The modem inter connects the landline and the personal computer. Telephone API: This accepts voice signals from the modem and passes them to the program Program: The program accepts signals from the telephone API and computes the action to be taken. This serves as the logic of the system. For instance, if the action to be taken is to get a voice clip from the knowledge base, the code does that. It does so the same way the logic would do it. This code can be written in any language, whether C++, C# or Java. Audio Files: These are the voice clips stored in the knowledge base. IVR supports the voice clips of any type.

4.3 IVR System Functional Requirements The functional requirements of the proposed IVR system are sub-divided into two functional processes: User Interactivity and System Interactivity. User interactivity requirements describe how the users will interact with the proposed system in order to access information from the pre-recorded voice clips. System interactivity requirements how describe the system will retrieve the pre-recorded voice clips from the knowledge database and play them to the users on request.

Morakanyane R.S. MSc.

Page 47

4.3.1 User Interactivity The proposed system is expected to have one type of user - the enquirer who is referred to as the user. The figure below shows tasks to be performed by this user, minimum requirements expected from the user as well as the level at which the user will be interacting with the system.

     Requirements      Level of  interaction Tasks

Table 5: User Task Analysis Summary User Log into the system by dialing Navigate through the audio menu by selecting presented options Query HIV AIDS and STDs module Query HIV AIDS and Nutrition module Query Counseling and Support module Should not be dumb Need not be literate Should have a basic GSM mobile phone Should be able to know and identify basic counting numbers Should be able to dial the center where the system lays High: the system will allow the user to select the different menu options and request repeat options where necessary

Pre-condition

 User’s mobile phone can access the GSM network in which the system lays

Postcondition Assumption

 User is listening to the information clip being played to them  The system is available in all GSM Networks in Botswana for free of charge

4.3.2 System Interactivity The proposed menu based IVR system will contain pre-recorded information clips that will be played to the user on request. The figure below shows tasks that the system should be able to perform, the minimum requirements expected from the system and the level at which the system will be interacting with the user.

Morakanyane R.S. MSc.

Page 48

Tasks

Requirements Level of interaction Pre-condition Post-condition Assumption

    -

Table 6: System Task Analysis Summary System Get options chosen by the user Get pre-recorded voice clips as requested by the user Present the voice clips to the user Play pre-recorded voice clips Communicate with a basic GSM mobile phone High

-

GSM Network is up and running

-

The system offers menu options for both Setswana and English

4.4 IVR System Non-Functional Requirement As much as the menu based IVR system has turned to be the most appropriate way of representing information for semi-literate and illiterate users on mobile phones, it is important to ensure that the solution remains user friendly to these users. The following non-functional requirements are taken into consideration when developing the technologies to represent information to the target users:

4.4.1 Recording of Voice Clips All voice clips in the IVR system should be of a homogenous file type to ensure easy manipulation of the clips. This will also make file concatenation easier which provides for more information, especially where more details are required (Lee et.al, 2000). The database that store the voice clips should be a relational database of any type. The database should contain flat files and database manipulation can be carried out using SQL commands. Voice files stored in the database can be of any file format. IVRusability.com (2009) emphasizes the importance of keeping storage formats of IVR voice clips consistent and of a homogenous type. It also advocates for use of the same professional voice-over artist to record the clips. This enhances system acceptance and reduces chances of user confusion.

4.4.2 Nomenclature of Voice Clips Naming of the voice clips should be consistent and easy to follow so that clips can be arranged well hence easy to identify and access.

Morakanyane R.S. MSc.

Page 49

4.4.3 Depth Level of Menu Options Menu levels for IVR Systems should be kept as minimal as possible. Hayes (1999) suggests that menu levels and menu options for IVR systems should not go beyond six levels especially where target users are the semi-literate and illiterate. This is so that there is less cognitive load required by the users when interacting with the system.

4.4.4 User Interface The interaction points between the system and the user will be on the mobile phone keypads, pressing of numbers on the keypad as well as listening to the voice clips being played. In the data collection, users have shown their ability to identify counting numbers on a mobile phone keypad. Numbers on the keypad should therefore be as clear as possible for the semi-literate and the illiterate user to identify. Voice clips should be clear and very audible so that the users do not have to wonder what was being said in the system. The language used in the system should accommodate all users especially the semiliterate and the illiterate. e.g. The pre-recorded voice clips played by the system could be in Setswana words and English numbers. This is because the target users proved to be familiar with counting numbers in English and could easily relate to them.

4.4.5 Security Requirements The proposed system should be an open information center to be accessed by any one who wants to enquire about HIV/AIDS information. Security issues should therefore not restrict users to access the system. However, it remains important to ensure that the system does not become vulnerable to any threats that may perhaps jeopardize its acceptance by the users.

4.5 Operational Requirements 4.5.1 Call Handling The system should be able to handle multiple calls at the same time. It should not take a long time to respond to the users.

Morakanyane R.S. MSc.

Page 50

4.5.2 Length of an Instructing Voice Clip The length of an instructing voice clip should be fairly short. This is to enable the user to remember only a few option items. Short voice clips will also enable faster retrieval of clips from the database. Ivrusability.com (2009) states that appropriate pacing of the system depends on the target user-group, the complexity of the feature set, and the regularity of use. It therefore advocates for user-testing to ensure the best way of determining the length of the clips.

4.5.3 Waiting Time Before Response Alpine Solutions Inc (2002) suggests that a waiting time before response of about 10 seconds should be allowed after presenting menu options to the user. This time is necessary to allow the user, in this case the semi-literate or illiterate time to think and select the option they want from the system. The semiliterate and illiterate are viewed as slow reactors therefore the 10 seconds delay will give provision for their slow reaction. When 10 seconds passes without the user giving feedback into the system, Alpine Solutions Inc (2002) goes on to advice that the clip that was just played will have to be replayed to allow the user to choose their preferred option in case they were delayed. This repeat action should be done three times, that is only when the system can automatically call the exit clip and exit the system. However, a gap of more than 3 seconds in between continuous voice clips can make users think that something has gone wrong (ivr-usability.com, 2009).

4.5.4 Error Handling Error messages should be played every time a user inputs a wrong option. These should also be able to guide the user with what to do next after the user has selected the wrong option.

4.5.5 Consistency Control: Certain standards should be adopted during development to ensure consistency of the system especially the clips that will be played to the users. E.g. * will be used to go back to the previous menu while # will be used to exit the system.

Morakanyane R.S. MSc.

Page 51

4.6 Summary A menu based interactive voice response system has been identified as the most appropriate solution to be developed for the semi-literate and illiterate users to enable them to access HIV/AIDS information with their mobile phones. Because of the nature of this solution, the IVR solution to be developed has to be user-centric. The user-centric approach to design allows the designer to interact with the target users frequently during the design process to allow the developer to meet the identified user requirements. In this design approach it is also important to ensure that the system being designed is tailor-made to meet the needs of the users without making the users to change their way of living to accommodate how the system operates.

The chapter also established that there are many different software that can be used to develop IVR systems. Depending on the target users of the proposed IVR system, there are also many requirements that the developer will need to consider when developing IVR systems. In the digital inclusion problem, it remains important to ensure that the system being developed remains simple and easy to use to the target users. Menu depths and waiting times before response need not over load the users’ memory with having to remember the many options being presented. Error handling messages and consistency control messages should also be in place to ensure that the user finds the system simple to use. The following chapter will describe the design process of the digital inclusion solution taking into consideration the stated requirements. A prototype of this system will also be developed to simulate the real digital inclusion solution.

Morakanyane R.S. MSc.

Page 52

CHAPTER 5: DESIGN AND IMPLEMENTATION

5.1 Introduction Hoffer et.al. (2002) define the design process as the time in the software development life cycle where the description of the recommended solution is converted into logical and physical system specifications. In this process, logical and physical designs of the proposed digital inclusion solution are described; where after a prototype of the solution is attempted.

5.2 Design Considerations The design principles used in developing the proposed digital inclusion solution are the user-centered; high-fidelity and the ethnographic contextual design principles (Preece et.al, 2002). These principles allow the designer to consistently consider the defined requirements as well as interact with the users while designing the proposed solution. With the nature of the solution being proposed, the defined requirements as well as the types of users targeted to use the solution; it is in order to adopt these design principles while developing the digital inclusion solution.

5.3 System’s Functional Design 5.3.1 Data Design The proposed digital inclusion solution should have clearly defined data structures containing HIV/AIDS information that the users will access. It should also have clear menus that the users will navigate with as less difficulties as possible. Design issues for the content and menus are discussed later in the non-functional design section. A typical menu that can be derived from the sample HIV/AIDS information that the users will be accessing is illustrated in the diagram below:

Morakanyane R.S. MSc.

Page 53

INTRODUCTION

HIV AIDS & STDs

HIV AIDS & NUTRITION

FAQs

FOOD ISSUES

STIs

EATING DISORDERS

COUNSELING & SUPPORT COPING WITH HIV PATIENTS SELF ACCEPTANCE

HIV & SEXUAL RELATIONSHIPS

AFTER EATING ISSUES

FIGHTING STIGMA

CONTACTS PLACES

CONTACT PLACES

CONTACT PLACES

Source: Author Figure 8: Conceptual View of the Interactive Voice Responsive System The diagram above shows a typical sample of the digital inclusion system that will provide HIV/AIDS information to the users. The system has three main menus, with each menu having 4 sub-menus. Each sub-menu then has pre-recorded information clips which will be played to the users on request. A detailed description of the sample menu is described later in the content design section.

Morakanyane R.S. MSc.

Page 54

5.3.2 Algorithm Design The figure below is an algorithmic representation of the system. The algorithm illustrates how the different function of the system will be communicating from the time a user makes a call into the system to the time an information clip is retrieved and played to the user. Table 7: System Algorithm Design FUNCTION

PROCESS

USER DIALS

User dials into the system I.e. dials 123456

INTRODUCTORY

Play Introductory Clip I.e. Welcome to the HIV/AIDS Call Center Play LanguageSelectionClip For Setswana Press 1 on your mobile phone; For English Press 2 on your mobile phone; To exit Press # on you mobile phone;

LANGAGE SELECTION

The system waits for a response from the user EVALUATE WHEN ResponseTime > 10 seconds, Replay Language Selection Clip WHEN option selected = 1, Call Setswana WHEN option selected = 2, Call English WHEN option selected = #, Call Exit ELSE Play Invalid Option Clip Replay Language Selection Clip END EVALUATE CALL MAIN MENU For more information on HIV AIDS and STDs Press 1 on you mobile phone; For more information on HIV AIDS and Nutrition Press 2 on you mobile phone; For more information on Counseling and Support Services Press 3 on you mobile phone; To exit the system Press # on you mobile phone;

MAIN MENU

Morakanyane R.S. MSc.

The system waits for a response from the user EVALUATE WHEN ResponseTime > 10 seconds, Page 55

Replay Main Menu Clip WHEN option selected = 1, Call Menu 1 WHEN option selected = 2, Call Menu 2 WHEN option selected = 3, Call Menu 3 WHEN option selected = #, Call Exit ELSE Play Invalid Option Clip Replay Main Menu Clip END EVALUATE; IF no response has been received AND Main Menu has been replayed 3 times THEN Call Exit END IF SUBMENU OPTIONS MENU 1

Call Menu1 Welcome to the HIV/AIDS and STDs Menu; For FAQs Press 1; For Information on STIs, Press 2; For Information on HIV and sexual relationships Press 3; For Information on contact places, Press 4; To return to the main menu, Press *; To exit the system, Press #; The system waits for a response from the user: EVALUATE WHEN ResponseTime > 10 seconds, Replay Menu1 Clip WHEN option selected = 1, Play FAQs Clip WHEN option selected = 2, Play STIs Clip WHEN option selected = 3, Play Sex Relationships Clip WHEN option selected = 4, Call Contact Places WHEN option selected = *, Return to Main Menu WHEN option selected = #, Call Exit ELSE Play Invalid OptionClip Replay Menu1 Clip END EVALUATE IF no response has been received AND Menu 1 has been replayed 3 times THEN Call Exit END IF

Morakanyane R.S. MSc.

Page 56

MENU 2

Call Menu2 Welcome to the HIV/AIDS and Nutrition Menu; For Information on Food Issues, Press 1 For Information on Eating Disorders, Press 2 For Information on After Eating Issues, Press 3 For Information on Contact Places, Press 4 To return to the Main Menu, Press * To exit the system, Press # The system waits for a response from the user: EVALUATE WHEN ResponseTime > 10 seconds, Replay Menu2 Clip WHEN option selected = 1, Play Food Issues Clip WHEN option selected = 2, Play Eating Disorders Clip WHEN option selected = 3, Play After Eating Issues Clip WHEN option selected = 4, Call Contact Places WHEN option selected = *, Return to Main Menu WHEN option selected = #, Call Exit ELSE Play Invalid Option Clip Replay Menu2 Clip END EVALUATE IF no response has been received AND Menu 2 has been replayed 3 times THEN Call Exit END IF

MENU 3

Call Menu3 Welcome to the Counseling and Support Menu; For Information on Coping with HIV Patients, Press 1 For Information Self Acceptance, Press 2 For Information on Fighting HIV AIDS Stigma, Press 3 For Information on Contact Places, Press 4 To return to the Main Menu, Press * To exit the system, Press # The system waits for a response from the user: EVALUATE WHEN Response Time > 10 seconds, Replay Menu 3 Clip WHEN option selected = 1, Play Coping Clip WHEN option selected = 2, Play Self Acceptance Clip WHEN option selected = 3, Play Fighting Stigma Clip WHEN option selected = 4, Call Contact Places

Morakanyane R.S. MSc.

Page 57

WHEN option selected = *, Return to Main Menu WHEN option selected = #, Call Exit ELSE Play Invalid Option Clip Replay Menu3 Clip END EVALUATE IF no response has been received AND Menu 3 has been replayed 3 times THEN Call Exit END IF

CONTACT PLACES MENU

Call Contact Places If you are calling from North East area press 1 If you are calling from the North West press 2 If you are calling from Southern press 3 If you are calling from East press 4 If you are calling from West press 5 If you are calling from Central area press 6 To return to the main menu press * To exit the system press # System waits for a response from the user: EVALUATE WHEN ResponseTime > 10 seconds, Replay WhereAreYouMenuClip WHEN option selected = 1, Play NorthEastClip WHEN option selected = 2, Play NorthWestClip WHEN option selected = 3, Play SouthernClip WHEN option selected = 4, Play EasternClip WHEN option selected = 5, Play WesternClip WHEN option selected = 6, Play CentralClip WHEN option selected = *, Return to Where Are You WHEN option selected = #, Call Exit ELSE Play Invalid Option Clip Replay Where Are You Menu Clip END EVALUATE NorthEast The contact places you can visit in North East are Francistown, Masunga, Ramokgwebana and Kalakamati. Please visit the place nearest to you when you want more information on HIV/AIDS or visit your nearest health center.

Morakanyane R.S. MSc.

Page 58

NorthWest The contact places you can visit in North West are Shakawe, Maun, Gumare and Kasane. Please visit the place nearest to you when you want more information on HIV/AIDS or visit your nearest health center. Southern The contact places you can visit in Southern are Gaborone, Molepolole, Mochudi, Ramotswa, Lobatse, Goodhope, Ramatlabama and Kanye. Please visit the place nearest to you when you want more information on HIV/AIDS or visit your nearest health center. Eastern The contact places you can visit in Eastern are Sefhare, Mahalapye, Lerala, Maunatlala, Bobonong, Mmadinare and Palapye. Please visit the place nearest to you when you want more information on HIV/AIDS or visit your nearest health center. Western The contact places you can visit in Western are Ghanzi, Tsabong, Kalkfontein, Karakubis, Ncojane, Tshane, Hukuntsi, Kang, and Verda. Please visit the place nearest to you when you want more information on HIV/AIDS or visit your nearest health center. Central The contact places you can visit in Central are Serowe, Tonota, Letlhakane, Nata, Sowatown, Sebina, Gweta, Rakops and Tutume. Please visit the place nearest to you when you want more information on HIV/AIDS or visit your nearest health center. CONTINUE?

EXIT

Morakanyane R.S. MSc.

Call Continue To return to the previous menu press *; To exit press #; System waits for a response from the user: IF ResponseTime > 10 seconds THEN Replay Would you like to Continue Clip IF option selected = * THEN Call SubMenu Options ELSE Call Exit Call Exit Thank You for using the IHISM system Exit

Page 59

5.3.3 Software Design On attempting to develop a prototype for the digital inclusion solution, few freely available software from the internet were reviewed. Out of these, Voicent Communication Software (voicent.com, 2009) was selected as the application that will be used to develop the prototype. There was no particular criteria used for selecting this application to be used to prototype the system.

Voicent Communication Software is a commercial telephony and email communication solutions provider which can be used by businesses, organizations as well as private practice. This solution provides technologies for computer telephony and communications. The major component of Voicent which is relevant to the proposed digital inclusion solution is the Voicent Gateway. This component is an Open Voice XML standard development platform for telephony program applications. It allows for the creation of inbound and outbound applications without low level telephony programming. Voicent Gateway is ideal for developing interactive telephony applications such as Auto Reminder and Broadcast by Phone applications. It is based on the W3C VoiceXML standard and it enables interactive voice access to the system from any telephone. Voicent Gateway can be deployed on any PC with a voice modem. The gateway can also supports multiple phone lines by using multiple modems on a single computer. For large scale applications, multiple gateways in different locations can be networked together through the web interface (voicent.com, 2009). The shareware version of this application was downloaded for free from the internet and used to develop the prototype of the digital inclusion solution. Different components of this shareware version which came as a complete package include Voicent Auto Reminder, Voicent Broadcast by Phone, Voicent Agent Dialer and the Voicent Gateway. Other devices required include a data/fax/voice modem; data input device such as microphone; a normal PC and a direct telephone line. Any modem and voice input device could be used but Voicent recommended its own manufactured products for best results.

Voicent Gateway Voicent Gateway can recognize voice over telephone and carry out commands accordingly. It also reads any text to the user over the phone as well as accept input in the form of DTMF tones from the user’s numeric keypad. To develop the graphical user interface for the IVR application the Voicent IVR Studio can be used. This tool enables flexible application development and it supports VoiceXML open standard, DTMF tones, text-to-speech as well as speech recognition. Figure 9 below illustrates how an

Morakanyane R.S. MSc.

Page 60

IVR application is developed using Voicent IVR Studio Shareware Edition. More details of the configuration and setup process are discussed later in the sector of implementing the prototype.

Figure 9: Graphical Illustration of the Voicent IVR Studio for the Proposed Solution After proposing the Voicent IVR system, it remains important to align the prototype with the design of the proposed digital inclusion solution. The figure below shows the design of the different function of the proposed system.

Morakanyane R.S. MSc.

Page 61

USER DIALS

INTRODUCTORY

LANGUAGE SELECTION ENGLISH

SETSWANA

LANGUAGE OPTIONS ERROR MSG

MAIN MENU

ERROR MSG

MAIN MENU

LANGUAGE OPTIONS

SUB MENU

MENU 1

MENU 2

MENU 3

CONTACT PLACES

SUB MENU OPTIONS ERROR MSG

PLAY CLIP

SUB MENU OPTIONS

CONTINUE?

SUB MENU OPTIONS

Source: Author Figure 10: System’s Functions

Morakanyane R.S. MSc.

Page 62

PLAY EXIT

EXIT

5.3.4 Library Specification The library specification shows voice clips that will be store in the system. These clips contain information that the users will be enquiring on. They include among others the following:

Table 8: Sample Clips to be used in the System Introductory Clip Main Menu Clip Menu1 Clip STIs Clip Contact Places Introductory Clip North East Clip Southern Clip Western Clip Menu2 Clip Eating Disorders Clip Menu3 Clip Self Acceptance Clip Would You Like To Continue Clip

Language Selection Clip Invalid Option Clip FAQs Clip Sexual Relationships Clip Where Are You Menu Clip North West Clip Eastern Clip Central Clip Food Issues Clip After Eating Issues Clip Coping Clip Fighting Stigma Clip Thank You Clip

5.3.5 Storage of Voice Clips The voice clips recorded will be of file type .wav as this is the most common data type used in Voicent IVR applications. Extra voice clips to be concatenated with the main clips will not be recorded. The knowledge base where the voice clips are stored will be a built-in Java Action Database and MySQL as the relational database.

The voice clips that will be accessed from the system are both in English and Setswana to allow users to select the language they will be comfortable with. Although Setswana is more relevant to the target users – semi-literate and illiterate, English language is included for any other user who may want to user the system who would be comfortable with the language.

Morakanyane R.S. MSc.

Page 63

5.3.6 Example Name Description Participating Actors Pre condition Entry Condition Flow of Events

Post Condition Impossible flows Interaction Special requirements

Morakanyane R.S. MSc.

Table 9: A typical data input activity of the system Enquire about Fighting HIV/AIDS Stigma This use case will be used to receive a query from the users User The user has already dialed into the system User has initialized the system by dialing into the system 1. The IVR system presents the ‘introduction’ clip and the ‘main menu’ options to the user through voice 2. The user selects option 3 i.e. the ‘Counseling and Support’ option by pressing 3 on their keypad 3. DTMF tone signals generated by the keypad are sent to the logic 4. The logic directs the signals to the knowledge base to pick the relevant sub-menu clip 5. The knowledge base sends the clip back to the logic 6. The logic sends the clip to the IVR system 7. The IVR system presents the sub-menu options under the ‘Counseling and Support’ option 8. The user selects option 3 i.e. the ‘Fighting HIV/AIDS Stigma’ sub option 9. Steps 3 to 7 are repeated, this time with the actual information clip, not sub-menu option. 10. After playing the clip to the user, the IVR system plays the exit clip 11. User presses * on their keypad 12. DTMF tone signals are sent to the logic then to knowledge back to pick good bye message and hang up the call. The system plays the how to ‘Fight the HIV/AIDS Stigma’ clip to the user The system cannot accept an option that does not exits in the system menu The user selects an option from the system The user has to pay attention not to supply incomplete or irrelevant information. The user must not take too long before they respond to the system The user must use the required format.

Page 64

The example above is a use case showing a series of events that will take place when a user dials into the system requesting for information on Fighting the HIV/AIDS Stigma. The system should therefore be able to play all other clips when requested by the user. The system should also respond to wrong input by playing error message clips when the user has selected an invalid or incorrect option.

5.4 System’s Non-Functional Design 5.4.1 Content Design Content obtained from the proposed digital inclusion solution will be providing information on HIV/AIDS issues. The specific topics that are covered in the system were selected randomly from the Ipoletse Training Manual (MRI, 2002). The selected topics are used as an illustration of how they system will work. Different or more topics could be added to increase the knowledge base of the system. It is therefore important to note that this research is concerned more with understanding the technology of the system rather than developing a full implementation of the solution. Below are typical categories of information that have been selected for the purpose of this particular research which users may request from the system.

HIV/AIDS and Sexually Transmitted Diseases: This category includes information on frequently asked question on HIV/AIDS such as what it is, how it is contacted, how to avoid contracting it and so on. It will have information on other Sexually Transmitted Infections (STIs) and how they relate to HIV/AIDS. Information on HIV/AIDS and sexual relationships is also available and the last sub option is the contact place where the patient will be asked where they are calling from and the physical information center nearest to where they are calling from will be provided.

HIV AIDS and Nutrition: This includes information on food issues relating to HIV/AIDS. Issues such as how to handle food for the HIV infected and how the HIV infected should handle food, how to prepare healthy food for the HIV and how to store the food are covered. It also includes information on dealing with challenges of eating disorders in people living with HIV/AIDS such as low appetite and swallowing problems. Information on after eating disorders such nausea, vomiting and diarrhea is also accessible. The option will also end with a contact places sub option. Morakanyane R.S. MSc.

Page 65

Counseling and Support: This includes information which patients may enquire on issues such as coping with HIV/AIDS patients in homes, positive living and self acceptance and fighting HIV/AIDS stigma in the community at large. The option also ends with a contact places sub option.

5.4.2 Menu Levels When creating a menu of options and menu levels the number of options to select from will not be more than 6 to reduce the cognitive load of the users. The prototype to be developed will only have 3 menu options and 3 sub-menus as shown in figure 8 above. To avoid having long menu options, some of the attributes will be grouped together. For instance, contact places will be group according to closeness of geographical area.

5.4.3 Audibility The importance of a clear audible voice when developing IVR systems cannot be over emphasized. Since the system to be developed is only a proof of concept prototype, a professional voice-over artist will not be used. However, the developer will try as much as possible to use a clear and audible voice. Also, although Voicent IVR application performs better when recorded using Voicent recommended devices, this was not possible as there was no budget to purchase a complete Voicent suite for developing the prototype.

5.4.4 Waiting Time before Response Voicent Gateway and IVR Studio allows for time frames to be configured. Therefore on configuring these components, a waiting time before response of about 10 seconds will be accommodated after presenting menu options to the user. This has been illustrated clearly by the extract of the algorithm below. With this, the semi-literate and illiterate should be able to use the system freely. Algorithm Extract: … The system waits for a response from the user IF ResponseTime > 10 seconds THEN Replay Language Selection Clip IF option selected = 1 THEN Call Setswana IF option selected = 2 THEN Call English IF option selected = # THEN Call Exit ELSE Play Invalid Option Clip Replay Language Selection Clip END IF … Morakanyane R.S. MSc.

Page 66

5.5 Implementation of the Ideal Digital Inclusion Solution This section describes the implementation of the ideal Digital Inclusion Solution in a real environment. It is important to note that this research has limited itself to a small prototype of the sample solution due to unavailability of the relevant resources required to fully implement the solution.

5.5.1 Configuration and Installation Process Ideally the digital inclusion solution is meant to be rolled out as a national project to enable communities, especially the semi-literate and illiterate communities in rural areas to access HIV/AIDS information from this system. The components of the ideal solution should therefore be of very high quality and standard to support nationwide requests. A typical ideal IVR system can be configured with Microsoft Speech Server (Microsoft, 2007).

To develop this ideal digital inclusion solution the incremental prototype method should therefore be used. This method allows the developer to develop and test one function of the system at a time before moving to the next function, hence ensuring thorough testing of the different system components (Sommerville, 2001).

5.5.2 Testing The testing process of the ideal digital inclusion solution can be looked at in different phases. During the development process, the developers can carry out unit tests which involve testing of an individual function which is being developed. This allows the developer to see if the system meets the requirements as were stated in the system requirements while still developing the function. After completing the function and being satisfied with the functionality, the developer may carry on into the next function which will then be integrated with the completed function. Tests carried out at this level are called system tests.

After integration of all functions, the developer may also test the integrated components as one entity to see if they cohere. When all components have been consolidated into a single system, then the developer can carry out a complete system test. At this stage the system will now be ready to be rolled out to the users. The tests that are carried out at this stage are the acceptance tests which check the system against the user requirements. This test is similar to systems testing in that the whole system is checked but the difference is that: systems testing checks that the system that was specified has been delivered while Morakanyane R.S. MSc.

Page 67

acceptance testing checks that the system delivers what was requested (Sommerville, 2001). This test is carried out by the users and not the developer. When all the tests have passed the developer can then develop a monitoring and maintenance plan that will be followed to ensure that the system keeps running all the time and it meets the requirements as expected.

5.5.3 User Training Since the ideal digital inclusion solution is intended to be rolled out for national usage, all components of the system need to be fully documented and thorough user training carried out. This documentation should be able to provide maintainers and further developers of the system will references to use when maintaining or developing further. Manuals explaining all processes of the system should also be prepared with acceptable standards of documentation (ISO 9000; 2005). User training on the other hand is also very essential. The developing team would need to develop a brand for the system, prepare a training plan and go out to the people to teach them how to use the system. A national campaign for the system through liaising with media, as well as seeking political buy-in would create a significant mileage for the system. This process should make user see the system as an everyday tool that should be part of their daily lives. The process should also involve teaching the users how to access the different menu options provided by the system and should ensure that users are informed about all changes that may occur in the system at all times.

5.5.4 System Location For easy maintenance and support ideally the system needs to be in a place where all the resources needed for maintenance and support are easily accessible. Such places can be one of the major health centers, either a hospital or a research center which will have resource to monitor the system’s progress. Dedicated developers and contact managers should also be available to keep developing and updating the system as well as improving on the content of the knowledge base.

5.6 Implementing the Prototype To implement the prototype of the proposed menu-based IVR system using Voicent IVR Studio Application, an intense study for how to deploy this application was carried out from the Voicent Software website. The process of implementation starts will firstly downloading the main Voicent Morakanyane R.S. MSc.

Page 68

Application and setting it up. Platform and all other requirements for this application are provided in the set-up manual.

After setting up the devices and downloading the application, the next step was to develop the actual IVR Application. The process involved downloading the Voicent IVR Studio Application, Installing it, Designing the Call Flow and Deploying the IVR Design as illustrated in Appendix 3 (Voicent, 2009). After successfully setting up the IVR Studio, the application was accessed by calling it through a fixed line to which the modem is connected. The fixed line number dialed to access the prototype of the system was: 3935983.

The following diagrams are snap shots of the different configuration step that were carried out when setting up the application.

Figure 11: Configuring Text-To-Speech (TTS) Commands The knowledge base where the voice clips are stored is a built-in Java Action Database with MySQL database. The system presents computer generated audio prompts using Voicent Natural TTS software where the designer had typed the words that the system reads as voice prompts to the user of the system. Morakanyane R.S. MSc.

Page 69

As the designer continues with deploying the application, they need to keep running the design checker to see if the configuration does not have any errors. Figure 12 below shows output of a successful design check screen.

Figure 12: Check Design Output The ones and twos on the side indicate the keypad option to be selected by the user when they want to access the clip next to it. E.g. on the HIV/AIDS Sub menu: Press 1 for FAQs Press 2 for Information on STIs

5.7 Challenges Encountered in Developing the Prototype The shareware version of Voicent is a free download of the Voicent IVR Application. This therefore blocks other components that may be required to develop a more complete system. For instance,

Morakanyane R.S. MSc.

Page 70

shareware edition limits the developer to develop up to only 2 levels of menu options. However, a caller can call the system using any medium, be it a cell phone or a fixed line.

The shareware also includes a pre-programmed voice clips which serves as a marketing clip for the product. This clip cannot be removed and will be played before the actual system is played every time when the system is being accessed. To remove this clip required buying a license for the other versions of the application (Professional or Enterprise Editions).

Recording of clips was also a challenge. To achieve clearly recorded clips, Voicent requires that the recording tools (modem; microphones; etc) be products of Voicent. Other devices are supported although the clips are not as clear as those obtained when the Voicent equipment was used.

The system also needs to be tested on the target users (semi literate and illiterate) so as to get a true picture of the usability of the system. This was not carried out as the system that was developed is just a simple prototype to simulate the real application, also because there was no budget for the project.

It is also important to note that the number dialed for accessing the system was not a leased toll free number. The developer had to bare costs of calling the number from a mobile phone for every trial call which was made. The system also does not cater for callers outside Botswana’s 267 code.

Morakanyane R.S. MSc.

Page 71

CHAPTER 6: CONCLUSION AND RECOMMENDATIONS 6.1 Conclusion Botswana as a country hard hit by the HIV/AIDS pandemic and low literacy rate especially in rural areas faces challenges of disseminating information to all communities in time. The most affected are the semi-literate and illiterate communities who also have limited access to other social services. The key contributing factors identified as the course of this problem are that these populations are mostly found in rural areas which unfortunately are the least served areas in terms of most social services. There are inadequate health outreach programs to these areas. There is also a serious shortage of health professionals. Limited resources and poor infrastructure in the rural areas are also contributing factors. As a result health information reaching these areas is either limited or is not represented in a manner in which these populations (semi-literate and illiterate) can easily absorb. It has also been established that Botswana on the other hand has a high mobile penetration which continues to grow year after year. Considering the high mobile penetration in Botswana, this research was therefore aimed at trying to explore how mobile technology can be used to bridge this digital divide. It did so by looking at how information can be represented for the semi-literate and illiterate populations on mobile phones, such that they can benefit from getting health information at their earliest convenience. From this, a survey was carried out where it was discovered that mobile technology is widely used and accepted in the society as indicated by the high numbers of mobile phone subscribers. It was also established that this technology is common to everyone regardless of level of literacy. Comparisons were then carried out on different ways of representing information on mobile phones. These were carried out using the semi-literate and illiterate users. To this, interactive voice response technology emerged as the most relevant way which could be used to represent information to cater for the needs of these communities, taking into consideration their level of literacy and their psychological perceptions towards use of technology. The results from these comparisons have been published in a paper which has been accepted at the Prato Community CIRN 2009 Conference, Monash University Prato Centre, Italy (See Appendix 4).

The research then went on to explore the different requirements of an IVR system taking into consideration the target users. A prototype of the IVR based solution was then designed and developed

Morakanyane R.S. MSc.

Page 72

using a free download of a commercial software – Voicent, which is used for developing IVR applications.

6.2 Recommendations From the findings and the developed prototype one can confidently recommend a voice technology based system to be developed as a way of representing information to the semi-literate and illiterate. The system can use DTMF tones as was designed in the prototype. To allow for full utilization of the system by the target users the system should use as minimal costs as possible. This can be ensured by asking for a toll-free number which all mobile network communication lines can call.

Voicent Software which was used to develop the digital inclusion solution prototype can be explored further in attempt to develop a more robust digital inclusion solution. Fund permitting, better versions of the software can be acquired and used to develop a more complete solution.

Testing of the already developed prototype can also be carried out to see how users will respond to this system. Although users showed some comfort with using other IVR based solutions during the data collection process, it still remains important to test the application with the users with the developed digital inclusion solution on HIV/AIDS information.

For further studies, other researchers can explore other means of disseminating HIV/AIDS information to the semi-literate and illiterate communities such as use of automatic reminders for care givers that need to take their patients to the hospitals, or take medication.

Also as further studies, it is recommended that options of representing HIV/AIDS information for the semi-literate and illiterate users using picture messaging should be explored further. Desk studies showed that graphical representation is the most appropriate way of representing information for the semi-literate and illiterate users, only it was not explored in this research due to costs incurred with using mobile phones, especially by the data collection participants.

Morakanyane R.S. MSc.

Page 73

REFERENCES 1. World Health Organization, (2005) HIV/AIDS (WHO/UNAIDS); United Nations. 2. Central Statistics Office (2005) Botswana AIDS Impact Survey II Statistical Report 2004 Central Statistics Office, Republic of Botswana. 3. President Festus Mogae, (2001) UN General Assembly Speech, United Nations. 4. Botswana Technology Center (2004), Telemedicine in Botswana Feasibility Study Report, Botswana Technology Center. 5. UNESCO (2004) Botswana Literacy Survey: 2003; UNESCO 6. Central Statistics Office (2002) The Central Statistics Office Population Survey Report of 2001; Central Statistics Office, Republic of Botswana 7. Botswana Telecommunications Authority (2009) Market Study of the Telecommunications and ICT sector in Botswana; Analysys Mason Limited, London, UK. 8. Joshi and Rasal, (2002) Interaction Paradigm for Semi-literates and Illiterates; Industrial Design Center IIT; Powai, Mumbai. 9. United Nations Development Programme (2005) Botswana Human Development Report: Harnessing Science and Technology for Human Development UNDP, Gaborone, Botswana. 10. Thatcher, A. et.al, (2002) Attitudes of Semi-literate and Illiterate Bank Account Holders to the Use of Automatic Teller Machines (ATMs); Psychology Department, University of the Witwatersrand; South Africa. 11. Tamocha, M. Telecommunications in Botswana; Mmegi Monitor Newspaper, 29 January 2007. 12. Orange Botswana (2008) Fact Book; Orange Botswana, Gaborone, Botswana. 13. Mascom Wireless (2008) Fact Book; Mascom Wireless, Gaborone, Botswana. 14. Parikh, T.S. (2006) Mobile Phones may be the Right Devices for Supporting Developing World Accessibility, but is the WWW the Right Service Delivery Model?, Edinburgh, UK. 15. Voicent Communication Software http://www.voicent.com/index.php [Accessed 14 February 2009] 16. Microsoft Corporation, (2005) Microsoft Speech Server 2004 R2, Management Pack for MOM Documentation, Microsoft Corporation 17. MedRescue Botswana, (2002) Ipoletse HIV/AIDS Call Center Operators Training Manual. MedRescue Botswana.

Morakanyane R.S. MSc.

Page 74

18. Van Niekerk .J, Meyer M. (1999) Managing Human Resource Development: An Outcomes Based Approach, Durban, Butterworths. 19. Chlebowska. K, (1990) Literacy for Rural Women in the Third World, UNESCO, Belgium. 20. Baecker. R. M. and Buxton. W. A. S, (1987) Readings in Human Computer Interaction: A Multidisciplinary Approach, Morgan Kaufmann Publishers, California. 21. BTA (2008). 22. Interactive Voice Response http://www.call-center-tech.com/interactive-voice-response.htm. [Accessed 21 March 2009]. 23. Lee, C.H. et.al. (2000) “On Natural Language Call Routing.” Speech Communication, 31(4): 309-320. 24. Hayes, A. (1999) The Interactive Response System and Kiosk Technology: Research Paper, Corporate Services Reform, New South Wales. 25. International Organization for Standardization (2005) ISO 9000; 2005 Quality Management Systems. ISO. 26. IVR Usability. http://www.ivr-usability.com. [Accessed April 2009]. 27. IVR Guidelines. http://ivr-usability.com/guidelines.htm [Accessed April 2009]. 28. Preece, J. et.al. (2002) Interaction Design: Beyond Human Computer Interaction, John Wiley & Sons Inc, USA. 29. Hoffer, J. et.al. (2002) Modern Systems Analysis and Design 3rd Ed, Prentice Hall, USA. 30. Sommerville, I (2001) Software Engineering 6th Ed, Addison-Wesley, England. 31. Joyce, M. et.al. (2007) Prospects of e-Advocacy in the Global South, Creative Commons Attribution. 32. Medhi I et.al. (2007) “Text-Free User Interfaces for Illiterate and Semiliterate Users” MIT Press; 4(1): 37-50 33. Kuun, C. et.al. (2007) OpenPhone Telephony System For HIV/AIDS Care-givers, Meraka Research Institute HLT Group, RSA. 34. Nokia Research Center (2005) Understanding Non-Literacy as a Barrier to Mobile Phone Communication, Nokia. 35. Common IVR Mistakes. http://www.angelmedicalsys.org/VrMistakes.htm [Accessed April 2009] 36. History of Mobile Phone. http://www.mamboservis.com/ [Accessed April 2009] 37. Microsoft External Research and Programs (2005) Digital Inclusion through Mobile and Wireless Technologies Research Funding Initiative Microsoft Corporation. Morakanyane R.S. MSc.

Page 75

38. Ramos, A.J.O. (2005) The Viability of Mobile SMS Technologies For Non-Formal Distance Learning in Asia; Makati City, Philippines. 39. Parikh, T.S. and Lazowska, E.D. (2006) Designing an Architecture for Delivering Mobile Information Services to the Rural Developing World; Seattle, WA, USA; 40. Butts, L., Cockburn, K. (2002) “An Evaluation of Mobile Phone Text Input Methods Procedure” Third Australasian User Interface Conference Pp. 55-59. 41. Arison, D. and Pries-Heje, J. (2005) Research in Information Systems - A Handbook for Research Supervisors and their Students, Butterworth-Heinemann Information Systems Series, Great Britain. 42. Sherwani, J. (2006) Are Spoken Dialog Systems Viable for Under-served Semi-literate Populations Carnegie Mellon University; Pittsburgh, USA 43. Curran, K., Woods, D. and Riordan, B.O. (2006) Investigating Text Input Methods For Mobile Phones; Northern Ireland, United Kingdom 44. Reid, F.J. and Reid, D.J. (2004); “Text Appeal: The Psychology of SMS Texting and Its Implications for the Design of Mobile Phone Interfaces” Campus-Wide Information Systems; Emerald Group Publishing Limited; 21(5): 196 – 200. 45. Mutula S. M. (2004) “Making Botswana an Information Society: Current Developments” The Electronic Library 22(2): 144 - 153 46. Gröschel J. et.al. (2002) Automated Speech Recognition For Time Recording In Out-Of-Hospital Emergency Medicine - An Experimental Approach; Institut für Medizinische Biometrie und Informatik, Medizinische Fakultät, Universität Heidelberg; Germany. 47. The Midweek Sun Newspaper; Issue of May 23 2007 48. Herbert, M. (1990) Planning a Research Project, Cassell Educational, London, UK 49. Dawson, W. D. (2000) The Essence of Computing Projects- A Student’s Guide, Pearson Education Limited, Essex, UK. 50. Dennis, A. et.al (2005) System Analysis and Design with UML Version 2.0; 2nd Ed, John Wiley and Sons Inc, USA.

Morakanyane R.S. MSc.

Page 76

APPENDICES Appendix 1: Botswana’s Literacy Survey Report Appendix 2: Data Collection Instruments Appendix 3: Voicent Manual Appendix 4: Conference Paper accepted at the Prato Community CIRN 2009 Conference

Morakanyane R.S. MSc.

Page 77

Appendix 1: Botswana’s Literacy Survey Report District Adult (15 years +) Literacy Rate by Sex Survey Population Aged 15+

Literate Population Aged 15+

District Literacy Rate (percent)

District

Male

Female

Total

Male

Female

Total

Male

Female

Total

Ngwaketse

33,926

46,786

80,712

26,280

36,788

63,068

77

79

78

Barolong

10,079

15,265

25,344

8,480

12,845

21,325

84

84

84

Ngwaketse West

2,606

3,830

6,436

1,738

2,790

4,528

67

73

70

South East

18,761

20,553

39,314

15,901

17,691

33,592

85

86

85

Kweneng East

50,875

59,754

110,629

39,049

50,150

89,199

77

84

81

Kweneng West

8,168

11,005

19,173

5,191

5,863

11,054

64

53

58

Kgatleng

20,056

24,081

44,137

14,081

19,149

33,230

70

80

75

Central Serowe/Palapye

42,073

55,450

97,523

32,463

42,137

74,600

77

76

76

Central Mahalapye

28,765

34,968

63,733

20,063

26,146

46,209

70

75

73

Central Bobonong

19,344

23,959

43,303

14,441

17,877

32,318

75

75

75

Central Boteti

15,344

15,571

30,915

11,213

10,830

22,043

73

70

71

Central Tutume

37,996

47,107

85,103

29,645

36,245

65,890

78

77

77

North East

11,020

11,220

22,240

9,658

9,715

19,373

88

87

87

Ngamiland East

18,852

24,402

43,254

16,993

21,304

38,297

90

87

89

Ngamiland West

11,656

14,933

26,589

7,709

9,752

17,461

66

65

66

Chobe

4,259

5,648

9,907

3,613

4,768

8,381

85

84

85

Ghanzi

8,147

8,437

16,584

5,239

4,694

9,933

64

56

60

Kgalagadi South

8,172

10,295

18,467

5,019

6,963

11,982

61

68

65

Kgalagadi North

3,831

3,812

7,643

2,674

2,828

5,502

70

74

72

Total

470,626 576,543 1,047,169 377,895

471,023

848,918

80

82

81

Morakanyane R.S. MSc.

Page 78

Appendix 2: Data Collection Instruments Data Collection Instrument Questionnaire Table 1 below is the sample questionnaire that will be used to collect participants’ profiles, their literacy levels, basic mobile phone usage as well as health services usage. Table 1: Sample Questionnaire PARTICIPANT’S PROFILE Name Date Of Birth Village Highest Level Of Education Received Cell Phone Make Mobile Phone Network Provider Does Phone support Picture Messaging

LITERACY AND MOBILE PHONE USAGE YES/NO Ability to identify counting numbers? Ability to send and receive SMSes? Ability to leave and access a voice message? Ability to send and receive picture messages? Would you think it is a fair assumption to say that most people in your neighbourhood have cell phones?

ACCESS TO HEALTH SERVICES How often do you visit the health facility? How far is the closest health facility from where you stay? What do you do if you have a health matter arising that you want to enquire about? Would you go to the nurse’s house at night to go and ask any arising issues? Would you think increasing the number of working hours in the health center will improve solve the problem?

THANK YOU FOR PARTICIPATING!!!

Morakanyane R.S. MSc.

Page 79

Interviews Table 2 and 3 below are samples of the interview questions used to collect participants’ perceptions and psychological behaviour. The different issues will be explained to the interviewee first before the interview is carried out. Table 2: Sample HCI Interview Questions No

1. 2. 3. 4. 5. 6. 7. 8.

Question: Which communication technology do you use most in your mobile phone? How often do you use the technologies? Where have you come across such technologies? E.g. local mobile phone providers, etc. Technology SMS; MMS Voice; Plan text messages Grouping by keyword Interactive Voice Response Video Menus

Usage

How Often

Table 3: Sample Psychological Behaviour Interview Questions No 1.

Question Would you be willing to spend money to make cell phone calls to make enquiries on HIV/AIDS information

2.

Would you accept it if you were to call a number on your cell phone to get information on HIV/AIDS?

3.

Would you trust the HIV/AIDS information you obtained from the cell phone?

Morakanyane R.S. MSc.

Responses

Page 80

Came Across

4.

Efficiency: how long would you expect the response from the system

5.

How would you feel about your privacy when accessing the system

6.

Accessibility: How often would you like to access the system

7.

Interactivity: how would you like to interact with the system

8.

Effectiveness: how available would you like to the system to be

Observation: all tasked will be assign a rank between 1 and 5 Basic Usage The purpose of carrying out these observations is to establish if the participant is able to use basic mobile phone features. The exercise will be able to help determine if the participant qualifies to be categorized as semi-literate or illiterate. EXERCISE 1: Ability to establish a mobile phone call; Ability to identify counting numbers; Ability to use mobile phone keypad NUMBER

ESTABLISH CALL

A IDENTIFY COUNTING USING KEYPAD NUMBERS

71894236 72439015 73524891 COMMENT

Morakanyane R.S. MSc.

Page 81

EXERCISE 2: Ability to identify numbers in the phone directory NUMBER

MANIPULATE THE IDENTITY THE NUMBER DIRECTORY

Resego 72172216 Masizana 717548625 Mpoeleng 72214698 COMMENT

EXERSICE 3: Ability to change phone settings Ability to manipulate the different menus with in the phone SETTINGS TO BE MANIPULATING CHANGING PHONE SETTINGS CHANGED MENUS Ring Tone Date and Time Set Reminder / Alarm Change Profile COMMENT

Text-Based Usage The purpose of carrying out these observations is to establish in more detail if the participant is able to use Text-Based Messaging technologies in their mobile phones. Full-Text SMS and Grouping by Keyword have been identified as some of the text based technologies that can be used to represent data for the semi-literate and illiterate users on mobile phones. The exercise will therefore be able to help determine if Text-based technology will be a relevant data representation technology that can be used for the semi-literate or illiterate users and if it is relevant, which is more appropriate between Full-Text SMS and Grouping by Keyword. Exercise 1: Ability to create a message The participant will be asked to create the following messages and send them to the number 72172216: MESSAGE TO CREATE

CREATING THE SMS

Tshedimosetso ka dijo tse di tshwanetseng go fiwa batho ba ba nang le mogare wa HIV Tshedimosetso ka go ikamogela seemo fa o na le mogare wa HIV Tshedimosetso ka ditla morago tse di ka go tlhagelang fa o sena go tsaya kgotsa go fa Morakanyane R.S. MSc.

Page 82

SENDING THE SMS

molwetsi diritibatsi tsa mogare COMMENT

The participant will be asked to send SMSes to the number 72172216 to ask for information on: - Food that should be given to people with HIV - Self-acceptance to people with HIV - How to respond to any side effects or after math that can take place after giving or taking medication Exercise 2: Ability to retrieve and read a message An SMS will be sent to the participant and the participant will be asked to retrieve and read it. The SMS to be retrieved and read will be:

MESSAGE TO BE READ

RETRIEVING THE SMS

READING THE SMS

Motho yo o nang le mogare wa HIV o tshwanetse go fiwa dijo tse di nang le dikotla tsotlhe tsa mmele mme e bile di le phepa. Sekai sa dijo tse a ka di fiwang e ka nna bogobe jwa mabele, morogo le nama, mme morago a fiwe namune kgotsa apole. Maungo a naga jaaka maretlwa le marula le one a siametse botsogo jwa motho yo o nang le mogare wa HIV. Fa o santse o na le potso e nngwe mabapi le dijo tse di fiwang banale mogare, o ka romela molaetsa o mongwe. COMMENT

Exercise 3: Ability to reply a message The participant will be asked to reply the retrieved message asking further what food items to be given to the HIV infected. The follow up question will be: MESSAGE TO BE TYPED AS TYPING THE REPLY REPLY SMS

SENDING THE SMS

A motho yo o nang le mogare wa HIV a ka nwa bojalwa? COMMENT

In this question the participant will be asked to reply to the just received SMS, enquiring if people with HIV can take alcohol. Exercise 4: Ability to use SMS info from local Mobile providers Morakanyane R.S. MSc.

Page 83

SMS info service from local mobile providers will be explained to the participants. The participants will then be asked to use the service to retrieve information. Orange subscribers will be asked to send their keyword to 160 SMS INFO SERVICE TO BE USED

KEYWORD TO TYPE

NEWS STOCKS WEATHER

bNEWS bihl bwWEATHER

USING SMS INFO SERVICE

COMMENT

Picture Based Usage The purpose of carrying out these observations is to establish in more detail if the participant is able to use MMS technology. Graphical representation and picture messaging have been identified as some of the technologies that can be used to represent data for the semi-literate and illiterate users on mobile phones. The exercise will therefore be able to help determine if MMS technology will be a relevant data representation technology that can be used for the semi-literate or illiterate users. Exercise 1: Ability to send picture messages The participant will be asked to send a picture message to the numbers below PICTURE MESSAGE TO BE SENT

NUMBER SENT TO

Pic 1 Pic 2

72172216 71458564

SENDING MESSAGE

THE

PICTURE

COMMENT Exercise 2: Ability to retrieve picture messages Picture messages will be sent to the participant’s phone and the participant will be asked to retrieve the picture message PICTURE MESSAGE TO BE RETRIEVED

RETRIEVING THE PICTURE MESSAGE

Pic 1 Pic 2 COMMENT

Morakanyane R.S. MSc.

Page 84

Exercise 3: Ability to store a picture message After retrieving the picture messages, the participant will be asked to store the picture message PICTURE MESSAGE TO BE STORED

STORING THE PICTURE MESSAGE

Pic 1 Pic 2 COMMENT

Voice Based Usage The purpose of carrying out these observations is to establish in more detail if the participant is able to use Voice technologies. Voice has been identified as one of the technologies that can be used to represent data for the semi-literate and illiterate users on mobile phones. The exercise will therefore be able to help determine which voice technologies will be relevant data representation technologies that can be used for the semi-literate or illiterate users.

Exercise 1: Ability to CREATE a voice message The participant will be asked to send voice message to the numbers below VOICE MESSAGE TO BE CREATED

NUMBER TO

Tshedimosetso ka dijo tse di tshwanetseng 72172216 go fiwa batho ba ba nang le mogare wa HIV A motho yo o nang le mogare wa HIV a ka 71485462 nwa bojalwa?

COMMENT

Morakanyane R.S. MSc.

Page 85

SENT CREATING THE VOICE MESSAGE

Exercise 2: Ability to retrieve a voice message Voice messages will be sent to the participant’s phone and the participant will be asked to retrieve the messages and repeat it by saying it out loud

VOICE MESSAGE TO BE RETRIEVED

RETRIEVING THE VIOCE MESSAGE

REPEATING THE MESSAGE

Motho yo o nang le mogare wa HIV o tshwanetse go fiwa dijo tse di nang le dikotla tsotlhe tsa mmele mme e bile di le phepa. Sekai sa dijo tse a ka di fiwang e ka nna bogobe jwa mabele, morogo le nama, mme morago a fiwe namune kgotsa apole. Maungo a naga jaaka maretlwa le marula le one a siametse botsogo jwa motho yo o nang le mogare wa HIV. Fa o santse o na le potso e nngwe mabapi le dijo tse di fiwang banale mogare, o ka romela molaetsa o mongwe.

COMMENT

THANK YOU FOR PARTICIPATING IN THIS EXERCISE!!!

Morakanyane R.S. MSc.

Page 86

Appendix 3: Voicent Manual: How to Setup IVR Studio Table of Content Part I: Basic Usage 

Overview o IVR Studio Components o Installation



Lesson 1: Call Flow Diagram o Create an element to collect password o Add prompt o Check correct password o Check incorrect password o Call Flow Diagram o Deploy application o Test Lesson 1



Lesson 2: Variable o Create a new variable o Update variable value o Use variable o Log caller ID o Test Lesson 2



Lesson 3: Record Voice Message o Create a record element o Create a choice element o Listen to recorded message o Test Lesson 3



Lesson 4: Run External Program o Create a run program action o Create a variable for the voice message o Use variable as program argument o Test Lesson 4 o Use Windows Batch File o Edit the run program action o Test Lesson 4 Again



Lesson 5: Design for BroadcastByPhone Autodialer o Create an outbound IVR application o Create a column variable o Use a column variable o Set audio message o Select Autodialer outbound IVR application

Morakanyane R.S. MSc.

Page 87

o o

Edit the Windows Batch File Test Lesson 5



Lesson 6: Communicate with a Website o Create an HTTP action o Edit the Windows Batch File o Deploy Inbound and Outbound IVR Applications o Test Lesson 6



Lesson 7: Improvements and Resources o Use a natural text-to-speech engine o Use pre-recorded audio o Use a multi-line system o Resources

Part II: Advanced Usage (For Developers) 

Lesson 8: Integrate with IVR Applications o Update the IVR Call Flow Diagram o Create a Java Action o Write a Java Class for Java Action o Compile and Run the Java Class o Use Action Return Variable o Update the Windows Batch File o Add a Run Program Action o Test Lesson 8



Lesson 9: Integrate with Relational Database o Learn more about relational database o The Java Class to check password using JDBC o Create a Java action to check password o Use Java action result to determine which transition to take o Get more information from the database o Use database action

Overview Interactive Voice Response (IVR) applications enable callers to interact with any software, such as query and modify database information, over their telephone. Callers can use their touch-tone pad to input requests or just say what they want to do, such as requesting account balance information. IVR systems usually employ text-to-speech software to read information back. Voicent IVR Studio is a complete IVR system, completed with an easy-to-use graphical user interface for application development. Voicent IVR Studio also supports outbound IVR applications through Voicent BroadcastByPhone Autodialer.

Part I & Part II Morakanyane R.S. MSc.

Page 88

This tutorial contains two parts: Part I Basic Usage and Part II advanced usage. Programming knowledge is required for Part II.

The Basic Sample for Part I This tutorial provides step-by-step procedures to create an IVR application, which enables callers to call in to record a voice message and broadcast the message to a selected call list. The sample for Part I is a complete but simplified version. It is described below: 1. The sample IVR application answers a call 2. Plays a welcome message, asking the caller to enter password 3. If the password is incorrect, go back to step 2. If the caller has tried more than 3 times, log the caller ID and hang up the call 4. Asks the caller to record a voice message 5. Asks the caller to listen to the recorded voice message, press 1 to continue and press 2 to record again 6. Start Voicent BroadcastByPhone with the recorded message and a fixed call list file 7. Hang up the phone

The Advanced Sample for Part II The complete sample is listed below. The difference is in step 6 and 7 where the caller is asked to select from a list of call list files. A Java program is used to obtain a list of available call list files. 1. The sample IVR application answers a call 2. Plays a welcome message, asking the caller to enter password 3. If the password is incorrect, go back to step 2. If the caller has tried more than 3 times, log the caller ID and hang up the call 4. Asks the caller to record a voice message 5. Asks the caller to listen to the recorded voice message, press 1 to continue and press 2 to record again 6. Asks the caller to select a call list for the broadcast 7. Start Voicent BroadcastByPhone with the recorded message and call list 8. Hang up the phone

IVR Studio Components IVR Studio consists of two components: IVR Studio GUI (graphical user interface) and Voicent Gateway. The GUI component is used to design IVR applications. Voicent Gateway is the component that actually handles telephone interactions. IVR Studio GUI

IVR Studio can be invoked from Windows Desktop or from Windows Start menu. Simply click the following IVR Studio icon to invoke the GUI component. Voicent Gateway

Morakanyane R.S. MSc.

Page 89

Voicent Gateway is the component that actually handles phone interactions based on IVR applications. You can find the gateway icon in the Windows toolbar notification area. Except for a few option settings, most of the time the gateway operates in the background. It runs as a Windows service, which automatically starts with Windows starts. Voicent Gateway can be used for both incoming and outgoing calls. The none-shareware edition of IVR Studio includes an inbound gateway license. If you are a developer and would like to develop your own interactive voice applications using Voicent Gateway, please take a look at Voicent Developer Network.

Installation You can download the latest release of IVR Studio from Voicent website: http://www.voicent.com/download Install IVR Studio

Please see IVR Studio Quick Start Guide For this tutorial, you also need to install Voicent BroadcastByPhone. Enable Incoming Call Handling

By default, Voicent Gateway does not answer incoming calls. To enable incoming calls, please select Setup > Options from the gateway main menu, choose the Inbound tab, and uncheck the box labelled "Do not pick up incoming calls".

Morakanyane R.S. MSc.

Page 90

Lesson 1: Call Flow Diagram The first three steps of this sample involves: 1. The sample IVR application answers a call 2. Plays a welcome message, asking the caller to enter password 3. If the password is incorrect, go back to step 2. If the caller has tried more than 3 times, log the caller ID and hang up the call This section shows how to capture the above interactions in a call flow diagram. A call flow diagram defines an IVR application in IVR Studio.

Create New IVR Application Start IVR Studio, select File > New from the program main menu. The initial IVR application is shown below. A top element/node labeled Incoming Call is automatically created. Morakanyane R.S. MSc.

Page 91

The rest of this section covers:       

Create an element to collect password Add prompt Check correct password Check incorrect password Call Flow Diagram Deploy application Test step 1

1.3 Check Correct Password Once a password is entered, the system needs to check the correctness of the password. In this sample, we'll assume the correct password is "123". You can certainly develop more secure mechanisms to check the password, such as comparing password stored in a database or file. In this step, we'll just inform the caller that the password is correct, and then hang up the phone. In later sections, we'll change the handling when a password is correct. We'll create two child elements to handle the password: one for correct password, and the other for incorrect password. Create a prompt element

A prompt element, as the name indicates, simply plays the prompts defined in the element and then transition to the next element (its child element). Right click on the Get Password element, select Add Next Element, then select Prompt. Enter the information as indicated below, then click the Next button.

Morakanyane R.S. MSc.

Page 92

From the Prompt page, click the New button to add a prompt item. Choose the Text-to-speech radio button, then enter the text like: "Your password is correct". Transition from Parent Element

The transition from Parent Element box defines the condition the element is to be activated. In this example, if the touch tone key entered (collected by the parent element Get Password) is "123", this element is activated. 1.4 Check Incorrect Password If a password is incorrect, the application should go back to the Get Password element to collect the password again. Create an Jump/Goto Element

A jump or go-to element simply transition the active element to the target element. Select the Get Password element, click the Add Goto element button on the toolbar. Enter the name as Password Incorrect, touch tone key "...", and select Get Password as the Jump to Element. Click the Next button to continue.

Morakanyane R.S. MSc.

Page 93

From the Prompt page, click the New button to add an prompt item. Select the Text-to-Speech radio button, enter text like: "Your password is incorrect". Transition from Parent Element

The transition from Parent Element box defines the condition the element is to be activated. In this example, the three dots "..." represents all other keys.

Morakanyane R.S. MSc.

Page 94

1.5 Call Flow Diagram The design so far is shown below:

As you can see, the IVR application is defined as a tree structure similar to the folders and files structure in the Windows environment. The tree structure is called a call flow diagram. A call flow diagram also looks like a user manual of a voicemail system. The tree node is called an element in IVR Studio. An element defines a certain stage of an IVR application. For example, the Get Password element is at the time right after an incoming call is answered. An element also defines a set of prompt items and activities. For example, the Get Password element plays the welcome message and waits for the caller to enter his or her password. The link between elements is called a transition. A transition defines a change of the active element. in order for a transition to happen, the condition of the transition must be satisfied. For example, the transition from Get Password element to Password Correct element happens when the caller enters the correct password "123". The call flow diagram defines the following interactions: 1. 2. 3. 4.

The sample IVR application answers a call Plays a welcome message, asking the caller to enter password If the password is incorrect, go back to step 2. If the password is correct, inform the caller the password is correct and hang up the call

1.6 Deploy Application The IVR application defined so far, though simple, is a complete application that can be deployed and tested . Morakanyane R.S. MSc.

Page 95

Validate Application

Click the Validate button on the program toolbar. You have to validate the design and fix all errors before an application can be deployed. Deploy Application

Once validated, the deploy/submit button is enabled on the toolbar. This is the same icon used for Voicent Gateway. Click the button to deploy the application to Voicent Gateway. Enter the application name as my_broadcast, and click the OK button.

Restart Voicent Gateway

Now re-start Voicent Gateway to have the new application take effect. To restart, click the gateway icon from the Windows toolbar notification area, press the Stop button and then press the same button (now labeled Start) again. You can check all deployed application from the Application main menu of Voicent Gateway. If you have multiple inbound application installed, please make sure the my_broadcast application is your default application to answer incoming calls. For more details, please see: IVR Manage Applications. *** Please make sure that answering incoming call is enabled. (Voicent Gateway > Setup > Options > Inbound)

Morakanyane R.S. MSc.

Page 96

Appendix 4: Conference Paper accepted at the Prato Community CIRN 2009 Conference Representing Information for Semi-Literate Users: Digital Inclusion Using Mobile Phone Technology Abstract Botswana as a developing country is one of those countries hard hit by the HIV/AIDS pandemic and the impact of HIV/AIDS on national development and socioeconomic transformations has made it a national development priority. The government is working hard to facilitate the provision of education and raising awareness concerning the pandemic. However, the country is facing a challenge of uneven population and development distribution with most population, developments and higher levels of literacy found in urban areas. As a result such challenges disadvantage rural populations by making them the victims of the digital divide and lagging behind when it comes to information accessibility. To improve this situation, ICT solutions are needed to bridge this gap In January 2005 the Department of Computer Science, University of Botswana received funding for a project, which is part of Microsoft Digital Inclusion initiative [1] aiming at bridging the gap between relatively welldeveloped and less-developed regions in the world. The project termed, IHISM (Integrated Healthcare Information System through Mobile telephony) [2] aims at bridging the digital divide by developing an HIV and AIDS public information portal accessible through mobile phones. HISM takes advantage of the country’s high mobile penetration to improve access to HIV/AIDS information. This research aims at devising the best technological solution to present HIV and AIDS information in ways that the semi-literate and illiterate populations can be able to comprehend and use effectively using mobile phones. Keywords: mobile phone; Digital HIV/AIDS; Semi-literate and illiterate Users;

Inclusion;

1.0 Background Botswana as a developing country is one of those countries hard hit by the HIV/AIDS pandemic with an estimated 330 000 to 380 000 of its adult population (aged 15-49) living with HIV/AIDS (about 35.5 to 39.1%) [3]. Because of the impact of HIV/AIDS on national development and socio-economic transformation, the pandemic has been described as an area of national development priority. In an attempt to fight this major challenge, the government through its different ministries, has devised

Morakanyane R.S. MSc.

Page 97

HIV/AIDS policies including the policy on providing antiretroviral therapy. The implementation of antiretroviral therapy started in 2002 and expanded to 23 sites in 2004 [4]. Treatment is provided free of charge in the public sector which has positively influenced the demand for voluntary counselling and testing. In addition, various education and communication programs have been created [5-7]. These programs aim at raising awareness of the disease and offer relevant information to the general public on the dangers and management of the disease. The government, through the Ministry of Health also provides public handouts and hosts public events such as HIV/AIDS fairs and awareness campaigns to try to educate and sensitize the public about the disease. Other fighting methods include forming various partnerships to combat HIV/AIDS [8.9]. Nevertheless, this response has seen Botswana exhibit high prevalence to HIV/AIDS infections over time as the general awareness of the dangers of the disease is raised among the population. On the other hand, it is also observed that rural populations in Botswana have many information needs which are not effectively addressed. Such information needs include access to government and social services, learning about empowerment initiatives that exist in other areas and more importantly access to crucial health information, such as HIV/AIDS education. According to Botswana Central Statistics Office Population Survey Report of 2001 [10], the country is facing a challenge of uneven population and development distribution with most population, developments and people with higher levels of literacy found in urban areas while rural areas experience less population, less developments and lower literacy levels (semi-literate and illiterate). This is supported by a survey Report on Botswana Literacy [11], which asserts that the literacy rate is lower in the rural areas as compared to the urban areas. With Botswana being hard hit by the HIV/AIDS scourge, these people therefore, remain vulnerable to contracting the disease as they have limited means of accessing and comprehending information being passed to them. Hence the most affected population remain and the semi-literate and illiterate communities in the rural areas [12]. The problems described give a clear indication that the complexity and incurability of HIV/AIDS requires that the government, in addition to providing clinical support, needs to also put an emphasis on the dissemination of

information about the disease, prevention methods, therapeutic methods and psychological support. With regard to availing formal education to all parts of the country, the government has attempted programs like adult education and distance learning. This is in attempt to avail the opportunity to those who could not have access to formal education and not necessarily inclusive of those who have not. In this age it is evident that in order to address this problem ICTs can be used to solve this problem need to be explored. The project termed IHISM aims at exploring the use of mobile phones as an access technology to a variety of HIV/AIDS related information as required by the general public.

someone has with using computer programs and other applications that are associated with computers. The computer literacy level is however strongly influenced by the level of education an individual has received. The semi-literate therefore refers to people who have very minimal information and communications technology skills while the illiterate refer to populations who have absolutely no information and communications technology skills. This research will therefore adopt the meaning of literacy as the ability of an individual to use information and communications technologies efficiently. It is also important to note that the terms semi-literacy and target users may, in some instances, be used to refer to both cases of partial literacy and total illiteracy henceforth. In their study Joshi and Rasal [14], realized that most semi-literate and illiterate populations live in rural areas. Other researches also shows that close to 4.8 Billion semiliterate and illiterate people live in developing countries and many of these live in rural areas, [15] which is also a case in Botswana.

Figure 2 IHISM Architecture

The problem addressed in this paper, henceforth referred to as the Digital Inclusion Problem, is to propose a system that will best serve semi-literate and illiterate users to access HIV/AIDS information using mobile phones.. 2.0 The Digital Inclusion Problem

This research aims at representing HIV/AIDS information for the semi-literate and illiterate members of the society on mobile phones. The definition of literacy differs depending on the different contexts in which the word is used. Collins dictionary [13] gives a general definition of literacy as the ability to read, write, communicate and comprehend. The semi-literate and the illiterate therefore are people who have never gone to school to have any form of formal education or people who have been to school at some stage of life but have lost touch with reading and or writing. These could also include people who have learnt selective terms at their own initiative or by peer learning. In computers, literacy generally refers to the knowledge and ability to use information and communication technologies. It can also refer to the comfort level

Morakanyane R.S. MSc.

Page 98

The low literacy level in rural areas means that most members of these communities have difficulty in understanding health information that the health professionals would otherwise want them to receive. With the main methods of disseminating health information in Botswana being newspapers, pamphlets and posters, these methods are not necessarily adequate for information dissemination as information being disseminated through these media is not necessarily customized for the needs of the semi-literate and illiterate communities in the rural areas. In addition, accessibility to such information may also be difficult due to great distances. Currently when the target user wants to enquire information on any health issue, they have to make a physical visit to the nearest health centre during operating hours. Traveling from home to the health centre obviously takes a significant amount of time. The queue that the user will join will consume even more time. Those who live further travel even more kilometres to seek medical advice. On the other hand, the country is currently facing serious shortage of health personnel which is a very risky situation in this HIV/AIDS pandemic. The Ministry of Health’s human resource levels indicate that in 2004 approximately 30 physicians and 262 nurses were available per 100,000 populations [16], which show a very low density considering the size of the country and the uneven population distribution. The government is going to great lengths to bring health services to rural communities, however, as a result of this shortage, there are generally fewer health personnel and services in the rural settlements leaving the communities to travel great distances to access health care. In an attempt to address issues of shortage of health professionals the Botswana Technology Centre undertook a Telemedicine Feasibility

Study to see if telemedicine could be a solution to Botswana’s shortage of health professionals [16]. According to the recommendations of this report, the Ministry of Health is currently developing a telemedicine link between two of the country’s hospitals – one primary hospital in the semi-urban area, one referral hospital in the urban area. The war on HIV/AIDS will not be won if some members of society are excluded not because they could not access the readily available information. Even though the above mentioned initiatives help in terms of alleviating problems of shortage of health personnel, they have not managed to totally address issues of information dissemination and representation for the semi-literate and illiterate populations to access and use HIV/AIDS information effectively. These communities also face other constraints which include lower purchasing power, poor connections to main lines of communication and infrastructure. As a result such challenges disadvantages these populations by making them the victims of digital divide, lagging behind when it comes to information accessibility. To improve this situation, there is therefore a need for technology solutions to bridge this gap. The application of ICT research and development to support health and medicine is an emerging research area with significant potential for use by semi-literate and illiterate communities. Major initiatives to improve the quality, accuracy and timeliness of healthcare data and information delivery are emerging all over the world. The Agency for Healthcare Research and Quality (AHRQ) [17] of the US Department of Health Services (HHS), awarded grants and contracts to promote the use of health information technology. The advent of HIV/AIDS has prompted the development of knowledge base systems in this problem domain. In the recent, researchers are developing Question and Answer systems on HIV/AIDS to assist mainly medical practitioners. Most of these systems differ in the way they were developed and take different forms of HIV/AIDS expertise delivery to the intended users. However, attempts to provide systems for the patients and care givers are now being more common. Temesgen; et.al. [18] developed a questions and answers support named CHESS which provides answers to questions logged in by people living with HIV/AIDS infection. Ybarra et al [19] on the other hand discuss yet another approach to these HIV/AIDS Question and Answer Systems. They discuss how cell phones and internet can be new trends of HIV Prevention and Intervention Programs in this era where an estimated 78% of Americans aged 12 years and older are now online and 35% of adults who have mobile phones can use text messaging. Other health sector players with HIV/AIDS issues at the core have developed Q&A systems [20, 21] with the objective of providing the public with online health (including HIV/AIDS) information Q&A systems for individual consumption.

Morakanyane R.S. MSc.

Page 99

Despite these challenges, one advantage is that Botswana has a high mobile penetration, including in rural areas where there are semi-literate and illiterate populations. In his article titled Telecommunications in Botswana [22], Tamocha verifies that mobile phone technology has taken off in Africa like no other technology ever did in the history of the continent. He states that Neil Ford of the Pan-African Business Magazine indicated that Africa has the fastest growing mobile phone market in the world with around 50 000 new subscribers every day. He goes further to state that the Botswana Telecommunications Authority [23] a body which licenses telecommunications service providers, puts the figure of mobile phone users in Botswana at 85 people for every 100. Mobile technology is after all the leading mode of telecommunications in Botswana and has reached a high level of acceptance by the users. Parikh, [24] emphasizes how mobile phones are the right devices for rural accessibility. He advocates for mobile phones as ‘rural computers’. The hardware is a great fit for rural conditions since it is battery-powered and networked at a low-cost. To take up this opportunity, this project aims at using mobile phones to improve access to HIV/AIDS information by representing the information in ways that the semi-literate and illiterate populations can be able to comprehend and use effectively. The objective is to develop a general HIV/AIDS information portal which would be queried by the users using mobile phone technology. The portal represents this information in the form of FAQ service where the user input a query on any of the subjects. Majority of the FAQ come from the Ipoletse Manual, [25]. Ipoletse is a Setswana term which means ‘Ask for yourself’. This means if you need information just pick the phone and have the information given to you. The call center was set up by Botswana Government in 2002 with a mission to provide free information on HIV/AIDS to the public. The manual consists of over 200 questions and answers on HIV/AIDS related information. The information service portal would allow the general public to request for information on topics related to HIV/AIDS such as descriptions, infection, testing, counselling and support, opportunistic diseases and paediatric care etc. This project also seeks to explore the different ways in how information can be represented in mobile phones for the semi-literate and illiterate users which is the aim of this paper. With the semi-literate and illiterate populations having different information needs that need to be urgently addressed, effectively representing information for them remains critical. The importance of representing health information, especially HIV/AIDS information, in a form that these populations can easily understand cannot be overemphasized. The research will address this by looking at the technical feasibility of using mobile phones

to improve accessibility to HIV/AIDS information as well as how the target users will perceive the technology being availed. Human Computer Interaction (HCI) is concerned with how human-beings interact with computing technology and relevant technologies. Interaction between users and ICTs occurs at the user interface level, which can be software, hardware or both. For example, in a mobile phone, the mobile phone screen and the mobile phone operating system are both user interfaces. Basic goals of HCI are to improve the interactions between users and systems by making the systems more usable and receptive to the user’s needs. Issues that need to be considered when attempting a solution for this digital inclusion problem are therefore: Feasibility: Establishing whether HIV/AIDS information can be represented and accessed through mobile phones as well as establish whether the semi-literate and illiterate would be able to use their mobile phones to access HIV/AIDS information. Accessibility: Establishing if the current telecommunications infrastructure would be adequate enough to support the solution that may be proposed. Interactivity: Ensuring that the proposed solution would allow the users to select different options in order to access the information they want Furthermore, the project proposes to test the psychological behaviour of the semi-literate and illiterate populations towards the solution to be proposed. In particular, the psychological behaviour issues to be considered will be: Acceptance: Establishing if the users will accept using mobile phones to access critical information like HIV/AIDS information as well as establishing if the users will not be intimidated by their inability to understand the use of technology when attempting to use the proposed solution Integrity: Establishing if the users will be able to trust the information being obtained from the proposed solution with out feeling that their privacy is being violated. The project wishes to establish whether the target users could afford to access the solution to be proposed. This expense could range from subscriptions to buying more powerful mobile phones to access multimedia information. In determining these factors, it is important to note that by nature, the semi-literate and illiterate populations are observed to be passive when it comes to issues of technology. Because of lack of exposure to technology, they may also become impatient and withdrawn when technology is being presented to them hence can easily

Morakanyane R.S. MSc.

Page 100

reject technology solutions being brought to them. In their study to explore attitudes of the semi-literate and illiterate towards the use of ATMs, Shaik et.al. [26] reveal how the semi-literate and illiterate may be too embarrassed to use ATMs for fear of making mistakes hence causing unnecessary delays and queues. On the other hand, Baecker and Buxton [27] describe how novice users, who are also somewhat semi-literate or illiterate, feel frustrated, insecure and even frightened when they have to deal with systems whose behaviour is incomprehensible, mysterious and intimidating. At the same time, these populations significantly contribute to the high levels of mobile phone penetration in Botswana as they have been observed to be among mobile phone users. This research therefore seeks to attempt a solution for the digital inclusion problem, taking these factors into consideration. 3.0 Information Representation Different information transmission technologies in mobile phones include Short Message Services (SMS), Multimedia Messaging Service (MMS), Voice Technology and Video Technology. SMS employs the sending of short messages between mobile devices. These could be plain full text messages or messages grouped by keywords.Within the context of the digital inclusion problem defined, users can be able to send full text SMS to the system to request for HIV/AIDS information as well as receive full text SMS from the system as responses.The system can also be used together with the faxes and emails to allow for files larger than the normal SMS file (160 Characters) to be sent to users as responses. MMS allows for sending of messages that have multimedia objects such as images, audio, video and rich text. This method of data representation involves grouping information or ideas and representing them by graphical images. In the digital inclusion solution the system can allow for sending of picture messages to users as reminders of different moments such as time to take medication. Voice Technology uses voice to access functions and delivers responses to mobile phones. Voice technologies include the normal voice to voice phone call; voice message, voice to text messages, interactive voice response technologies, among others. Typical examples for the digital inclusion solution include the system sending voice reminders at the times which the user has to take medication; a user calling in and talking with the system to enquire about any information they may be looking for; where the user can call the system and follow the menu based voice instructions by pressing the mobile phone keypads; where the user can call the system and follow the voice instructions they are getting through accessing the menu being presented by voice. Video Technology involves representing data using a combination of images and sound over time, therefore this

method is developed from both MMS and voice technologies. Typical video applications include the system sending video clips of general health information to the users.

pre-recorded voice clips. The system may then extract information relating to ‘fighting the stigma’ and send it to the number that sent the request. In this example, issues to do with the HIV/AIDS stigma such as HIV/AIDS in the Work Place, Living Positively, etc. can be grouped under one key word – ‘fighting the stigma’

In terms of messages grouping by keywords, different ideas or information can be grouped into single general keyword to represent ideas that the system may want to Different solutions that can be obtained from different communicate to the users. This way of representing data combinations of these technologies include among others: can be developed from the SMS technology as well as plain text SMSes; menu-based speech recognition; textutterance in voice technology. A typical example of to-speech; speech synthesis; menu-based SMS grouping grouping by keyword in the digital inclusion solution can by keyword; picture messaging; video messaging. Table be where a user sends the phrase ‘fighting the stigma’ to 1 below shows the technologies explored and the the system and the system picks the keyword sent and advantages and disadvantages associated with each search the knowledge base of the pre-saved text files or method of representation. Table 1: Comparisons among Mobile Phone Data Representation Techniques Data Representation

Strengths

Weakness

Plain Full Text Messages

Relatively cheap and affordable , Allows for up to 3 concatenation of messages, Allows for message broadcasting

Usually limited to 160 characters, Most keypads do not support local languages, Requests users to type and read responses hence may be a constraint to the target users

MMS

Good where users do not have the ability to read or write , If the users are well trained in the meanings represented by the graphics then graphics require less cognitive load. Graphics can take any form

Use of graphics poses incompatible content adaptation, difficulties in sending to distribution lists, inefficiency in sending bulk messaging and incompatible handset configurations. May require a lot of user training such that users know exactly what each image means so as to avoid misinterpretations of information. Image transmissions require special networks and may be carried out at high costs.

Grouped by Keyword (Text and Voice)

Voice utterance is ideal for target users who cannot read and type, Well organized and classified information easily retrievable from the database,. Requires less cognitive load than full text messages as the users has to remember only the keywords used in the system

Not ideal for target users – who cannot read and type, May require intense training so that the users know exactly the keywords used in the system and avoid miscommunications, Keywords tend to generalize information a lot making it difficult to obtain precise meaning and detail of what one is looking for

Voice and Speech Technologies

Uses less cognitive load, Can be used where users cannot read or write

Video Technology

Video may tend to be more effective as it provides both images and sound

Technology more complex to develop, Recognition is generally more difficult when vocabularies are large or in similar-sounding words, Require complex technologies to transmit, Occupy large spaces and large files take time to transmit especially with limited bandwidth

4.0 Field Survey and Data Collection In order to pursue the digital inclusion problem, the above different techniques were explored through a field study on a sample of rural community members in Botswana. The survey was restricted to two regions in Botswana. These areas are Ngamiland West and Kweneng West regions. These regions were selected based on them being

Morakanyane R.S. MSc.

Page 101

the regions with the lowest literacy level, according to Botswana’s Literacy Survey Report [28]. Data was collected in one main village or a few villages in each of the regions. The selection of the villages was influenced by the availability of mobile phone networks as well as willingness of the community to take part in the survey. To gain access into these communities Rural Administration Centres, Tribal Administration

Authorities, Village Development Committees and other village administration bodies were involved. This process was necessary so that the participants got encouraged by seeing people they already know as part of the study. A total of 32 hours was spent at each village where data was being collected. The process was used to ensure that there was intense interaction with the target users so that they felt relaxed when interacting with the interviewers. Through this interaction it was hence easier to establish

the real needs, traits and responses of the participants. In total 127 participants volunteered to participate in the exercise. The participants had age range of 16 to 77 years with 8 participants not knowing their ages. Majority of the participants had received education up to primary school level. 21% had not received any form of formal education while 64% had not received secondary, hence a true representation of semi-literate and illiterate population. Figure 2 below is a graphical illustration of these results

Figure 2: Results on Participants Profile The main purpose of this field survey was to collect data that will help determine how best to represent HIV/AIDS for semi-literate and illiterate users on mobile phones. The data collection process therefore involved finding out about the current situation of the participants in their villages, how the participants use their mobile phones on a day to day basis as well as the perceptions and psychological issues the participants would have with regard to mobile phone based solution providing HIV/AIDS information. The data collection tools used in

Morakanyane R.S. MSc.

Page 102

this process included questionnaires, interviews and general observations as the participants used their mobile phones. The questionnaires were used to collect participants’ profiles; the interviews were used to record participants’ perceptions while observations were used to record reaction to the technology. Other supporting data collection media to be used includes video and still picture cameras. Table 2 summarizes the results obtained on usability, interactivity and perception of the participants towards using mobile phones to access HIV/AIDS Information.

Table 2: Results on Usability, Interactivity and Perception of the Participants

No. 1 2 3 4 5 6 7 8 9 10 11 12 13 No 1 2 3

Usability and Integrity Responses Ranked in Percentages (%) Variable Measured Poor Below Average Average Above Average Good Establish a call 5 5 4 17 69 Create SMS 32 28 16 14 10 Send SMS 31 6 7 21 35 Retrieve SMS 17 10 10 35 28 Read SMS 8 42 38 10 2 Reply SMS 17 10 28 30 15 Grouping by Keyword 28 35 23 12 2 Send Pic Msg 25 35 22 10 8 Retrieve Pic Msg 15 35 36 10 4 Use Picture Messages 3 4 7 21 65 Create Voice Msg 11 46 29 13 1 Retrieve Voice Msg 31 24 27 12 6 Use IVR 5 9 24 29 33 Perception, Psychological Acceptance and Affordability, Ranked in Percentages (%) Variable Measured Responses Willing to Pay Willing: 60; Not Willing: 23; Undecided: 17 Privacy and Trust Trusting: 87; not Trusting: 1; Undecided: 12 How Often the User Will Access the System Frequently: 4; Often: 62; Rarely: 34

5.0 Discussion The main observation from this sample was that majority (70%) of participants had mobile phones therefore the solution proposed would have readily available potential users. In addition 87% were able to identify counting numbers and 86% could establish a simple call rendering that if the solution to be proposed has a dial in function most of the users should be able to use the function. However the more important factor was to determine whether the participants could actually use the phones as intended, that is, to send and retrieve HIV information to the IHISM portal. In terms of creating and retrieving SMS messages, more people could send and retrieve messages and yet less could create it which is the main function for FAQ input as shown in Table 2. As long as they could not create the SMS, other capabilities alone are not adequate enough to qualify usage and thus SMS technology would remain a challenge for the majority of the users. Majority of participants could not use a grouping by keyword function as they failed to either read and or interpret responses obtained from sending keyword messages. This means if the solution to be proposed is to have the grouping by keyword function, more users would find it difficult to use. With picture messages, majority of participants did not have mobile phones that supported picture messaging to start with. In addition, participants failed to store or send a picture message when they were provided with

Morakanyane R.S. MSc.

Page 103

a phone that supports picture messaging. Mobile phones with technologies that use graphics, images and icons require special infrastructure may be too expensive to be used by the semi-literate and illiterate people. However a significant majority responded extremely well when given an opportunity to follow a picture menu message once the pictures were retrieved for them. Menu based representation is a representation in which users traverse through some form of pre-programmed menu in order to access the system In observing the use of voice technologies, majority of participants could not confidently create nor retrieve a voice message. However, a significantly large majority were able to interact with a prerecorded voice messages which are accessed through menu based voice prompts. A typical application of a menu based system include the common mobile communications providers’ prepaid service which the customer dials and access a menu to access different services such as checking their balance, playing a voice message, recharging their airtime. Therefore an IVR based solution may be relevant for the target users. In terms of perception, participants seemed to embrace technology (87% Trust) and agree that having a mobile phone based solution would expedite the process of seeking for medical help. They further assert they would use the service often enough (62%)

and would be willing to pay for the service as opposed to the long distance travel to health centres. However to cater for the others not willing not to pay, the government could be approached for a subsidiary collaboration with the mobile service providers. In conclusion, participants welcome to the idea of a mobile phone based solution. They own mobile phones, can establish simple mobile phone calls and are able to identify counting numbers on mobile phone. This is a positive indication that participants would be able to use their mobile phones to dial into the IHISM portal. It also emerged that menu based representation for both pictures and voice are the best technologies to use for semi literate users. This is because the users feel more confident if they are guided through the process which has less cognitive load on them. However, beside expense, most participants’ phones did not support picture messaging. The option seemed not viable. Plain text messaging was also eliminated as it required users to be able to create messages, something majority could not do. Grouping by keyword was eliminated as it used keywords and outputs that may be complex for semi-literate people. The only viable option seemed to be a voice based solution. However the participants could not create or retrieve voice messages unless they were guided. Nevertheless, participants found it easy to maneuver through the IVR system. The familiarity with existing IVR system that are generally available and used by the mobile phone service providers in the country gave a an edge to the recommendation of this solution.

6.0 The Proposed Solution

Figure 3: Information Flow Logic of the Proposed System START

USER DIALS

PLAY INTRODUCTION

ENGLIS H

SELECT LANGUAGE

SETSWA NA

PRESENT MENU

SELECT MENU OPTION

PLAY CLIP

A menu based interactive voice response system has emerged to be the most appropriate solution to the defined digital inclusion problem. The voice technology presents representations that are closest to the natural human language and do not require any level of literacy. In this context, a menu based representation would involve pre-programmed menus which the user can interact with through dialling into the system, listening and pressing numbers on their mobile phone keypad. Parikh [24] observes that numeric keypads and the utility of voice communications are familiar to billions of users and therefore many users in rural places, including the semi-literate and the illiterate are generally comfortable with this service. The voice used to record the information clips should also be clear and smooth for the users to hear.

Morakanyane R.S. MSc.

In the proposed system users will be able to make calls using their normal GSM mobile phones and follow guided instructions. On receiving the call, the system will retrieve and play introductory clip to the user and present different options from the menu, such as ‘for information on HIV/AIDS and Nutrition press 2’. It will then get the option that the user has selected, searches it through the knowledge base and play the relevant clip to the user. Figure 3 below is a graphical illustration of the flow logic of how the user would interact with the proposed system.

Page 104

STOP

In this solution, the Mobile Phone is used by the user to connect to the system through the GSM network. The IVR consisting of an E1 transceiver card connects to the GSM network and the logic and receives DTMF tones generated by the mobile phone. The IVR also contains the menu options that the user will be accessing. The logic receives information from the GSM card about what to do. It will then play static pre-recorded menu clips as well as collect information clips from the database and play them to the user. The logic also manages the GSM card. The Knowledge Base is a normal relational database that stores information clips that the users will be accessing. It should be updated regularly with new

information when it is available. Information that is accessed through the system would be classified in different categories. These categories are a representation of issues the users normally enquire about on day to day basis. The information is not restricted to these categories only. It can be improved to include more categories when need arises. 7.0 System Development

the IVR application the Voicent IVR Studio was used. IVR Studio is ideal for developing the IVR applications through drawing call flow diagrams. This tool enables flexible application development and it supports VoiceXML open standard, DTMF tones, text-to-speech as well as speech recognition. Figure 5 below shows the development environment of the call flow diagram for the digital inclusion solution using Voicent IVR Studio.

Speech Based Software come in different packages. There is the free open source and the commercially developed software. Examples of speech based software in open source include XVoice, Console Voice Control, DialogPalette to mention but a few while examples of commercial speech or voice software include IBM ViaVoice, Microsoft Speech Server, etc. Any of these software can be used to develop an ideal digital inclusion solution. However, the solution proposed for the digital inclusion problem in this study was developed using Voicent Communication Software [29]. Voicent is a communication solutions provider which uses speech synthesis, computer telephony and internet technologies. The major components of Voicent which are relevant to this research are the Voicent Gateway and IVR Studio. Voicent Gateway is an Open VoiceXML standard development platform for telephony program applications. It recognizes voice over telephone and carry out commands accordingly. It also reads any text to the user over the phone as well as accept input in the form of DTMF tones from the user’s numeric keypad. To develop the graphical user interface for

Figure 4 Snapshot of Knowledgebase

8.0 Conclusion Botswana as a country hard hit by the HIV/AIDS pandemic and low literacy rate especially in rural areas faces challenges of disseminating information to all communities in time. The most affected are the semiliterate and illiterate communities who also have limited access to other social services. As a result health information reaching these areas is either limited or not represented in a manner in which these populations (semiliterate and illiterate) can easily absorb. This project aimed at trying to understand how mobile technology can be used to bridge this digital divide. It did so by looking at how information can be represented for the semi-literate and illiterate members of the community such that they also benefit from getting health information at their

Morakanyane R.S. MSc.

Page 105

earliest convenience. From this it was discovered that mobile technology is widely used and accepted in the society as indicated by the high numbers of mobile phone subscribers. It is also clear that this technology is common to everyone regardless of level of literacy. Voice technology emerged to be the most appropriate way which could be used to represent information to cater for the semi-literate and illiterate communities, taking into consideration their level of literacy, standards of living, psychological behaviour and all other factors. The research ends with a prototype of a menu based IVR system being implemented as a sample of a recommended solution to representing information to the semi-literate and illiterate. As a way forward the system needs to be tested by the users and verified as the next research issues.

9.0 References 1.

2. 3. 4.

5.

6.

7.

8.

9. 10.

11. 12.

13. 14.

15.

16.

17.

18.

Microsoft Digital Inclusion Program http://research.microsoft.com/ur/us/fundingopps/ rfps/DigitalInclusion_2005_RFP.aspx IHISM Project website www.ub.bw/ihism World Health Organization; HIV/AIDS (WHO/UNAIDS); United Nations; 2005. Summary Country profile for HIV AND AIDS treatment scale-up, Botswana, WHO HIV AND AIDS Plan January 2004-June 2005 Botswana National Strategic Framework for HIV AND AIDS", NACA, 2003. The NACA Project, http://www.naca.gov.bw/about.htm Accessed 22 April 2008 MASA - The National Antiretroviral Therapy Programme of Botswana", http://www.moh.gov.bw/index.php?id=192, Botswana Ministry of Health, accessed April 2008. IPOLETSE Call Centre for HIV AND AIDS, http://www.mri.co.bw/html/call-ipoletse.htm, Botswana Ministry of Health, accessed April 2008 African Comprehensive HIV AND AIDS Partnerships (ACHAP), http://www.achap.org/profile.html Accessed 22 April 2008. Botswana Harvard Partnership http://www.hsph.harvard.edu/bhp/ The Central Statistics Office Population Survey Report of 2001; Botswana Central Statistics Office; 2002 Botswana Literacy Survey: 2003; UNESCO; December 2004 Botswana AIDS Impact Survey II Statistical Report 2004; Republic of Botswana; Central Statistics Office; October 2005. Collins Dictionary: www.collinslanguage.com Joshi, A. and Rasal, S.; Interaction Paradigm for Semi-literates and Illiterates; Industrial Design Center IIT; Powai, Mumbai; 2002. United Nations Development Programme; Botswana Human Development Report Harnessing Science and Technology for Human Development; UNDP, Gaborone, Botswana; 2005. Telemedicine in Botswana Feasibility Study Report; Botswana Technology Center; September 2004 Agency for Healthcare Research and Quality (AHRQ) http://www.ahrq.gov/RESEARCH/hitfact.htm Accessed January 20 2009 Temesgen, Zelalem1; Knappe-Langworthy, James; Marie, Mary; Smith, Becky; Dierkhising, Ross, A Comprehensive Health Enhancement Support System (CHESS) for People with HIV

Morakanyane R.S. MSc.

Page 106

19.

20.

21.

22. 23. 24.

25.

26.

27.

28.

29.

Infection, Springer, AIDS and Behaviour, Volume 10, Number 1, January 2006 , pp. 35-40 Ybarra, M.L.; and Bull, S.S.; Current Trends in Internet and Cell Phone–based HIV Prevention and Intervention Programs; Current HIV/AIDS Reports 2007, 4:201–207 CareFirst Inc. Website; http://carefirst.staywellsolutionsonline.com/Relat edItems/40,HIVAIDSQuiz accessed 19 February 2009 American Social Health Association Website; http://www.ashastd.org/learn/learn_hiv_aids.cfm Accessed 19 February 2009 Tamocha, M.; Telecommunications in Botswana; Mmegi Monitor; 29 January 2007. Botswana Telecommunications Authority; Fact book; 2004 Parikh, T.S.; Mobile Phones may be the Right Devices for Supporting Developing World Accessibility, but is the WWW the Right Service Delivery Model?; May 2006, Edinburgh, UK IPOLETSE Call Centre for HIV AND AIDS, http://www.mri.co.bw/html/call-ipoletse.htm, Botswana Ministry of Health Shaik, F.; Thatcher, A. and Zimmerman C.; Attitudes of semi-literate and literate bank account holders to the use of automatic teller machines (ATMs); University of the Witwatersrand, South Africa; 15 September 2004. Baecker and Buxton, R.M. Baecker and W.A.S. Buxton, Readings in Human Computer Interaction A Multidisciplinary Approach, Morgan Kaufmann Publishers, California, 1987. Botswana’s Literacy Survey Report: The Report of the First National Literacy in Botswana, 1993 www.moe.gov.bw/dnfe/research/entry.htm Voice INT reference: www.wineexpertpro.com/voice_win_int_5_1.ht m

Morakanyane R.S. MSc

Page 107