Knowledge-Based Learning Environments for Construction - CiteSeerX

Journal for Education in the Built Environment, Vol.1, Issue 1, March 2006 pp. 51-72 ISSN: 1747-4205 (Online)

Knowledge-Based Learning Environments for Construction Mustafa Alshawi & Jack Steven Goulding: University of Salford, UK Ihsan Faraj: Manchester Metropolitan University, UK

Abstract Developments in information technology have revolutionised the delivery of online learning to such an extent that interoperability, scalability, adaptability and mass-customisation are now becoming practical solutions for pan global delivery. However, whilst advocates of virtual learning environments, and advanced learning management systems often extol the virtues of e-learning per se, they often fail to articulate the limitations of such systems, especially concerning the ‘personalisation’ of the learning process and incompatibility with pedagogic needs. This paper presents an analysis of previous research in the field of knowledge based learning environments using the context of construction as an exemplar. A conceptual framework for developing a fully-customisable knowledge-based learning environment is proposed which uses knowledge objects linked to an object oriented database, the concept of which embraces interoperability, intelligent tutoring, shareability (learning content), and an intelligent interface to manage advanced learning object metadata.

Keywords: Construction Training, Distributed Learning Environments, Interactive Learning Environments, Learning Strategies, Pedagogy

51 Copyright © 2006 CEBE

M. Alshawi, J.S. Goulding & I. Faraj: Knowledge-Based Learning Environments for Construction

Introduction Online learning has continued to develop and evolve, the market of which is currently worth around £2.6bn in Europe, and £16.7bn worldwide (Watson and Ahmed, 2004). Rapid technological advances have made online courses increasingly more accessible through the World Wide Web (WWW), and an increased proliferation of pedagogic-rich e-learning material has started to change society regarding training and learning (Santos and Ramos, 2004). The specific advantage e-learning has to offer is now considered an important tool for meeting the challenges of the 21st century, as new and developing technologies are revolutionising the way instructional content is presented, delivered, and shared. Online learning therefore presents learners with many greater opportunities to learn online, the provision of which is also much more likely to meet learners’ needs, as instructional content can now be specifically tailored to meet individual learner needs. Furthermore, this provision can now be extended to provide a range of different services in order to support online learning – from monitoring the learning process, through to managing assessment process. However, the process of defining and developing e-learning material for a virtual learning environment (VLE) or managed learning environments (MLE) is often expensive to produce (especially in a single context setting), making the return on investment difficult to quantify. One of the major challenges over the last few years has been to make learning resources 'interoperable' across different systems in order to improve efficiency (Boyle, 2003). This has raised several other issues, most notably, that of cost, as developing high quality content often consumes considerable resources, particularly learning developers’ time. A key emerging factor arising from this is, is that of protecting content through copyright and digital rights management (Santos and Ramos, 2004), especially as the ease of copying and distributing information (e.g. peer-to-peer file sharing) is increasingly prevalent in today’s society. There seems to be a growing dichotomy associated with online learning, specifically: bridging the gap between technology and pedagogy. Technology has advanced to such a level that resource issues are now becoming a focal point for discussion, and learning objects have been mooted as a possible solution due to their inherent characteristics of interoperability and reusability. However, certain limitations still exist concerning their adaptability and shareability (Watson and Ahmed, 2004), especially concerning technology standards (Advanced Distributed Learning Initiative (ADL), 2005; Learning Technology Standards Committee (LTSC), 2006). From a pedagogical perspective however, ardent supporters argue that pedagogy is far more important than technology, and that the process of learning should be the key locus, not just the technology itself. This paper introduces the central pedagogical issues associated with online learning, along with associated developments in instructional design theory, instructional design models, and recent pedagogical advances. These issues are then discussed from a technological perspective covering current understanding of learning objects, Learning Management Systems, and e-learning standards. A Knowledge Based Learning Environment conceptual

52 Journal for Education in the Built Environment Vol. 1, Issue 1, March 2006 Copyright © 2006 CEBE


framework is presented as a way forward to develop high quality personalised learning environments.

Pedagogical Developments in E-Learning Current e-learning literature encompasses several pedagogical theories, from constructivism, through to systems theory, socio-cultural and behaviourism. Much of these issues have historically centred on the distinct characteristics of these approaches and applicability to the e-learning domain. Early developments have been presented on experiential issues (Kolb, 1984), through to reflection and learning (Jarvis et al., 2003), and each of these theories have their relative strengths and weaknesses. For example, Bloom’s Taxonomy identified classification levels of intellectual behaviour associated in the learning process (Bloom, 1956); whereas, Piaget (1952) identified four stages of logical and conceptual learning growth; and Gagne (1965) determined the ‘Conditions of Learning’, outlining the relationship of learning objectives to appropriate instructional designs. More recently however, Conole et al. (2004) noted that from an e-learning perspective, there was very little evidence to link these models against pedagogical effectiveness i.e. mapping pedagogical approaches against any specific characteristics of learning. Technological developments have now made it much more important to concentrate on learning styles and preferences (Weber and Brusilovsky, 2001), and it is also considered important to specifically respond to learners’ needs (Dimitrova, 2003). In this context, the overall effectiveness of virtual learning still needs to be evaluated (Stonebraker and Hazeltine, 2004). Moreover, the next stage is to challenge learners’ using meta-cognitive techniques in order to facilitate individual preferences regarding inquiry learning and reflective thinking activities (Wen et al., 2004). These issues are currently being ‘mapped’ against technological developments, the process of which is creating very powerful knowledge-based learning environments that can be tailored to suit individual learner needs (blending learning content and learner styles). These approaches are expanding the overall learning experience, and are increasing the rate at which learning is achieved (Clark, 2003; Rosson and Carroll, 1996).

Instructional Design Theory Instructional Design (ID) theories focus on the learning process through the implementation of cognitive approaches i.e. the sequential steps for effective learning. They use a set of perceptions for determining the appropriate instructional strategies needed in order to enable learners to acquire instructional goals (Merrill et al., 1996). This approach can also prescribe a variety of instructional methods depending on the type and nature of the subject matter, individual learning style, assessment procedures etc. Five ID theories are presented for discussion – see Table 1. In principle, ID theories can provide a foundation for effective mapping between pedagogy and technology. However, Gagne's (1965) conditions of learning note that different instructional goals require different delivery strategies; which means that for a learner to effectively acquire a given set of instructional goals/skills, the type of content (a descriptive theory of knowledge) should be appropriately matched to the method of delivery in order to 53 Journal for Education in the Built Environment Vol. 1, Issue 1, March 2006 Copyright © 2006 CEBE


promote the learning process (a descriptive theory of strategy); and that the relationship between these two components should be carefully addressed through a prescriptive theory. However, Gagne's conditions of learning, Merrill's (1983) component display theory, and Reigeluth’s (1999) elaboration theory do not provide a sufficiently complete set of prescriptions that are capable of supporting a computerised learning environment. For example, Merrill's ID theory supports learning objects and computerisation, and identifies several types of knowledge objects, the interrelationships between them, and rules for sequencing knowledge objects – the approach of which also introduces different instructional transactions and rules for sequencing. This has been clearly demonstrated by the development of an instructional design expert system which includes rules for automatically configuring transactions given certain learner characteristics. Table 1 Key Instructional Design Theories Theory

Main Features

Conditions of Learning (Gagne, 1965; Gagne et al., 1992)

Signifies different types or levels of learning

Component Display Theory (Merrill, 1983)

Classifies learning into: content (facts, concepts, procedures, and principles) & performance (remembering, using, generalising)

Each type requires different types of instructions Outlines nine instructional events & their cognitive processes Learning tasks can be organised into a hierarchy to identify prerequisites that should be completed to facilitate learning at each level

Specifies four primary presentation forms: rules, examples, recall and practice Lessons consists of objectives followed by some combination of rules, examples, recall, practice, feedback, help, and learning tasks For a given objective and learner, there is a unique combination of presentation forms that results in the most effective learning experience

Elaboration Theory (Reigeluth, 1999)

Instruction is organised in increasing order of complexity for optimal learning

Instructional Transaction Theory (Merrill et al.,1991; 1996)

This is an extension to the conditions of learning (Gagne) and component display theory (Merrill)

Learners need to develop a meaningful context into which subsequent ideas and skills can be assimilated Proposes seven major strategy components; an elaborative sequence; learning prerequisite sequence, summary, synthesis, analogies, cognitive strategies and learner control

Consists of a descriptive theory of knowledge, a descriptive theory of strategy, and a prescriptive theory of instructional design Descriptive theory of knowledge consists of knowledge objects ‘learning objects’ and knowledge interrelationships Descriptive theory of strategy includes rules and messages for selecting, sequencing and representing knowledge objects Prescriptive theory consists of algorithms that enhance the descriptive theory of knowledge



Technological Developments in E-Learning Developing cost effective e-learning material is now considered as one of the most important factors, especially concerning the return on investment equation (Boyle, 2003). This intrinsic cost can often be compounded by the level of interactivity used, as the higher the level of interactivity, the greater the demand on resources (from a resource development perspective). Therefore, the cost of developing high levels of learner interactivity can often be quite considerable, with figures of around 300 hours for every 1 hour of e-learning content developed being quoted (Collaborativelearningsystems, 2005). Furthermore, there are other costs to consider e.g. the operational costs, content delivery costs, maintenance costs etc. One particular solution to this problem is the potential use (and reuse) of learning objects.

Learning Objects The concept of ‘learning objects’ are grounded in the object-oriented field of computer science, including the behaviourist, and cognitivist instructional paradigms. The key notions behind learning objects, are that they can be used to enable developers to build instructional components that can be used and reused in different (and multiple) learning contexts (Dahl and Nygaard, 1966; Reigeluth and Nelson, 1997). Learning objects can therefore be used by instructional designers in a variety of different combinations in order to create individual learning outcomes (or entire courses) by linking these objects together through an ‘intelligent interface’ to deliver instructional goals. Learning objects can thus be an effective way of creating content for e-learning environments, as they allow high rates of reusability using heterogeneous material from content providers. These objects can be independent (encapsulated), indivisible, interactive, reusable, and can also embody rich pedagogical techniques, such as interactive scaffolding, i.e. task definition with sequencing to match learners’ needs, (Rosson and Carroll, 1996). Furthermore, learning objects are able to support inheritance, i.e. support inheritance rules, methods, specifications etc. from their parent object, and can allow aggregation (i.e. be combined with others). This added functionality thereby affords learners the ability to ‘call’ a grouped learning object (or one specific learning object within a course), providing learners with unprecedented levels of unique customisation. This ‘knowledge pool’ of learning objects could therefore allow learners to pick specific learning objects independently of the way they were embedded into the learning environment (ARIADNE, 2005). However, “…a well-articulated set of principles, guidelines or model for their instructional use has yet to emerge” (McGee, 2003), and the structure and definition of learning objects still needs further clarification (Polsani, 2003). This must also be tempered by the reusability paradox limitations, as Wiley (2000) noted that while the most decontextualized learning objects are reusable in the greatest number of learning contexts, they are also the most expensive and difficult for instructional designers to reuse. It is therefore important “to develop common instructional models which can guide the instructional presentation of learning objects for personalized instruction….to dynamically adapt content to fit instructional objectives” (Martinez, 2000). This naturally embraces



learning orientations (pedagogy) and instructional design theory, in a learner-centric environment. The key principle behind this concept is to take a learning object and provide an electronic envelope, or 'wrapper' around it. The wrapper describes the overall structure of the object, and includes the descriptive metadata within it. The metadata is used to allow fast and effective searching to take place, in order to retrieve the appropriate learning objects, and the metadata tags are used to describe the instructional approach within the object. The overall effectiveness of learning objects can be further enhanced by data mining (Hanna, 2004; Coenen et al., 2004), to reinforce the principles of knowledge discovery. Learning objects are governed to some extent by the advanced distributed learning initiative (ADL, 2005). This initiative was set up in 1997 to manage learning initiatives and to develop standard protocols e.g. SCORM, and now extends to embrace other initiatives, such as the Learning Technology Standards Committee (LTSC, 2006) for technical standards – the remit of which also extends to include recommended practices and guidelines for learning technology. Whilst it can be argued that there has been a general insurgence of e-learning material developed to meet learners’ needs, these initiatives have not yet fully matured, especially in the field of self-adaptive e-learning systems (Atif et al., 2003). This field of research could provide substantial benefits to several stakeholders, especially where the delivery strategy can be matched to an intelligent interface in order to deliver personalised learner programmes.

Learning Management Systems A Learning Management System (LMS) is a software-driven platform which hosts blended learning content within a collaborative learning community. This is able to manage, track and report on learners’ performance, and is typically able to integrate with other systems e.g. administration, finance, student records etc. in order to facilitate automation. The functionality of a LMS varies considerably, from providing basic training management, through to delivering sophisticated enterprise-wide learning management systems that include competency management with enhanced pedagogical features. Current LMS include Blackboard©, Oracle© and WebCT©, all of which support a range of features, e.g., content assessment authoring, assembly of curricula from existing learning content, student collaboration tools, skills/gap analysis, etc. Watson and Ahmed (2004) identify four main aims of LMS, specifically: •

to deliver training content

•

to track participants’ performance

•

to manage online learning (courses and trainees)

•

to provide tools for student collaboration

Over the past few years, personalised learning has increasingly becoming more popular. This was reinforced by Sampson and Karagiannidis (2002), who identified that instruction should not just be restricted by time, place or any other barriers, but should be tailored to the



continuously modified individual learner’s requirements, preferences, background knowledge, interests, skills, etc. This signifies a radical departure from educational theory and technology, from ‘traditional’ interactive learning environments, to more personalised (and customisable) learning environments, and four key themes seem to be emerging. The first theme is that ‘traditional’ learning environments have adopted the ‘one-to-many’ learning model, whereas personalised learning environments are based on the ‘one-to-one’ or ‘manyto-one’ learning concept (i.e. one or more tutors for one learner). The second theme is that traditional learning environments usually pose a number of constraints in relation to the learning setting; personalised learning environments on the other hand, facilitate learning independent of time, location, etc. The third theme to note is that traditional learning environments are usually designed for the ‘average learner’; while, in personalised learning environments, the learning material and sequencing, learning style, learning media, etc, depend on the individual learner’s characteristics (and needs). Finally, the fourth theme to observe is that within the traditional learning environment, the curriculum, learning units etc, are determined by the tutor, while in the personalised learning setting, they are based on the learner’s requirements (i.e. self-directed learning). The high proliferation of LMS now presents many problems to content developers and providers - the most notable aspect of which is the general lack of interoperability between the various systems (Watson and Ahmed, 2004). Furthermore, as content development is a major cost, more efficient delivery approaches need to be assessed unilaterally, as continued development of different incompatible specifications by different organisations is likely to further hinder the integration of new technologies.

E-Learning Standards Standards play an important role in the innovation, development, evolution and adoption of any product. In the e-learning industry, organisations need therefore to protect their learning content and increase their return on investment. This precursor is increasingly being noted as a central issue (Boyle, 2003), the nuance of which is that if systems cannot grow, be sustained, maintained, and delivered efficiently and effectively, then this investment will be wasted (or be ineffective). Standards with the e-learning environment are therefore fundamentally important in this equation, as they allow a variety of products to co-exist. This convergence of technologies is essential, especially for e-content developers and end-users (consumers). Therefore, where products adhere to ‘agreed’ standards, these will in turn naturally provide consumers with wider product choices, and e-content developers (or providers of learning) stand a better chance that the products in which they invest will be precluded from immediate obsolescence. The e-Learning Consortium (2003) has identified certain standards and mandates in order to help promote the following issues: Interoperability •

to mix and match content from multiple sources (and within multiple systems)



•

to enable multiple systems to communicate, exchange, and interact transparently

Re-usability •

to enable content and code to be assembled, disassembled, and re-used (quickly and easily)

•

to enable content objects to be assembled/used in a context other than that originally designed

Manageability •

to allow systems to track appropriate information about the learner and content

•

to facilitate management of the complex selection and assembly of “tailored content”

Accessibility •

to enable learners to access appropriate content, at the appropriate time, on the appropriate device

Durability •

to ensure buyers are not ‘trapped’ by a particular vendor's proprietary learning technology

•

to ensure that no significant additional investment is required for re-usability and interoperability

Scalability •

to enable learning technologies to be configured in order to have expanded functionality to serve broader populations and organisational purposes

•

to enable an organisation’s return on investment in e-learning products to be increased, especially if they can be leveraged beyond their original scope

Affordability •

to ensure learning technology investments are economically viable and risk adverse

To date, there are no official standards in e-learning that content providers must adhere to – just a collection of disparate specifications by various bodies. This situation presents considerable challenges for e-learning developers. However, many leading organisations have now started to share these concerns and are collaborating to develop an official standard with interchange capability. This is likely to incorporate aspects of work undertaken by several organisations into one official International Standardisation Organisation (ISO) standard (Watson and Ahmed, 2004), and one key aspect of this progression, is that of the Shared Content Object Reference Model standard (SCORM).



The SCORM Standard The SCORM standard assumes the existence of a set of LMS services to: launch learning content, keep track of learner progress, establish the order (sequence) of learning objects to be delivered, and reports on learners’ progress. SCORM is not a standard itself, but rather a reference model that serves to test the effectiveness and real-life application of a collection of individual specifications and standards. It provides a foundational reference model upon which anyone can develop models of learning content for subsequent delivery. This framework details an implementation reference standard to enable content, technology, and systems using SCORM to communicate with each other (E-Learning Consortium, 2003). SCORM addresses two main issues: •

Content Aggregation Model: this includes how to put learning content together (so it can be imported and reused).

•

Run Time Environment: this describes how content is launched, and how learners’ progression is tracked and reported back.

The SCORM Content Aggregation Model (CAM) defines how learning content can be identified and described in order to facilitate discovery and re-use. It can therefore be used to specify how learning objects can be assembled from smaller information blocks and raw assets (such as single illustrations, animations, text or audio) and how these objects can be aggregated into a larger collection such as a lesson or course. The CAM also enables these objects to be moved from one LMS to another, or between Learning Content Management System repositories. The CAM has three primary building blocks: Raw Content Item, Sharable Content Object (SCO), and Content Aggregation Metadata. A raw content item is the most atomic form of learning content e.g. electronic representations of text, images, sound, Web pages, etc. Whereas, a SCO represents a collection of one or more raw content items, representing the lowest level of granularity of content that an LMS is able to track. Finally, the Content Aggregation Metadata is a hierarchal representation of a SCO, aggregated to form higher-level units of instruction. The SCORM Run Time Environment is used to enable learning resources to be shared across a LMS platform. This allows content to be interoperable, and permits learning resources to be ‘pooled’, thereby facilitating learning resources to communicate with the LMS in a predefined communication language. The SCORM Run Time Environment consists of three basic aspects: Launch Mechanism, Applications Program Interface (API), and Data Model (ADL, 2005). The Launch Mechanism defines a common way for the LMS to start the Web-based learning resources by defining the procedures and responsibilities for the establishment of commutation between the delivered learning resources and the LMS. In this context, the API is used as a communication mechanism to inform the LMS about the state of the learning resources i.e. retrieving setting data (e.g., score, time limits, etc.) between the LMS and the SCO. The final component (the Data Model) identifies and defines a standard set of data elements regarding the communication of information. This is used to facilitate data transfer between the LMS and SCO i.e. what each need to know about each other (protocols). The LMS is therefore used to maintain the state of the required data



elements across each session, thereby ensuring the learning content utilises only the permitted predefined data elements (if reuse across multiple systems is to occur).

Knowledge-Based Learning Environments There has been significant progress in research and development of e-content, standards, and e-learning programmes. However, these have not fully attained their potential due to a variety of factors – including: 1) the lack of exchangeability between learning materials; 2) the delivery mechanisms being incompatible with the pedagogical design; 3) relatively low levels of student interaction insensitive to pedagogic learning processes; 4) the absence of intelligent online programme advice and guidance; 5) overall inflexibility – to meet learners’ diverse needs; 6) ineffective implementation strategies; 7) technology being used mechanistically, rather than being an informatics driver. These issues have culminated in a catalogue of intransigence, and have stifled developments. In this context, future knowledge-based learning environments should address these issues directly, providing learners with unparalleled levels of support and guidance, tailored specifically to their pedagogic learning style and preferences.

Knowledge Based Learning Environment: Definition and Features A Knowledge Based Learning Environment (KBLE) is a flexible and robust computerised learning environment which provides highly relevant and cost-effective personalised learning. It can be used to deliver individual, bespoke, and highly personalised training matched to learners’ needs. The KBLE approach thereby enables: •

Accessibility: to content over a wide range of delivery platforms (including mobile technology)

•

Interactivity: with learning content (including assessment)

•

Maintainability: supports high levels of content maintenance

•

Shareability: enables content between various environments/ organisations to be exchanged

•

Flexibility: to adapt to various delivery strategies

•

Scalability: to enable content to be scaled easily and appropriately

The constitution and delivery aspects of a KBLE should address the parameters of (and relationships between) learners’ needs and requirements; learners' learning process, and the learning/knowledge repository - the approach of which is identified in Figure 1. This is an important dimension, as learners are continually searching for opportunities that fulfil their individual learning needs, and in this context, are influenced by their individual histories and preferences. This mandate would need to embrace learners’ access and exit points within the learning cycle, their level of desired attainment, including the availability and acceptability of the learning experience. Furthermore, learning traits often differ by preferred learning style and environment as well as by motivational and transformational factors. Thus, attaining a match between learner needs and the ‘supply side’ will require replacing traditional educational approaches with alternatives that emphasise the primacy of the 60 Journal for Education in the Built Environment Vol. 1, Issue 1, March 2006 Copyright © 2006 CEBE


learning materials architecture in order to deliver effective personalised learning programmes. An alternative approach is therefore needed which addresses these issues (by reversing) the traditional top-down order of designing training programmes, to enable a bottom-up aggregation system to be developed. This would focus directly on learners’ needs, and would be capable of leveraging advantage through a constructed ontology which enables learners to search and discover specific and relevant knowledge (to them) through a personalised training programme.

Knowledge Repository

Learner Centric

Learners’ Needs & Requirements

Learning Process

Figure 1 Delivery Trichotomy From a delivery perspective, the learning opportunity is related to the capacity of the learning process to facilitate the achievement of the specified learning outcomes. Therefore, the performance of the selected learning process is predicated on the effectiveness of the underpinning instructional design theory, the nuance of which is represented by a process model that defines the sequence of activities, operations and tasks. Thus, the learners' learning process must consider how 'optimal' learning can occur, and a prerequisite of this is to match learning outcomes with the most appropriate learning and assessment strategy, thereby providing learners with individual and personalised feedback/progress reporting facilities. Finally, the efficiency of the delivery process depends on the economies to be achieved in creating, searching, reproducing, assembling and delivering the requisite body of knowledge in support of the learning experience. These issues are fundamentally important, as there is a substantial investment cost associated with creating learning materials in a usable and acceptable format. In this context, knowledge representation is now becoming the focal point for discussion. At micro level (minimal size), knowledge representations can be described with a high degree of precision, thereby enabling a training portfolio to be assembled in the



form of a modularised and customised programme. This has prompted the use of learning objects, and the formulation of common object standards. Taken one step further, the creation of a learning/knowledge repository would constitute the assembly and relationship of these knowledge objects (as a reusable digital resource), the process of which would naturally embrace the relationship of object technology with pedagogy and learning. This approach would thereby enable the key characteristics of knowledge objects to be meaningfully tailored, matched, and applied to individualised learner packages.

Knowledge Based Learning Environment: Three-Parameter Relationship There is a strong and fervent relationship between the knowledge to be acquired by learners, the learner’s learning process, and the learners’ needs and requirements. This is a ‘learner centric’ environment, which is dynamic and which changes and adapts from one learner to another (as each learner’s needs can not be efficiently catered for with one generic blanket pedagogic strategy), or as the learner progresses through the learning cycle (e.g. from one stage/level to another). In this context, the knowledge repository is the main engine for the provision of relevant content to learners – the design of which should be robust, yet flexible enough to deal with a variety of learners’ needs. For example, it must be able to have the capacity to efficiently ameliorate (and respond to) the relevant delivery strategies and all the interrelated activities. However, the knowledge repository should be treated independently from any specific delivery strategy in order to give it the flexibility to respond to other related delivery strategies without significantly changing its structure. This indicates that the there is a strong bond between delivery strategies (instructional design theories) and the design of the knowledge repository. This relationship can be adequately captured and facilitated by an object technology approach – the mechanics of which have been significantly developed over the past decade, and various structured methodologies have now been developed for the design and development of object oriented environments. However, whilst these methodologies are fully capable of balancing captured data (objects) and their usage (processes), the majority of these developments have been predominantly business-focused, and have not been developed for the cognitive field of pedagogy. The following section outlines a methodology for the design and development of a knowledge based learning environment.

Knowledge Based Learning Environment: A Built Environment Conceptual Methodology Using the built environment as a context, the KBLE could be used to ‘build’ courses; from single learning outcomes, through to complex interlinked teaching and learning packages. For example, consider the topic of surveying as a context. Naturally, this is a very wide area, covering such issues as mathematics, engineering principles, instrumentation etc. The first part of the development process would be to select a small finite area to focus upon e.g. ‘accuracy’. This would then be used as a basis for developing the course material at learning object level, with individual learning outcomes identified at each sub-element of this topic



area (e.g. precision, tolerances, errors etc). From an instructor or courseware developer’s perspective, each learning outcome would then need to be matched with an ‘appropriate’ pedagogical technique. Each of these sub-elements would then be rigorously tested with ‘learner sets’ and domain experts to ensure the desired outcomes are achieved. When the testing process has been completed, the context and sequencing is set, and all the learning objects are then bound together with a ‘wrapper’ for the object-oriented database. The concept of the methodology is drawn from a balanced approach between object technology methodologies and Instruction Design Theories identified in Table 1. This conceptual methodology can be seen in Figure 2. The left side of the diagram reflects the design and development of the knowledge objects, while the right hand side represents the ID theory delivery strategies. The concept is represented by unfolding four layers of developments (from top to bottom): the domain analysis layer; specification layer; design layer; and implementation layer. The first layer aims to establish the specific training programme(s) and the relevant training priorities and content – identified from analysing the learners’ needs. The second layer (specification) involves the identification of the discipline pedagogical requirements, expected learning outcomes, learning targets and the associated relevant delivery strategies. The third layer (design) aims to produce the design of the ontological structure, i.e. the data schema of the content, as represented by knowledge objects and their relationships, along with the most effective presentation of knowledge to learners – as defined by the sequencing of knowledge objects and the contexts within which the knowledge will be presented.

Knowledge Domain

Learning Strategies

Knowledge Content

Specification Layer

Sequencing & Context 1

Wrapper

Binding

OO Hierarchy

Delivery Strategies

Relationship

Domain Analysis Layer

Tailored Context

Sequencing & Context 2

Design Layer

Sequencing & Context n

Interface

SCORM Intelligent OO integrated learning environment

Implementation Layer

OODB

Figure 2 Conceptual Design and Development of the KBLE



The knowledge objects would then be formulated with wrappers (using metadata) which provide guidelines on the behaviour of the objects with regard to the learning outcomes, delivery, and learning context. This will facilitate their manipulation to assemble tailored training programmes. The fourth layer (implementation) is used to create knowledge object classes and their hierarchy to satisfy multiple abstractions i.e. capable of responding to various needs of learners and learning styles. Object classes will be implemented in an object oriented environment with interfacing standards protocols such as SCORM to facilitate the portability of the contents to multiple delivery platforms such as the learning management systems. Once the object oriented environment is populated with sets of learning objects, learning outcomes, and delivery strategies, learners will then be supplied with a dynamically unfolding sequence of learning objects that will facilitate the completion of their personal training programmes. For example, using the analogy of the surveying context; the learner would be iteratively diagnosed and presented with learning material dynamically tailored to suit their individual needs (not the generic needs of multiple learners).

Knowledge Based Learning Environment: Implementation Framework The conceptual methodology should be implemented using the principles drawn from the Instructional Transaction Theory (Merrill et al., 1996) and the principles of the "Collaborative System Design" approach suggested by the ADL Guidelines – see Figure 3.

Using ADL Guidelines

Using ADL Guidelines

Analyse

Specify & Design

Implement Evaluate

SCORM Standards

Finished product/ Performance impact

Figure 3 Collaborative System Design Approach Using a built environment context as an exemplar; the design and implementation of this methodology should consider five key steps: Evaluation of the Knowledge Domain (built environment); Design and Development; Implementation; Evaluation and Feedback; and Prototyping. The technical design and implementation of the KBLE and the instructional architecture (Instructional Design Theory) should be considered concurrently in each of these five stages. Furthermore, continual coordination and feedback should be constantly



assessed to ensure that a balanced approach between technology and instructional design theory is achieved.

Step 1: Evaluation of the Knowledge Domain: Built Environment Aim: Defining Learning Outcomes and Learning Strategies This step (see Table 2) should be undertaken using a range of domain experts and core stakeholders. The constitution of this panel would naturally embrace the following personnel: construction industry representatives; clients; policy-makers (professional institutions, governmental bodies etc); subject specialists (domain experts, academia etc); providers of econtent (e-learning specialists); publishers (software); and technology providers (multimedia and communication industries). These workshops should be supplemented with a series of interviews and discussion groups to achieve the following: •

Define the subject area for training (domain knowledge)

•

Define the training priorities

•

Define the associated pedagogies and learning outcomes

•

Identify and match the instructional design theories

This process would ameliorate and match learners’ needs to the subject material in order to maximise success. For example, it may be more appropriate after the observations have been undertaken to use a different pedagogical technique than first anticipated e.g. signposting rather than exploratory. These observations help form the structure of the learning environment vis-à-vis content analysis, performance outcomes and measurement. Table 2 Evaluation of the Knowledge Domain Learning Environment Design Aims

Instructional Design Theory Aims

Determine the structure of the environment

Conduct needs assessment

Identify the learner taxonomy

Perform content analysis

Establish standards and protocols

Identify performance objectives and outcomes

Establish security standards

Develop performance measures

Step 2: Design and Development of the KBLE Aims: Determine Delivery Strategy and Supporting Activities; Create Knowledge Objects Hierarchy This step (see Table 3) should be carried out in order to establish the delivery strategy needed to effectively achieve the predefined learning outcomes (from step 1). The elements of this strategy need to embrace the following issues: •

Sequencing of knowledge objects

•

Identification of the context which each sequence of objects delivers

•

Determination of the entry requirements for each context

•

Identification of the level and type of learners' interaction 65 Journal for Education in the Built Environment Vol. 1, Issue 1, March 2006 Copyright © 2006 CEBE


•

Establishment of the feedback policy

•

Identification of the knowledge portrayal process and type of the required delivery support

•

Metadata of the knowledge objects which facilitate the communication, identification and manipulation of objects within the environment

Once the delivery strategy is determined, the next stage is to define the reusable knowledge objects and hierarchy. The Unified Modelling Language (UML) approach and its associated “User Cases” can be used to model the objects, their behaviour, structure and relationship with each other. The creation of these objects should be based on specifications such as those defined by the joint Advanced Distributed Learning Co-Laboratory (ADL, 2005); definitions of learning objects by the Learning technology Standards Committee (http://ieeeltsc.org/ [2006]); instructional content objects (Merrill et al., 1996); learning object criteria by Cisco's systems (Barron, 2000) etc. This will enable clear specifications for the total learning environment to be identified (including its behaviour), and would include selecting an appropriate LMS which best matched the delivery strategy; the communication protocols between the LMS and the knowledge objects database; the underlying Web and communication technology standards; and specifications for an intelligent interface that facilitated the delivery of personalised programmes. Table 3 Design and Development of the KBLE Learning Environment Design Aims


Identify an appropriate learning management system

Identify the delivery strategy & the supporting activities

Identify the required Web technologies

Determine knowledge object sequence performance objectives

Allocation of functions to instructors & computers

Specify context for the selected learning strategies

Design the environment and interface

Design knowledge objects

Determine level of interactivity with learners

Design instructional knowledge hierarchy

Specify metadata (in line with standards e.g. SCORM)

Step 3: Implementation of the KBLE Aims: Develop an OODB; Implementation of the Intelligent Interface In this step (see Table 4), the previously developed 'User Cases' and their findings would be used to develop 'learning scenarios', which would lead to the identification of the main learning environment’s functions. These functions would facilitate the operation of the environment for both the instructors and the learners, and would establish the level of interaction between learners and the KBLE. Communication standards (e.g. SCORM) should be adopted in the implementation of the knowledge objects in order to ensure interoperability. This process should be accompanied by the implementation of the intelligent interface, which would be the main vehicle for communication with the knowledge objects (in



an Object Oriented environment) and the selected LMS. The technology used for the interface would be capable of satisfying all communication requirements – being compliant with all the Internet-based standard communication protocols, and could facilitate the implementation of a distributed object oriented environment (where shareability and exchangeability of objects are the main requirements). Table 4 Implementation of the KBLE Learning Environment Design Aims


Develop an Object Oriented Database (OODB)

Provide information to populate the database

Populate the database

Specify and design the interface

Develop an intelligent interface

Perform preliminary tests and feedback with domain experts

Link to learning management system (if necessary)

Step 4: Evaluation and Feedback Aim: Evaluate the KBLE Step four (see Table 5) should be carried out using a representative sample of learners, the aim of which is to evaluate the KBLE on a full training cycle. The criteria developed by the Institute for Simulation and Training (2005) could be used to measure the quality of instructional materials, user interface, user satisfaction, organisational impact, and overall learning effectiveness; and extend to embrace the ADL checklist (http://www.jointadlcolab.org/) and other such metrics. Feedback from this step should then be analysed and taken into the final prototyping stage. Table 5 Evaluation and Feedback Learning Environment Design Aims


Design checks

Evaluate construction (formative and summative)

Use of feedback Software log file analysis

Step 5: Prototyping the KBLE Aim: Prototype the KBLE The final prototyping step (see Table 6) should be undertaken in accordance with the guidelines identified in Figure 3, starting from the analysis of requirements (Step 1) to evaluation (Step 4). Feedback from Step 4 should be incorporated into the design and development of the knowledge structure, which would require mapping into the OODB. This development would then be re-evaluated, and the prototyping cycle repeated until the prototype learning environment was complete.



Table 6 Prototyping the KBLE Learning Environment Design Aims


Modify functionalities of the KBLE

Modify sequencing activities

Modify knowledge object model

Modify knowledge portrayal

Modify user interface

Modify learning cycle

Produce working prototype

From a learner’s perspective, the adoption of the KBLE would provide significant benefits, as the learning style would be tailored to suit individual preferences, making it pedagogically enhanced to suit their needs. Furthermore, as the introduction and sequencing of learning material and activities is matched to the learner’s progress, the rate and pace at which learning takes place can be better monitored and assessed. This approach not only maximises the learning process, but also allows reflective learning to be embraced (Brockbank and McGill, 1998), specifically, developing reflective practice skills beyond academic learning into professional practice.

Conclusion The significant value of the e-learning market (Watson and Ahmed, 2004), along with the advances in technology, has increased the demand for new and innovative approaches to design and develop cost-effective and high quality e-learning environments which can efficiently respond to learners’ needs and requirements. Over the past decade, research has attempted to address key areas in this field, such as, automation of the learning process; improving the portability of e-learning materials; pedagogy; learning objects; and e-learning standards. Many organisations have been created to further develop and maintain these areas, e.g. the Learning Technology Standards Committee, Advance Distributed Learning, and the ARIADNE Foundation. However, addressing pedagogy in learning environments has still not been fully addressed; yet, ‘mapping’ pedagogical requirements against technological developments have the potential to create powerful knowledge-based learning environments that can be tailored to suit individual learner needs (blending learning content and learner styles). This paper explained the relationship between pedagogy and technology in the context of the design and development of personalised e-learning environments. Instructional design theories were utilised as a vehicle to map pedagogy with technology – culminating in the development of a conceptual model. The technology issue was addressed in terms of learning objects and standards together with their role in providing interoperability between delivery platforms, reusability and manageability of e-learning materials and accessibility. Whereas, from a pedagogical perspective, the design and implementation of a knowledge based learning environment was aligned to pedagogical requirements using the concept of learning objects. This balanced approach between object technology and ID theory was presented by unfolding four layers of developments; the domain analysis layer, specification layer, design layer and implementation layer.



The methodology was supported by an implementation framework which was drawn from the Instructional Transaction Theory (Merrill et al., 1996) and the principles of the "Collaborative System Design" approach, as identified by the ADL Guidelines. This conceptual framework identifies five steps: evaluation of the knowledge domain, design and development of the KBLE, implementation of the KBLE, evaluation and feedback, and prototyping the KBLE. Each of these steps has clear aims and deliverables (in order to make the implementation process easier to control and manage). This methodology is part of the overall efforts to further understand the relationship between pedagogy and technology, in an attempt to procure highly personalised learning environments which can dynamically respond to learners’ pedagogical requirements – especially concerning learner expectations and learner process design (Ul-Haq et al., 2003). Furthermore, this approach could offer developers and learning content providers in the built environment new opportunities to embrace unparalleled levels of content reusability, shareability and mass customisation. The learnercentric approach puts the learner first, and in doing so, maximises the trichotomy of: learning, knowledge, and learner needs.



References ADL. (2005). Advanced Distributed Learning Initiative. URL: http://www.adlnet.org/ [Accessed 4 July 2005]. ARIADNE. (2005). ARIADNE Foundation for the European Knowledge Pool. URL: http://www.ariadne-eu.org/ [Accessed 4 July 2005]. Atif, Y., Benlamri, R. & Berri, J. (2003). Learning Objects Based Framework for Self-Adaptive Learning. Journal of Education and Information Technologies, 8 (4), 345-368. Barron, T. (2000). Learning Object Pioneers. URL: http://www.learningcircuits.org/2000/mar2000/barron.html [Accessed 4 July 2005]. Blackboard®. URL: http://www.blackboard.com/ [Accessed 4 July 2005]. Bloom, B. S. (Ed.). (1956). Taxonomy of Educational Objectives: The Classification of Educational Goals, Handbook I: Cognitive Domain. New York: McKay. Boyle, T. (2003). Design Principles for Authoring Dynamic, Reusable Learning Objects. Australian Journal of Educational Technology, 19 (1), 46-58. Brockbank, A. & McGill, I. (1998). Facilitating Reflective Learning in Higher Education, Society for Research into Higher Education. UK: Open University Press. Collaborativelearningsystems. (2005). URL: http://www.collaborativelearningsystems.com/co/publications/elearningcost.pdf [Accessed 21 March 2006]. Clark, A. (2003). Natural-Born Cyborgs: Minds, Technologies, and the Future of Human Intelligence. Oxford: Oxford University Press. Coenen, F., Goulbourne, G. & Leng, P. (2004). Tree Structures for Mining Association Rules. Journal of Data Mining and Knowledge Discovery, 8 (1), 25-51. Conole, G., Dyke, M., Oliver, M. & Seale, J. (2004). Mapping Pedagogy and Tools for Effective Learning Design. Journal of Computers and Education, 43 (1-2), 17-33. Dahl, O. J. & Nygaard, K. (1966). SIMULA - An AGOL Based Simulation Language. Communications of the ACM, 9 (9), 671-678. Dimitrova, V. (2003). STyLE-OLM: Interactive Open Learner Modelling. International Journal of Artificial Intelligence in Education, 13, 35-78. E-Learning Consortium. (2003). Making Sense of Learning Specifications and Standards a Decision Maker's Guide to their Adoption. 2nd Ed., November, 2003, The Masie Center. URL: http://www.masie.com/standards/s3_2nd_edition.pdf [Accessed 4 July 2005]. Gagne, R. M. (1965). The Conditions of Learning. New York: Holt, Rinehart & Winston.



Gagne, R., Briggs, L. & Wager, W. (1992). Principles of Instructional Design. 4th Edition, Holt, Rinehart, & Winston. Hanna, M. (2004). Data Mining in the E-Learning Domain. Campus-Wide Information Systems Journal, 21 (1), 29-34. Institute for Simulation and Training. (2005). URL: http://www.ist.ucf.edu/overview.htm [Accessed 4 July 2005]. Jarvis, P., Holford, J. & Griffin, C. (2003). The Theory and Practice of Learning. London: Kogan Page Ltd. Kolb, D. (1984),. Experiential Learning: Experience as the Source of Learning and Development. New Jersey: Prentice Hall. Learning Technology Standards Committee. (2006). URL: http://ieeeltsc.org/ [Accessed 21 March 2006]. Martinez, M. (2000). Designing Learning Objects to Mass Customize and Personalize Learning. URL: http://reusability.org/read/chapters/martinez.doc [Accessed 4 July 2005]. McGee, P. (2003). Learning Objects: Bloom’s Taxonomy and Deeper Learning Principles. Association for the Advancement of Computing in Education Conference [E-Learn 2003]. Phoenix, Arizona, USA, Nov., 2003 URL: http://educ3.utsa.edu/pmcgee/nlii/McGee_eLEARN_2003.doc [Accessed 4 July 2005]. Merrill, M. D. (1983). Component Display Theory. In Reigeluth, C. M. (Ed.). Instructionaldesign theories and models: An overview of their current status. Hillsdale, New Jersey: Lawrence Erlbaum Associates. Inc. Merrill, M. D., Li, Z. & Jones, M. (1991). Instructional Transaction Theory: An Introduction. Journal of Educational Technology, 31 (6), 7-12. Merrill, M. D., Drake, L., Lacy, M. & Pratt, J. A. (1996). Reclaiming Instructional Design. Journal of Educational Technology, 36 (5), 5-7. Oracle©. URL: http://www.oracle.com/applications/human_resources/intro.html [Accessed July 2005]. Piaget, J. (1952). The Origins of Intelligence in Children. New York: International Universities Press. Polsani, P. R. (2003). Use and Abuse of Reusable Learning Objects. Journal of Digital Information, 3 (4) URL: http://jodi.ecs.soton.ac.uk/Articles/v03/i04/Polsani/ [Accessed 4 July 2005]. Reigeluth, C. M. & Nelson, L. M. (1997). A New Paradigm of ISD? In Branch, R. M. & Minor, B. B., (Eds.). Educational Media and Technology Yearbook. Englewood, CO, Libraries Unlimited, Inc.



Reigeluth, C. (Ed.). (1999). Instructional Design Theories and Models: A new paradigm of Instructional Theory. Hillsdale, NJ: Lawrence Erlbaum Associates. Rosson, M. B. & Carroll, J. M. (1996). Scaffolded Examples for Learning Object-Oriented Design. Communications of the ACM, 39 (4), 46-47. Sampson, D. & Karagiannidis, C. (2002). Personalised Learning: Educations, Technological, and Standardisation Perspective. Interactive Educational Multimedia, 4, 24-39. Santos, O. A. & Ramos, F. M. S. (2004). Proposal of a Framework for Internet Based Licensing of Learning Objects. Journal of Computers and Education, 42 (3), 227-242. SCORM. (2004) URL: http://www.adlnet.org/scorm/index.cfm [Accessed 4 July 2005]. Stonebraker, P. W. & Hazeltine, J. E. (2004). Virtual Learning Effectiveness: An Examination of the Process. The Learning Organization Journal, 11 (3), 209-225. Ul-Haq, R., Stiles, J. & Pond, K. (2003). Learning Expectations and Learning Process Design. Active Learning in Higher education, 4 (2), 168-180. Watson, J. & Ahmed, P. K. (2004). Learning in the Age of Global Information Technology: Development of a Generic Architecture for an Advanced Learning Management System. Campus-Wide Information Systems Journal, 21 (1), 4-21. WebCT. URL: http://www.webct.com/ [Accessed 4 July 2005]. Weber, G. & Brusilovsky, P. (2001). ELM-ART: An Adaptive Versatile System for Web-based Instruction. International Journal of Artificial Intelligence in Education, 12, 351-384. Wen, M. L., Tsai, C., Lin, H. & Chuang, S. (2004). Cognitive-Metacognitive and ContentTechnical Aspects of Constructivist Internet-Based Learning Environments: A LISREL Analysis. Journal of Computers and Education, 43 (3), 237-248. Wiley, D. (2000). Connecting learning objects to instructional design theory: a definition, a metaphor, and a taxonomy. In Wiley D. A. (Ed.). The instructional use of learning objects Online version. URL: http://reusability.org/read/chapters/wiley.doc [Accessed 21 March 2006].
