2011 11th IEEE International Conference on Advanced Learning Technologies
A Semantic Web-enabled Tool for Self-Regulated Learning in the Workplace Melody Siadaty1,2, Jelena Jovanović3, Kai Pata4, Teresa Holocher-Ertl5, Dragan Gašević2,1, Nikola Milikić3 1
Simon Fraser University, Canada 2 Athabasca University, Canada 3 University of Belgrade, Serbia 4 Tallinn University, Estonia 5 Centre for Social Innovation/Technology and Knowledge, Austria
[email protected],
[email protected],
[email protected],
[email protected],
[email protected],
[email protected] based on the generalization of worker’ experience, and the guidance that individual workers receive from the organizational knowledge. However, this model does not target the importance of SRL in the workplace as a motivational element. According to the elaborated knowledge building model [7], SRL in an organizational context should also consider that in order to perform intrinsically motivated learning, learners have to align their learning activities to (i) their organizational learning goals, (ii) the learning activities of other members of the organization and (iii) their own learning goals. To be able to do this, the learners need personalized information – i.e., information relevant to them and their present learning context – about the organization’s objectives and expectations; the learning activities and achievements of co-workers; and learners’ own progress w.r.t. their current learning goal(s). However, access to this kind of information is far from straightforward, primarily due to the fact that today’s knowledge workers often use diverse tools for their everyday working and learning practices; thus the traces and outcomes of their activities are dispersed among different and often heterogeneous tools. To be turned into information relevant for learners, the traces (i.e., data) about learners’ learning activities and created knowledge objects (KOs) have to be structured, organized and well annotated, so that they can be (re-)discovered and (re-)used inside the organization [8]. To address this challenge, we propose the use of a network of ontologies since ontologies have proven as highly suitable for integrating data originating from different, often dispersed and heterogeneous sources [9]. They are, among other things, an excellent means for the integration of: 1) traces about individuals’ activities across diverse tools and services, and 2) individuals’ knowledge into shared, organizational knowledge. [10]. Specifically, the network of ontologies that we developed within the IntelLEO EU project [11] enables formal representation, and seamless integration of data about individuals’ learning experiences (i.e., learning activities and their context), knowledge being shared as well as different kinds of annotations (tags, comments, ratings and the like) that capture either personal or community reflections on the shared content/knowledge. By enabling this, the ontologies provide the basis for all the functionalities of Learning Pal – our tool aimed at scaffolding SRL processes in personal learning spaces. Particularly, via Learning Pal,
Abstract—Self-regulated learning processes have a potential to enhance the motivation of knowledge workers to take part in learning and knowledge building activities, and thus contribute to the resolution of an important research challenge in workplace learning. An equally important research challenge for successful completion of each step of a self-regulatory process is to enable learners to be aware of characteristics of their organizationallyembedded learning context. In this paper, we describe how a combination of pedagogy and Semantic Web-based technologies can be utilized to address the above two challenges. Specifically, we demonstrate the proposed solution through the Learning Pal tool which leverages ontologies to support self-regulation in organizational learning. Keywords-ontologies; learning planning
personal
&
organizational
learning;
I. INTRODUCTION Learning in the workplace is commonly informal and autonomous [1]. Ideally, this would mean that an individual knowledge worker is aware of his/her learning needs based on the requirements of a task, project, duty or any other of his/her responsibilities in the organization. At the same time, (s)he should be able to define relevant learning goals and has enough motivation to get engaged in proper collaborative learning and knowledge building activities to reach these goals, reflect upon and share how (s)he gained the required knowledge, so there would be less hassle for others in need of the same competences. This ideal image of motivated and proactive individuals, however, rarely happens in everyday work environments. Unless provided with structured learning scenarios in formal settings, most people are not proactive enough to initiate a learning process or simply do not know how to learn [2]. Self-regulated learning (SRL) [3] contains the motivational elements to address this challenge. Empirical research brings evidence that, individuals who are oriented towards the self-regulated enhancement of competences and the excellence within their tasks, show high intrinsic motivation, high task persistency, and high selfefficacy beliefs [4][5]. In our work, we consider SRL in the workplace as a part of organizational learning. In particular, we base our research on a well-known organizational knowledge building model proposed by Nonaka and Takeuchi [6]. This model highlights the renewal of organizational level norms and visions 978-0-7695-4346-8/11 $26.00 © 2011 IEEE DOI 10.1109/ICALT.2011.27
66
and demonstrate how each of them contributes to supporting SRL in the workplace. To better illustrate the harmonizing nature of our solution, we split these functionalities into three main sets (based on [7], cf. Sec. II): a) harmonization with organizational goals, b) aligning to organizational members’ learning goals, and c) aligning to individual learning goals.
knowledge workers can benefit from recommendations to better identify their competence gaps; find the most fitting learning strategies, extracted from the organizational memory, to reach the missing competences; monitor their learning progress, share and document their learning experiences and, compare their self-observed performances against organizational benchmarks and/or the performance of their colleagues. In this paper, we first elaborate on the pedagogical foundation of our work (Sec. II) and describe the functionalities of Learning Pal (Sec. III). Then, the role of ontologies in data integration is explained and exemplified (Sect. IV). After presenting the results of the initial evaluation of Learning Pal, we conclude the paper by comparing our work to the related work and by outlining directions of the further research.
A. Harmonization with Organizational Goals Contextual Recommendation of Available Competences. Organizational objectives are reflected through organizationally established and valued competences. In Learning Pal, users can view these competences from different perspectives (e.g., duties, roles, projects), each reflecting a need for the user to achieve a particular competence (e.g. to obtain a competence that is a prerequisite for a certain duty). By providing visual and textual clues along those competences, the tool aids users to identify those competences that are of particular importance to them. For instance, under the Projects category, a learner would see the available competences, prioritized based on their relevancy for the project(s) the user participates in (Fig.1 A). Similarly, the Duties and Roles categories opened by the user prioritize the competences that are prerequisites for the duties and roles assigned to each individual. This functionality supports the forethought phase of SRL [3], and accordingly helps the user to plan and define proximal learning goals, choose competences for these learning goals, and bring these goals in accordance with the organizational objectives. Recommendation of Learning Paths, Learning Activities and Knowledge Assets. A learning path is comprised of a sequence of learning activities along with descriptions (metadata) of knowledge assets required for performing those activities. By recommending learning paths, along with the contextual recommendation of available competences, Learning Pal supports users in developing strategies to reach a certain competence (i.e., the forethought phase of SRL). However, learners are not limited to the paths recommended by the tool; they can adapt any suggested learning path to their learning needs by selecting and adding learning activities and knowledge assets that they wish to perform, e.g. only those of a higher priority, or adding new ones created by themselves. Once documented as a learning path used to achieve a certain competence, such a user-(re)defined learning path is saved in the organizational memory, so that it can be later pulled and recommended to other users. Our ontology framework (cf. Sec. IV) significantly assists in this task by semantically aggregating learning activities from the tools in which they are preformed (e.g., an LMS) and associating them with the learner’s personal learning profile.
II. PEDAGOGIGAL FRAMEWORK The majority of conventional interpretations of SRL are based on an individualistic perspective [12], where the impact of the social context is often assumed to be less significant than individual-based factors. Such perspectives contradict the nature of the workplace, where individuals’ work and learning activities are highly social and community centered. The role of the social context becomes bolder when it comes to defining and evaluating learning goals, adapting one’s strategies and actions to social/organizational norms, and receiving incentives or experiencing inhibitors from the communities the learner belongs to. In our research, we have extended the approach proposed in [7] with SRL practices (i.e. aligning to one’s own learning goals), social embeddedness of workplace learning (i.e. aligning to the learning goals of other members of the organization), and harmonization of individual leaning goals with those of the organization. All these extensions play an important role in enhancing the motivation of individuals to take part in learning and knowledge building activities. This further implies that while planning his/her learning goals, a learner should harmonize his/her goals with organizational goals and expectations; at the same time (s)he should be aware that his/her individual learning contributions will benefit the whole organization's performance. The learner should also monitor the learning process of other learners (his/her colleagues), and rely on their successful learning activities as a guidance for his/her own learning process. Being aware that other people know about, and make use of his/her learning success increases a learner’s motivation. Also, knowing that the gained knowledge is of value to others would heighten the learners’ sense of self-efficacy and further motivate them in pursuing their learning process. Moreover, awareness of the utility of the shared knowledge is one of the major factors affecting one’s motivation in imparting knowledge within an organization [8]. Finally, a learner has to monitor his/her own progress and make corrections w.r.t. the planned goals. There is an extensive evidence in the literature that encouraging learners to systematically monitor their own performance (either through self-recording or external recoding) positively affects their skill acquisition, motivation, and self-efficacy [13][3].
B. Aligning to Organizational members’ Learning Goals Knowledge Sharing Profiles. Users can monitor the extent to which they share their learning experiences, e.g. defined learning goals, acquired competences, performed learning activities, etc., within the organization and compare it with those of other users within the same group, project, or the entire organization. They can also explore the extent to which others in the organization have used their shared learning experiences. This functionality supports the self-reflection phase of SRL as well as the social embeddedness of learners in an organization (cf. Sec. II).
III. LEARNING PAL Here, we describe the main functionalities of Learning Pal
67
nizational roles) who have successfully finished this path or a revision of it, and their average completion times. Comments and ratings of the colleagues who have already accomplished a certain learning activity plus statistical information such as their average completion time in finishing the activity further helps a user to decide whether to include this activity as part of his/her learning strategy or not. The ultimate goal is to further assist users in their planning and goalsetting tasks of the SRL process.
Provision of Usage Information. Being derived from the organizational memory, this information both supports the provided recommendations and allows users to further probe into the organizational memory around a recommended resource. For instance, each available competence is accompanied by information such as the number of users who have acquired it, their roles in the organization and the comments and discussions about it. The recommendation of a learning path is further augmented with the number of users (or orga-
Figure 1. A screenshot of the Learning Pal tool
Social Wave informs users about the latest updates of their learning goals and the accompanying learning resources. Such updates allow for tracking how a goal and any of the competences, activities and knowledge objects included in it have been used, ranked or commented on by other members of the organization, thus increasing the social embeddedness of the learning process.
C. Aligning to Individul Learning Goals Progress-o-meter aids users to monitor their own learning progress in the organizational context. It shows users their progress in achieving their defined learning goals, in terms of the completeness of each competence included within a goal and the completeness of learning activities performed toward achieving each of these competences; thus, it supports the self-reflection phase of SRL. It also provides users with a comparison of their progress with their colleagues who have the same learning goal (e.g., a goal shared by the members of a project). Observing oneself within the social context of the organization helps learners to monitor their progress toward their goals, thus assisting them in the performance phase of SRL. Activity Streams provide users with an overview of their learning activities such as annotating and sharing knowledge resources, adding new competences to a goal, commenting on a resource, requesting help, etc. It helps user to keep track of all their learning and knowledge building activities. This stream of activities is not limited to our Learning Pal tool, but thanks to the use of ontologies and linked data principles, those activities are also collected from other different systems learners regularly use.
D. Usage Scenario To better illustrate the functionalities of Learning Pal from the perspective of the challenges discussed in Sec. II and the role of the underlying ontologies, let us walk through a typical scenario for workplace learning involving a newcomer in a large organization. Brian is a newcomer in a company who plans to start his learning and knowledge building activities at his new workplace. However, like many other newcomers, he is concerned about gaining/enhancing the competences required for the duties assigned to his new organizational position. Our previous research [8] shows that lack of familiarity with organizational needs, policies and expectations, is one of the major challenges that newcomers face in larger companies. Likewise, Brian is not sure wherefrom he can obtain the information about the competences related to his new duties.
68
The Contextual Recommendation of Available Competences feature of Learning Pal (cf. Sec. III A) is under the category Duties (Fig. 1 A). Brian can spot the competences which are valued by his company, and required for accomplishing his duties. Thanks to the linked data support provided by the ontologies (cf. Sec. IV), Brian can benefit from the personalized visual hints which indicate those competences of higher importance for him, considering his current state of expertise w.r.t. the duties he is responsible for. Having identified the competences he is missing, Brian creates a new learning goal in his Learning Pal (Fig. 1 B). Now, he needs to obtain information about the best ways to achieve these competences. For each recommended competence in his Learning Pal, Brian can glance over the Recommended Learning Paths, Learning Activities and Knowledge Assets for that competence (Fig. 1 C), and also explore their Usage Information (cf. Sec. III B), such as visual representations showing the number of users, along with their organizational positions, who have been successful in achieving a certain competence, the average time that took other people to complete a recommended learning path for this competence, or indicators representing how ‘lively’ this competence has been recently (e.g., the number of comments, rankings, tags, and submitted help requests for it), Fig. 1 D. All in all, this information enables Brian to better make a decision whether to include a certain competence in his learning goal and which resources to allocate for it. Again, the integrated set of ontologies (cf. Sec. IV) is the main enabler for the induction of all this diverse information based on the activities of various users in different working environments. Once Brian has chosen competences for his new learning goal, he can simply follow the selected learning path toward achieving each competence. At this level, Progress-o-meter (cf. Sec III.C) enables him to monitor his learning process (Fig. 1 E). Further, the updates provided by Social Wave enable Brian to better adapt his learning strategies w.r.t the social context of the organization. Observing his Activity Stream, is another facility provided by Learning Pal that Brian can benefit from at this phase of his learning process to reflect over his learning goals and/or document and share his learning experiences. Having all these activities tracked and gathered in one place is supported by the underlying ontologies, as explained in the following section.
These annotations allow for interlinking KOs based on their semantics. In addition, we are currently experimenting with ontology elements for expressing different kinds of relationships between KOs. For example, the alocom-ct:reflection property allows for relating a user’s (externalized) reflection and the KO that the reflection was based upon. This kind of relationship enables us to pull together KOs that discuss/comment/critique another KO (e.g., a blog post) authored by a certain user and to provide the user with feedback regarding the usage of his/her knowledge (cf. Knowledge Sharing Profiles, Sec. III), thus contributing to his/her motivation for further knowledge sharing. Reflections are also relevant from the pedagogical perspective as they are an important part of the SRL process (cf. Sec. II). In order to gain access to and make use of the data about users’ activities within different software tools (required for Learning Pal’s functionalities), we needed to define a mapping between the tools’ specific data storage format (i.e., their native database schema) and the appropriate elements (classes and properties) of our ontologies. So far, these mappings have been defined for Moodle and MediaWiki, and we are now working on the mappings for the Elgg social networking platform.
IV. THE UNDERLYING ONTOLOGIES Fig. 2 presents a simplified example of the role played by the ontologies1 in integrating data about a learner’s activities and reflections originating from different online environments. In particular, the user (um:User) has authored (foaf:maker) a wiki page (alocom-cs:WikiPage), a blog post (alocomcs:BlogPost) and has posted a message (a:ForumPost) as a reflection on the content of the blog post. Although being created in different tools, these KOs and the (meta)data about their context of creation can be seamlessly integrated using our ontologies. Furthermore, the wiki page and the blog post are semantically annotated with the DBpedia (http://dbpedia.org) topic Pedagogy, whereas the blog post has been annotated with tags (ann:hasTag, ann:Tag) as well.
The ontologies also allow for the semantic-rich representation of both personal and organizational goals and expectations, and thus support Learning Pal’s functionalities that aim at motivating users through harmonization of learning goals on the individual and organizational level (cf. Sec. III A). For example, recommendation of competences to be acquired can be based on the requirements of the organizational position an individual would like to be promoted to. As indicated in Fig. 3, each organizational position (org:Organizational Position) has one or more assigned duties (org:Duty). A duty defines a set of expected behaviors, rights and obligations and imposes certain requirements (org:CompetenceRequirements) in terms of competences required for fulfilling that duty. An individual holding a certain position needs to satisfy the competence requirements of all the duties related to the desired position (as indicated in the
1
Figure 2. The use of ontologies for integrating the data about a user’s activities and KOs from different sources
All the ontiologies are available at http://bit.ly/f2Azq4
69
nized in a controlled test-setting using scenarios to guide users through selected tasks. Second, we will deploy Learning Pal in the application cases as a part of their regular practice.
Usage Scenario, Sec. III D). The missing competences are identified by the system by comparing the required competences with the individual’s competence profile (represented through one or more c:CompetenceRecord), and used for suggesting competences to be acquired.
ACKNOWLEDGMENT This publication was partially supported/co-funded by the European Community/European Union under the Information and Communication Technologies theme of the 7th Framework Program for R&D. This document does not represent the opinion of the European Community, and the European Community is not responsible for any use that might be made of its content. REFERENCES [1] [2]
Figure 3. The ontology elements used for computing recommendations that harmonize organizational and individual needs and expectations.
V.
[3]
RELATED WORK AND CONCLUSIONS
Different aspects supporting SRL and enhancing the motivation of individuals to take part in learning and knowledge building activities at workplace have been the subject of several research efforts recently. The APOSDLE project aimed at enhancing users’ productivity in informal selfdirected workplace learning by making individuals aware of available knowledge sources for a task at hand in the context of their everyday work processes [15]. Utilizing the Knowledge Maturing model [16], the MATURE project focused on individual, interpersonal and work-context motivational barriers, and suggested that for organizational agility the intrinsic motivation of employees should be incentivized by engaging them in collaborative learning activities with new forms of organizational guidance. The TenCompetence project built on the self-directedness of learners in a learning network, emphasizing the importance of social exchange between community members, in particular the ability to generate and sustain “connections” between users. It may be concluded that SRL in work environment is a highly socially-mediated process (rather than being individually-based), structured by and deeply integrated within work tasks and priorities as well as the performance measurement and promotion criteria [2]. As an advancement to abovementioned research initiatives, in this paper we have demonstrated how the Learning Pal tool, which leverages ontologies to support selfregulation in organizational learning, can provide autonomous learners at workplace both with the organizational and individual contexts and guide the intrinsically motivated learning. In the future work, we will improve Learning Pal’s support for documenting and sharing learning experiences, and collaboration. Once extended, it will be evaluated in three heterogeneous application cases by a mix of quantitative and qualitative evaluation methods in two iterations. First, a walkthrough of the running prototype will be orga-
[4] [5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14] [15] [16]
70
P., Tynjala, “Perspectives into learning at the workplace,” Educational Research Review, Vol.3, No.2, 2008, pp.130-154. A., Margaryan, et al., “Self regulated learning and knowledge sharing in the workplace,” In Proc. Int’l Conf. Organisational Learning, Knowledge and Capabilities, 2009. B. J., Zimmerman, “A Social Cognitive View of Self-Regulated Academic Learning.” J. of Educational Psychology, Vol.81, No.3, 1989. B. J., Zimmerman, "Self-Regulated Learning and Academic Achievment: An Overview," Educational Psychologist, Vol.25, No.1, 1990, pp. 3-17. M. V., Covington, “Goal Theory, Motivaiton, And School Achievement: An Integrative Review.” Annual Review Psychology, Vol.51, 2000, pp.171-200. I. Nonaka, & H. Takeuchi, “The knowledge-creating company: How Japanese companies create the dynamics of innovation.” Oxford University Press, 1995. K., Pata, & M., Laanpere, “Supporting cross-institutional knowledgebuilding with Web 2.0 enhanced digital portfolios,” In Proc. of the 8th IEEE Int’l Conf.Advanced Learning Tech. 2008, pp.798 - 800. M., Siadaty, J. Jovanović, D. Gašević, Z. Jeremić, T. Holocher-Ertl, “Leveraging Semantic Technologies for Harmonization of Individual and Organizational Learning,”, In Proc. 5th Euro. Conf. on Tech.enhanced Learning, 2010, pp. 340-356. D. Allemang, & J. Hendler. “Semantic Web for the Working Ontologist: Effective Modeling in RDFS and OWL.”, Morgan Kaufmann, 2008. Z. Jeremic, J. Jovanovic, D. Gasevic, “Semantic Web Technologies for the Integration of Learning Tools and Context-aware Educational Services,” In Proc. 8th Int’l Semantic Web Conf., 2009, pp. 860-875. IntelLEO D1.4 (2010). IntelLEO - Intelligent Learning Extended Organisation, Deliverable D1.4 IntelLEO Public Concept, FP7 Project no. FP7-ICT-231590. Available at: http://bit.ly/eXxKvj T. Jackson, J. Mackenzie, S. Hobfoll, “Communal aspects of selfregulation”. In M. Boekaerts, P.R. Pintrich & M. Zeidner (Eds.), Handbook of self-regulation (pp. 275-296). Academic Press, 2000. M., Boekaerts and E., Cascallar, “How far have we moved toward the integration of theory and practice in self-regulation?” Educational Psychology Review, Vol.18, No.3, 2006, pp.199-210. J. M. Carroll, “Five Reasons for Scenario-Based Design.” Interacting with Computers, Vol.13, No.1, 2000, pp.43-60. S. Lindstaedt, & H. Mayer, “A Storyboard of the APOSDLE Vision,” In Proc. 1st Euro. Conf. on Tech. Enhanced Learning, 2006. R. Maier, & A. Schmidt, “Characterizing Knowledge Maturing – Conceptual Process Model for Integrating E-Learning and Knowledge Management,” In Proc. 4th Conf. Professional Knowledge Management, 2007, pp. 325-334.
