See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228375485

E-learning agents

Article in The Learning Organization · June 2007

DOI: 10.1108/09696470710749245

CITATIONS READS
44 2,280

1 author:

Dawn G. Gregg
University of Colorado
58 PUBLICATIONS 1,910 CITATIONS

SEE PROFILE

All content following this page was uploaded by Dawn G. Gregg on 30 May 2014.

The user has requested enhancement of the downloaded file.

 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0969-6474.htm

TLO
14,4 E-learning agents
Dawn G. Gregg
University of Colorado and Health Sciences Centre, Denver, Colorado, USA

300 Abstract
Purpose – The purpose of this paper is to illustrate the advantages of using intelligent agents to
facilitate the location and customization of appropriate e-learning resources and to foster collaboration
in e-learning environments.
Design/methodology/approach – This paper proposes an e-learning environment that can be used
to provide customized learning. It utilizes a set of interacting agents that can personalize instruction
based on an individual’s prior knowledge as well as their cognitive and learning needs. The e-learning
agents monitor the e-learning environment and improve learning and collaboration based on learners’
prior knowledge, social characteristics and learning style.
Findings – E-learning agents should allow the discovery of new learning objects more easily, allow
learners to customize materials presented to improve learning outcomes, and improve collaboration in
the e-learning environment.
Originality/value – Little prior research has been done on the use of agents in e-learning
environments. This paper proposes a set of e-learning agents that, if implemented in online education
or training environments, should provide tangible benefits to organizations.
Keywords E-learning, Intelligent agents, Education, Training, Worldwide web
Paper type Conceptual paper

Introduction
E-learning represents one of the most dramatic changes to impact education and
training in the history of education. The internet and multimedia technologies are
reshaping the way knowledge is delivered allowing e-learning to emerge as a solution
to the problem of lifelong learning and workforce training (Zhang et al., 2004). What
makes e-learning content different from other educational materials is that it can be
disassembled as individual learning objects, tagged, and stored for reuse in a variety of
different learning contexts (Harris, 2005). These learning objects can be assembled into
different configurations depending on the requirements of an individual educational
situation. This reuse of educational content is one of the core values of e-learning
systems (Harris, 2005). While current learning technology can be used to effectively
manage the delivery of courses through the web, it does not support learning object
reuse (Sicilia and Lytras, 2005).
Research in the areas of intelligent tutoring systems, virtual mentors, and adaptive
hypermedia has produced techniques and tools that can provide improved learning
outcomes (Brusilovsky, 2000; Melis et al., 2006; Zhang, 2004). However, current
systems are either domain specific or non-adaptive and do not support learning object
reuse. This has resulted in the search for open intelligent e-learning infrastructures that
can be used with standard web technology (Sicilia and Lytras, 2005). Software agents
have great potential for addressing the limitations of current e-learning systems by
The Learning Organization supporting learning processes that target and deliver just-in-time learning materials
Vol. 14 No. 4, 2007
pp. 300-312 required by the individual learners. Software agents can be used to support instructors
q Emerald Group Publishing Limited and domain experts with both course design and delivery. They can also support
0969-6474
DOI 10.1108/09696470710749245 individual learners by personalizing course materials based on learning objectives,
learner characteristics, and prior learner knowledge, and facilitating learner E-learning
interaction. agents
Organizational learning environments can include instructor-led synchronous
courses, courses developed for asynchronous use, and self-directed learning. The
e-learning agent architecture proposed in this paper supports all three of these learning
scenarios. The e-learning agent system utilizes “semantic web” metadata descriptions
to locate and classify web learning resources (Berners-Lee et al., 2001), planning and 301
scheduling for intelligent course tailoring, personalization techniques to develop an
individualized curriculum for each learner, and active monitoring to facilitate
collaboration among learners and between instructors and learners.
This paper is organized as follows: the next section reviews related work with
learning support systems and autonomous agents. The sections that follow describe
the various agents that could be a part of an e-learning agent system and discuss their
functionality. The final section summarizes the results, and discusses potential
drawbacks of the proposed system and future research directions for agents operating
in e-learning domains.

Learning support systems and agents

Learning support systems are designed to support learners and provide improved
learning outcomes (Brusilovsky, 2000). Some learning support systems have been
proposed for use in traditional e-learning domains. For example, the ActiveMath
system is a complex web-based adaptive learning environment with a number of
components and interactive learning tools (Melis et al., 2006). Another system designed
for higher education courses allows students to conduct searches of external materials
(the ACM digital library) from within the learning environment (Gašević and Hatala,
2006). Finally, the Virtual Mentor system is a complete multi-media-based e-learning
environment that enables well-structured, synchronized, and interactive virtual
instruction. Testing of a prototype Virtual Mentor system indicates that learners that
used the Virtual Mentor system had better learning outcomes than learners in
traditional classroom settings (Zhang et al., 2004). This research suggests that carefully
crafted e-learning systems can benefit learners and provide improved learning
outcomes.
Learning support systems have also been implemented in organizational settings.
One system that has been successfully implemented is the inter-organizational
learning network implemented in the North Carolina wet-processing industry. The
system links multiple stakeholders involved in North Carolina’s Toxicity Program and
successfully allowed them to collaborate and learn from shared publications, trade
associations, and consultants (Manring and Moore, 2006). Within three months of
beginning the project, significant improvements occurred from suggested process
changes and chemical substitutions. Other e-learning systems have been successfully
implemented in organizations to perform ongoing sales force training related to new
product releases (Chelan, 2006) and to perform skill-based or job-based training
(Whitney, 2004). These e-learning systems have demonstrated measurable business
results to the organizations in which they were implemented.
One of the problems with existing e-learning systems is that most support only one
or two e-learning functions (e.g. material delivery and collaboration). Current systems
do not support the development of completely integrated online learning environments
TLO that meet the needs of individual students and groups. Most existing systems require
14,4 an enormous amount of time on the part of instructional designers to develop
instructional videos, presentation slides, lecture notes and other multimedia materials
which can then be compiled into an integrated knowledge repository (Zhang et al.,
2004). The time commitment required to develop a fully integrated course is not
feasible for many e-learning situations because small class sizes make the time
302 commitment impractical, or because the material being taught changes frequently.
This paper proposes that e-learning systems utilize intelligent agents to support
instructional design, the retrieval of relevant learning materials, and the processing
and analysis of data to enable meaningful e-learning recommendations to be made to
instructors and learners. Intelligent agents are programs designed to assist end-users
in different ways. Agents can hide the complexity of difficult tasks, perform tasks on
the user’s behalf, train or teach the user, help users collaborate, or monitor events and
procedures (Maes, 1994). Agents operating in heterogeneous networked computing
environments are being used as: intelligent search engines (Hsinchun et al., 1998;
Lawrence and Giles, 1998; Menczer, 2003; Selberg and Etzioni, 1997; Yang et al., 2000);
to track user activities and data to improve end-user decision making (Bui and Lee,
1999, Gandon et al., 2002, Hess et al., 2000); and to find and classify specific types of
information on the web (Bui and Lee, 1999; Etzioni, 1997; Gregg and Walczak, 2007).
These agents do not attempt to map or understand the entire web. Instead, they
attempt to process specific types of content about which the agent has some prior
knowledge.
The opportunities for using agents in e-learning applications are enormous. Agent
characteristics like autonomy, abilities to perceive, reason and act in specialized
domains, as well as their capability to cooperate with other agents makes them ideal
for e-learning applications (Papazoglou, 2001). However, many current agents do not
provide a comprehensive service to end-users. For example, many previously proposed
information retrieval agents are capable of retrieving specific information but do not
have the ability to filter, analyze, or make recommendations based on that information
(Shaw et al., 2002). Agents are not being used in existing e-learning applications (Woolf
and Eliot, 2005). Some researchers have proposed agent-based e-learning systems (e.g.
Huang et al., 2006a, b); however, the agent-based systems currently being proposed do
not support all aspects of e-learning.
Unlike previous e-learning support systems, the e-learning agents presented in the
next sections support instruction design, learning object reuse, personalization and
collaboration. They should allow learning objects to be assembled more quickly, while
supporting the mass customization of learning materials to meet individual needs.

E-learning agents
Ongoing research into effective learning support systems suggests that in order to
support ubiquitous, collaborative, experiential, and contextualized learning in dynamic
virtual communities, an e-learning environment should provide the following features
for learners: (Allison et al., 2005):
.
Experiential active learning. Learning resources should be interactive, engaging,
and responsive, with active learning and knowledge formation emphasized
above simple information transfer.
 .
Personalized. The learning environment should be customized to the individual E-learning
learners learning styles and educational needs with the quality of the learning agents
experience continually validated and evaluated. This includes customizing
accessibility to meet unique learner needs (e.g. to support screen readers,
language translation or alternative devices automatically), and dynamically
creating appropriate learning contexts.
.
Collaboration socio-constructivist. Both solitary and group work should be 303
supported.

In addition to the requirements suggested above, e-learning environments can also be

used to support learning design, including:
.
Lesson planning. Agents can be used to perform information gathering and
sophisticated reasoning necessary to determine appropriate learning sequences
(Woolf and Eliot, 2005; Cassin et al., 2003, Sicilia, 2006).
.
Resource location. With hundreds of thousands of educational resources
available on the web, anyone assembling e-learning materials requires assistance
to locate appropriate resources online (Gašević and Hatala, 2006; Woolf and Eliot,
2005). Agents can facilitate location of a wide variety of learning materials,
including those that support active learning.

The e-learning environment, shown in Figure 1, includes the agents discussed above. It
includes three agents designed to support the learning design process:
(1) an Instruction Agent;
(2) a Lesson Planning Agent; and
(3) a Resource Location Agent.

It also includes a Learner Centered Agent, and a Personalization Agent, to customize

the individual learning process. Finally, it includes Collaboration Agents, designed to
facilitate and encourage interaction among learners and between instructors and
learners. The following sections describe the functionality of these e-learning agents.

Instruction Design Agents

Instruction Design Agents allow the instruction designer to tailor the learning material
selection, assist them in identifying appropriate search terms to locate learning
materials based on specific concepts, and learning styles, and communicates with the
collaboration agents to improve interaction between the instructor and learners in
environments with instructor-led learning.

Lesson Planning Agents

One of the most important tasks required when planning a course or training module is
the development of the learning design. The learning design includes the assumptions
and guidelines used to formulate learning objectives, as well as the selection and
scheduling of specific learning activities (Sicilia, 2006). Lesson planning agents can be
used to assist with the design of the course structure as well as with the selection of
appropriate learning materials. For example, pedagogical agents have been proposed
to determine “best-practice” course structures using online catalog descriptions and
TLO
14,4

304

Figure 1.
E-learning agent
architecture

syllabi of similar courses (Woolf and Eliot, 2005). These agents assume there is
regularity to the order in which subjects are taught (Cassin et al., 2003). The agents
select the most frequently occurring topics, then use pair wise frequency to determine
which topics appear before or after another most often, and finally propose an
appropriate course structure (Woolf and Eliot, 2005).
Once an appropriate course structure is developed, the next step is to dynamically
generate recommended instructional resources and schedules, which can be
customized for individual learners. This involves both planning and scheduling
processes. Planning the course content involves selecting the materials that support a
particular educational topic or sub topic. The Lesson Planning Agents retrieve both
local and web-based resources and determine if the resources are interactive, engaging,
and responsive (to support active learning) and if the resources support a variety of
learning styles. If not, the Resource Location Agent can be used to identify additional
resources for the specific topic. The materials selected during the planning process
then are ordered into a comprehensive linear schedule, based on constraints, such as
course timeline and usefulness of materials to prior learners. The Lesson Planning
Agent decomposes complex topics into simpler subtopics to facilitate the ordering of E-learning
the materials. agents
The lesson planning process is an iterative one, with the Lesson Planning Agents
providing candidate solutions to instruction designers and then allowing them to alter
parameters and produce alternate lesson plans. Instruction designers can change the
ordering of topics and materials, specify some materials as required, and
recommending others for learners that experience difficulties related to a specific topic. 305

Resource Location Agents

Resource discovery can occur in two modes in e-learning environments. The first
occurs when learners explore the digital environment on their own to assemble
learning materials based on a self-perceived learning need, and the second occurs when
domain experts or instructors assemble the reusable learning objects necessary to
support a course or learning module. The web provides access to a huge repository of
learning materials; however, anyone wishing to use learning materials available on the
web is faced with major challenges. The web is highly dynamic and volatile, with the
constant addition of new materials, as well as frequent updating of existing materials
and the disappearance or removal of materials (Baeza-Yates and Ribeiro-Neto, 1999).
Thus, anyone utilizing web-based learning materials is constantly facing the prospect
that important learning resources may be unexpectedly eliminated or the way that they
are used may change. In addition, most of the web’s content today is designed for
humans to read, not for computer programs to manipulate meaningfully. Thus, it is
difficult for any computer program to classify new instructional resources correctly as
they appear on the web. One answer to this problem is semantic web technologies that
can provide the means to describe resources and services, and compose them in virtual
learning environments (Berners-Lee et al., 2001).
The purpose of e-learning resource location agents is to locate educational resources
available on the web, then evaluate those resources to determine what topics they
address, and what learning styles they support. Semantic web technologies provide a
semantic infrastructure that enable these agents to accomplish this goal. Providers of
learning materials can now deploy resources and services using semantic web
metadata that describe the learning content in a uniform way. One of the most widely
recognized metadata standards is the IEEE Learning Object Metadata (LOM) (Duval,
2002). It defines relevant characteristics of the learning object including general, life
cycle, meta-metadata, educational, technical, educational, rights, relation, annotation,
and classification categories.
Resource location and maintenance agents can utilize semantic web metadata to
help locate and classify new learning materials as well as to monitor existing resources
to determine if they have been changed, moved, or eliminated. These agents are similar
to information retrieval agents that have been used in other domains (e.g. Abasolo and
Gomez, 2000; Cazalens et al., 2000; Gregg and Walczak, 2007; Rhodes and Maes, 2000;
Tan et al., 2002; Walczak, 2003). However, resource location agents need to allow
semantic searches of online content to facilitate discovery of appropriate materials
based on the conceptual meaning of the subject material (Huang et al., 2006a, b; Shafrir
and Etkind, 2006). The main processes that are involved in semantic information
retrieval include (Korfbage, 1997):
TLO .
a querying process, where the user specifies the types of information to be
14,4 located using natural language or terms connected by Boolean operators;
.
an indexing process, where a document representation is created based on
semantic web metadata and word usage; and
.
an evaluation process, were a matching between the user query and the
document representation is performed using concept parsing algorithms (a
306 generic semantic procedure that identifies the lexical labels and building blocks
of concepts) (Shafrir and Etkind, 2006) or ontology mappings (Gašević and
Hatala, 2006).

The development of systems utilizing reusable learning objects and the semantic web
is in its initial stages; however, there are a few implemented systems that show the
promise these technologies hold. For example, the ActiveMath system utilizes the
semantic XML for mathematical documents (OMDoc) to create self-describing learning
objects that can be selected and linked based on the specific learning contexts. A
second semantic web system, Fedora, is designed to allow the storage, management,
and dissemination of complex digital objects and the relationships among them
(Lagoze et al., 2006). This system is not specifically designed to support learning
objects; however, it is designed to support the integration and reuse of digital materials
in different contexts, suggesting that domain independent resource location systems
are possible using semantic web technologies.
Utilizing an agent based resource location process can help both instructors and
individual learners locate and utilize learning resources that facilitate active learning
and knowledge formation (e.g. video, Java applets and Flash materials that provide
animated demonstrations and/or interactivity). Materials that are deemed appropriate
can then be automatically added to lesson plans as needed to support specific learning
objectives.

Learner Centered Agents

Learner Centered Agents are responsible for making the learner’s interaction with the
e-learning environment smooth and effective. They buffer malicious or sub-par
performance by resources (e.g. a buggy application) and assemble various educational
resources together into as coherent a whole as possible. The Learner Centered agent is
responsible for soliciting feedback from learners regarding the effectiveness of specific
learning materials; and continuously monitors learning outcomes. It is responsible for
communicating with the Personalization Agent and the Collaboration Agent to
improve learning.

Personalization Agents
Research on the cognitive information processing model of learning suggests that
customizing learning materials based on the individual’s preferred learning style or on
personality can provide a measurable benefit to the learner (e.g. improved learning
outcomes) (Heinström, 2000; Leidner and Jarvenpaa, 1995; Bovy, 1981; Wilson, 2000).
Personalization agents can be used to create a personalized learning model and
pathway tailored to individual learner knowledge and personality traits (Huang et al.,
2006a, b).
 Learners begin courses or training sessions by taking knowledge surveys and/or E-learning
answering questions related to their cognitive, affective, and social characteristics. agents
This allows personalization agents to customize the available learning objects to meet
an individual learner’s needs. For example, if a learner lacks the prerequisite
knowledge for a particular learning objective, additional learning objects can be
recommended to correct for the deficit. Personalization agents should be able to select
learning materials and optimize schedules for individual learners based on cognitive 307
style, personal preferences, and accessibility needs in addition to prior knowledge and
desired knowledge. Personalization agents should accommodate learners with a wide
range of skills and backgrounds; allowing learners who are strong in one area to move
ahead quickly, while other learners obtain extra assistance.
Personalization agents continuously monitor individual learners progress and use
of materials. Learners rate the usefulness of learning materials to them and answer
questions that assess their understanding of the subject matter being taught. This
information is used to determine the usefulness of materials for teaching specific topics
as well as to update the individual lesson plan as the learner interacts with the system.

Collaboration Agents
Collaborative learning has a major role in constructive cognitive development (Piaget,
1928, 1932). Collaboration allows learning to occur in relatively realistic, cognitively
motivating, and socially enriched learning contexts. There are a number of
experimental studies and implemented systems that emphasize the importance of
collaboration to e-learning (e.g. Durfee et al., 1989; Blaye et al., 1990, 1990; Chan and
Baskin, 1988; Chan, 1991). Researchers studying e-learning environments have
highlighted the need for a variety of collaboration tools including: tools to support
socialization between learners and their instructors, community building tools to
support the process of building cohesion in a group, and discussion supporting tools
(Yli-Luoma and Naeve, 2006).
A collaborative learning system facilitates the refinement and integration of the
subject knowledge of learners with the help of the collaborative partners. However,
numerous studies show collaboration is more difficult in e-learning environments.
These studies cite such components as physical separation, reduced sense of
community, disconnectedness, isolation, distraction, and lack of personal attention as
contributors to lack of success in various virtual programs (Kerka, 1996; Besser and
Donahue, 1996; Twigg, 1997; Stonebraker and Hazeltine, 2004).
Collaboration agents can be used to encourage collaboration between e-learning
participants and improve the efficiency of that collaboration. This can include
suggesting collaboration where appropriate or taking steps to improve the
collaboration. Collaboration agents can monitor the e-learning environment and
suggest synchronous “chat” between learners working on similar problems at the same
time. They can also point learners to appropriate discussion threads throughout the
learning process. These collaboration agents can also identify learners that are having
difficulty with a particular topic and facilitate their interaction with the instructor,
before they become too lost or spend too much time misunderstanding a particular
topic.
One factor that has been demonstrated to influence the efficiency of collaborative
learning is the composition of the group (Salomon and Globerson, 1989). If a
TLO collaboration agent observes too much homogeneity among the group members, it can
14,4 modify some conditions in order to activate collaboration. The agent may allocate roles
to participants, which creates conflict or provides them with contradictory information,
facilitating collaboration. Collaboration agents facilitate e-learning by increasing the
effective interactions between e-learning participants and improving the timing of
those interactions.
308
Evaluation
Research has shown that organizational learning does play an important role in the
generation of organizational performance (Garcı́a-Morales et al., 2006) and that
properly managed e-learning programs do provide measurable benefits to
organizations, for example Tyco Fire and Security (Whitney, 2004), sales force
performance (Chelan, 2006), physicians (Jamieson, 2005), and the textile wet-processing
industry (Manring and Moore, 2006). However, most organizations need to be able to
measure outcomes so that they can determine the return on investment for their
training programs (Whitney, 2004). Thus, an important capability of any
organizational e-learning system is the ability to evaluate learning and to provide
information that can be used to improve course quality and measure value provided to
the company.
The e-learning agent systems proposed in this paper are capable of monitoring
individual learning within a course and the usefulness of specific learning objects.
Agents can be used to generate learning progress reports against predefined goals and
can document learning efficiency as well as learning effort (Huang et al., 2006b). This
should allow organizations to manage their e-learning programs to provide the
materials and courses that are of the greatest benefit to the organization. The ability to
measure the impact of the organization’s e-learning program is another important
benefit of implementing an agent-based e-learning system.

Conclusions
This article describes a set of interacting e-learning agents that have the capability of
assisting instructors with online course design, course scheduling, and learning
material location. E-learning agents can also be used to personalize instruction based
on learner’s prior knowledge (e.g. from knowledge surveys), learning style, and
accessibility needs. These agents have the capacity to select and customize resources,
problems, and hints. Finally, agents can be used to foster effective collaboration in the
e-learning environment.
E-learning agents can be used to provide support to educate instructors and support
learning object reuse by helping them locate existing e-learning content. In the
dynamic world of the internet, the learning materials available on a particular topic are
constantly changing, and hence there is a significant need to continuously monitor
existing materials and search for new ones so that the most appropriate
course/training materials can be selected. One of the benefits provided by an agent
based e-learning system is it can continuously retrieve the most up-to-date educational
materials available when creating customized lesson plans for learners. Another
advantage of an agent based e-learning system is that it can assist instructors in
monitoring learner progress and facilitate interactions between the instructor and
learners that are struggling with a particular topic. They can also be used to optimally E-learning
place learners in groups formed to solve specific problems. agents
Introducing agents into the e-learning environment will fundamentally change the
way online education is conducted and the outcomes for both learners and instruction
designers. As online learners and instructors increase their use of intelligent agents to
automate information gathering, lesson planning, learning material customization, and
collaboration the outcome for both learners and instructors will be improved. With 309
sufficient information, agents should be able to the select the most appropriate learning
materials for individual learners based on both topics covered and learner
characteristics, thus improving learning outcomes. Agents can also monitor learning
effectiveness so the benefits of the e-learning program can be assessed by the
organization.

References
Abasolo, J.M. and Gomez, M. (2000), “MELISA: an ontology-based agent for information retrieval
in medicine”, paper presented at the ECDL 2000 Workshop on the Semantic Web, available
at: http://scholar.google.com/scholar?hl ¼ en&lr ¼ &q ¼ cache:TSbjIzjfkXIJ:www.csee.
umbc.edu/, cost/courses/691A_sp03/papers/melisa.pdf þ %22information þ retrieval
þ agent%22 (accessed December 2, 2005).
Allison, C., Cerri, S.A., Ritrivato, P., Gaeta, A. and Gaeta, M. (2005), “Services, semantics and
standards: elements of a learning grid infrastructure”, Applied Artificial Intelligence,
Vol. 19, pp. 861-79.
Baeza-Yates, R. and Ribeiro-Neto, B. (1999), Modern Information Retrieval, Addison-Wesley,
Reading, MA.
Berners-Lee, T., Hendler, J. and Lassila, O. (2001), “The semantic web”, Scientific American,
Vol. 284 No. 5, pp. 34-43.
Besser, H. and Donahue, S. (1996), “Introduction and overview: perspectives on . . . distance
independent education”, Journal of American Society for Information Science, Vol. 47
No. 11, pp. 801-4.
Blaye, A., Light, P., Joiner, R. and Sheldon, S. (1990), “Collaboration as a facilitator of planning
and problem solving on a computer-based task”, CITE Report 90, Institute of Educational
Technology, Open University, Milton Keynes.
Blaye, A., Light, P.H., Joiner, R. and Sheldon, S. (1991), “Joint planning and problem solving on a
computer-based task”, British Journal of Developmental Psychology, Vol. 9, pp. 471-83.
Bovy, R.C. (1981), “Successful instructional methods: a cognitive information processing
approach”, ECTJ, Vol. 29 No. 4, pp. 101-19.
Brusilovsky, P. (2000), “Adaptive hypermedia: from intelligent tutoring systems to web-based
education (invited talk)”, in Gauthier, G., Frasson, C. and VanLehn, K. (Eds), Intelligent
Tutoring Systems, Lecture Notes in Computer Science, Vol. 1839, Springer Verlag, Berlin,
pp. 1-7.
Bui, T. and Lee, J. (1999), “An agent-based framework for building decision support systems”,
Decision Support Systems, Vol. 25 No. 3, pp. 225-37.
Cassin, P., Eliot, C., Lesser, V., Rawlins, K. and Woolf, B. (2003), “Ontology extraction for
educational knowledge bases”, in van Elst, L., Dignum, V. and Abecker, A. (Eds), Agent
Mediated Knowledge Management, Lecture Notes in Artificial Intelligence, Vol. 2926,
Springer Verlag, Berlin, pp. 297-309.
TLO Cazalens, S., Desmontils, E., Jacquin, C. and Lamarre, P. (2000), “A web site indexing process for
an internet information retrieval agent system”, First International Conference on Web
14,4 Information Systems Engineering (WISE’00), Vol. 1, p. 0254.
Chan, T. (1991), “Integration-kid: a learning companion system”, in Mylopoulos, J. and Reiter, R.
(Eds), Proceedings of the 12th International Conference on Artificial Intelligence, Sydney,
Vol. 2, pp. 1094-9.
310 Chan, T. and Baskin, A. (1988), “Studying with the prince: the computer as learning companion”,
Proceedings of the Intelligent Tutoring Systems Conference, Montreal.
Chelan, D. (2006), “Revving up elearning to drive sales”, EContent, Vol. 29 No. 2, pp. 28-32.
Durfee, E.H., Lesser, V.R. and Corkill, D.D. (1989), “Cooperative distributed problem solving”, in
Barr, A., Cohen, P.R. and Feigenbaum, E.A. (Eds), The Handbook of Artificial Intelligence,
Vol. 4, Addison-Wesley, Reading, MA, pp. 83-148.
Duval, E. (Ed.) (2002), 1484.12.1: IEEE Standard for Learning Object Metadata, available at:
http://ltsc.ieee.org/wg12/files/LOM_1484_12_1_v1_Final_Draft.pdf (accessed June 8,
2006).
Etzioni, O. (1997), “Moving up the information food chain: deploying softbots on the world wide
web”, AI Magazine, Vol. 18 No. 2, pp. 11-18.
Gandon, F., Poggi, A., Rimassa, G. and Turci, P. (2002), “Multi-agent corporate memory
management system”, Applied Artificial Intelligence, Vol. 16 Nos 9/10, pp. 699-720.
Garcı́a-Morales, V.J., Lopez-Martı́n, F.J. and Llamas-Sánchez, R. (2006), “Strategic factors and
barriers for promoting educational organizational learning”, Teaching & Teacher
Education, Vol. 22 No. 4, pp. 478-502.
Gašević, D. and Hatala, M. (2006), “Ontology mappings to improve learning resource search”,
British Journal of Educational Technology, Vol. 37 No. 3, pp. 375-89.
Gregg, D. and Walczak, S. (2007), “Exploiting the information web”, IEEE Transactions on
System, Man and Cybernetics Part C, Vol. 37 No. 1, pp. 109-25.
Harris, J. (2005), “E-learning strategy: repurposing content”, Chief Learning Officer, October,
pp. 24-8.
Heinström, J. (2000), “The impact of personality and approaches to learning on information
behaviour”, Information Research, Vol. 5 No. 3, available at: http://informationr.net/ir/5-3/
paper78.html (accessed June 8, 2006).
Hess, T.J., Rees, L.P. and Rakes, T.R. (2000), “Using autonomous software agents to create next
generation decision support systems”, Decision Sciences, Vol. 31, Winter, pp. 1-31.
Huang, W., Eze, E. and Webster, D. (2006a), “Towards integrating semantics of multi-media
resources and processes in e-learning”, Multimedia Systems, Vol. 11 No. 3, pp. 203-15.
Huang, W., Webster, D., Wood, D. and Ishaya, T. (2006b), “An intelligent semantic e-learning
framework using context-aware semantic web technologies”, British Journal of
Educational Technology, Vol. 37 No. 3, pp. 351-73.
Hsinchun, C., Yi-Ming, C., Ramsey, M. and Yang, C. (1998), “An intelligent personal spider for
dynamic Internet/Intranet searching”, Decision Support Systems, Vol. 23, pp. 41-58.
Jamieson, A. (2005), “Ubiquitous computing, the semantic web and the future of e-learning”,
Work Based Learning in Primary Care, Vol. 3 No. 1, pp. 50-7.
Kerka, S. (1996), “Distance learning, the internet, and the world wide web”, ERIC Digest, ERIC
Document Reproduction Service No. ED 395 214, The Ohio State University, Columbus,
OH.
Korfbage, R.R. (1997), Information Storage and Retrieval, Wiley, Mississauga.
Lagoze, C., Payette, S., Shin, E. and Wilper, C. (2006), “Fedora: an architecture for complex objects E-learning
and their relationships”, International Journal on Digital Libraries, Vol. 6 No. 2, pp. 124-38.
agents
Lawrence, S. and Giles, C. (1998), “Context and page analysis for improved web search”, IEEE
Internet Computing, Vol. 2 No. 4, pp. 36-48.
Leidner, D.E. and Jarvenpaa, S. (1995), “The use of information technology to enhance
management school education: a theoretical view”, MIS Quarterly, Vol. 19 No. 3, pp. 265-91.
Maes, P. (1994), “Agents that reduce work and information overload”, Communications of the 311
ACM, Vol. 37 No. 7, pp. 31-40.
Manring, S.L. and Moore, S.B. (2006), “Creating and managing a virtual inter-organizational
learning network for greener production: a conceptual model and case study”, Journal of
Cleaner Production, Vol. 14 Nos 9-11, pp. 891-9.
Melis, E., Goguadze, G., Homik, M., Libbrecht, P., Ullrich, C. and Winterstein, S. (2006),
“Semantic-aware components and services of ActiveMath”, British Journal of Educational
Technology, Vol. 37 No. 3, pp. 405-23.
Menczer, F. (2003), “Complementing search engines with online web analysis agents”, Decision
Support Systems, Vol. 35 No. 2, pp. 195-212.
Papazoglou, M.P. (2001), “Agent-oriented technology in support of e-business”, Communications
of the ACM, Vol. 44 No. 4, pp. 71-7.
Piaget, J. (1928), Judgement and Reasoning in the Child, Harcourt Brace, New York, NY.
Piaget, J. (1932), The Moral Judgement of the Child, Routledge and Kegan Paul, London.
Rhodes, B.J. and Maes, P. (2000), “Just-in-time information retrieval agents”, IBM Systems
Journal, pp. 685-704.
Salomon, G. and Globerson, T. (1989), “When teams do not function the way they ought to”,
International Journal of Educational Research, Vol. 13 No. 1, pp. 89-100.
Selberg, E. and Etzioni, O. (1997), “The metacrawler architecture for resource aggregation on the
web”, IEEE Expert, Vol. 12 No. 1, pp. 8-14.
Shafrir, U. and Etkind, M. (2006), “E-learning for depth in the semantic web”, British Journal of
Educational Technology, Vol. 37 No. 3, pp. 425-44.
Shaw, N.G., Mian, A. and Yadav, S.B. (2002), “A comprehensive agent-based architecture for
intelligent information retrieval in a distributed heterogeneous environment”, Decision
Support Systems, Vol. 32 No. 4, pp. 401-15.
Sicilia, M.A. (2006), “Semantic learning designs: recording assumptions and guidelines”, British
Journal of Educational Technology, Vol. 37 No. 3, pp. 331-50.
Sicilia, M.A. and Lytras, M.D. (2005), “The semantic learning organization”, The Learning
Organization, Vol. 12 No. 5, pp. 402-10.
Stonebraker, P.W. and Hazeltine, J.E. (2004), “Virtual learning effectiveness: an examination of
the process”, The Learning Organization, Vol. 11 No. 3, pp. 209-25.
Tan, B., Foo, S. and Hui, S.C. (2002), “Web information monitoring for competitive intelligence”,
Cybernetics & Systems, Vol. 33 No. 3, pp. 225-51.
Twigg, C.A. (1997), “Is technology a silver bullet?”, Educom Review, Vol. 31 No. 2, pp. 28-9.
Walczak, S. (2003), “A multiagent architecture for developing medical information retrieval
agents”, Journal of Medical Systems, Vol. 27 No. 5, pp. 479-98.
Whitney, K. (2004), “Jim LaRocco: blazing a path to learning success”, Chief Learning Officer,
available online at: www.clomedia.com/content/templates/clo_cloprofile.
asp?articleid ¼ 517&zoneid ¼ 4 (accessed June 8, 2006).
TLO Wilson, E.V. (2000), “Student characteristics and computer-media communication”, Computers &
Education, Vol. 34 No. 2, pp. 67-76.
14,4 Woolf, B. and Eliot, C. (2005), “Customizing the instructional grid”, Applied Artificial Intelligence,
Vol. 19, pp. 825-44.
Yang, C.C., Yen, Y. and Chen, H. (2000), “Intelligent internet searching agent based on hybrid
simulated annealing”, Decision Support Systems, Vol. 28 No. 3, pp. 269-77.
312 Yli-Luoma, P.V.J. and Naeve, A. (2006), “Towards a semantic e-learning theory by using a
modeling approach”, British Journal of Educational Technology, Vol. 37 No. 3, pp. 445-59.
Zhang, D., Zhao, J.L., Zhou, L. and Nunamaker, J.F. (2004), “Can e-learning replace classroom
learning?”, Communications of the ACM, Vol. 47 No. 5, pp. 75-9.

Corresponding author
Dawn G. Gregg can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints

View publication stats