Conference PaperPDF Available

Competence Analyser: A portable GUI tool for modelling domain and learner knowledge

Conference Paper

Competence Analyser: A portable GUI tool for modelling domain and learner knowledge

Abstract and Figures

Learner models form the basis of adaptive learning systems, representing what the system knows about a learner. Knowledge about a learner's expertise within a learning domain plays an essential role when recommending learning objects as well as when supporting a learner's reflection and awareness. In this paper we present the Competence Analyser tool to construct knowledge structures as a foundation for learner models in accordance with a psychologically sound framework, the Competence based Knowledge Space Theory (CbKST), which supports performance assessment in user-adaptive learning systems. The tool enables the learner to actively influence the learning profile creation by setting learning goals and defining prior knowledge in the context of learning domains. A first prototype has been developed and evaluated within the frame of the EC-funded TARGET project that provides a game-based learning environment with adaptive story recommendation. The paper outlines how to foster awareness and self-regulation while building learner and domain models for adaptive learning systems using a graph-based GUI tool.
Content may be subject to copyright.
Wong, L.-H. et al. (Eds.) (2013). Proceedings of the 21st International Conference on Computers in Education.
Indonesia: Asia-Pacific Society for Computers in Education
Competence Analyser: A portable GUI tool for
modelling domain and learner knowledge
Simone Kopeinika*, Michael Bedeka, Georg Öttla & Dietrich Albertab
a Knowledge Technologies Institute, Graz University of Technology, Austria
b Department of Psychology, University of Graz, Austria
*simone.kopeinik@tugraz.at
Abstract: Learner models form the basis of adaptive learning systems, representing what the
system knows about a learner. Knowledge about a learner’s expertise within a learning
domain plays an essential role when recommending learning objects as well as when
supporting a learner’s reflection and awareness. In this paper we present the Competence
Analyser tool to construct knowledge structures as a foundation for learner models in
accordance with a psychologically sound framework, the Competence based Knowledge
Space Theory (CbKST), which supports performance assessment in user-adaptive learning
systems. The tool enables the learner to actively influence the learning profile creation by
setting learning goals and defining prior knowledge in the context of learning domains. A first
prototype has been developed and evaluated within the frame of the EC-funded TARGET
project that provides a game-based learning environment with adaptive story recommendation.
The paper outlines how to foster awareness and self-regulation while building learner and
domain models for adaptive learning systems using a graph-based GUI tool.
Keywords: Learner Modelling, Knowledge Modelling, Graph-based Visualization,
Prerequisite-relations, Competence based Knowledge Space Theory, Critical Reflection
1. Introduction
In user-adaptive learning systems information about the learner is used to identify and recommend
learning objects, or more concretely to select and adapt instructional strategies to the learner (Kay &
Kummerfield, 2012). A learner's specific needs may depend on various factors, including prior
knowledge (e.g. expertise levels), motivation, learning styles, learning goals and learning efficiency
(Klašnja-Milićević, Vesin, Ivanović & Budimac, 2011). Some of these factors, such as knowledge or
motivation, typically vary within a single learning cycle. Hence, to support adaptivity during a
learning cycle the learner model needs to be continuously updated. To this end, information about the
learner may be collected through a combination of implicit performance assessment and explicitly
requested user input (Brusilovsky & Millán, 2007), and then strategically integrated into the learner
model. The psychologically sound framework behind our approach is the Competence-based
Knowledge Space Theory (CbKST) (Heller, Steiner, Hockemeyer & Albert, 2006). The CbKST is a
framework for representing the conceptual organisation of a given body of competences. It extends
the behaviourist Knowledge Space Theory (Doignon & Falmagne, 2011), which focuses on the set of
problems or tasks a person is able to master. Beyond that, the CbKST concentrates on the
identification of latent skills and competences and prerequisite relations between them: If a person
wants to develop competence x, he or she has to develop y beforehand. These competences and their
prerequisite relations represent a competence structure which forms the basis of learners’ domain
models and, furthermore, for updates on these models.
Technology-enhanced learning (TEL) applications building upon the CbKST require the
definition of the competence structure and the competences a learner exhibits and aims to develop at
the very beginning of the learning process. Therefore, we introduce the Competence Analyser (CA), a
tool that allows defining initial competence information needed to provide CbKST based assessment
Wong, L.-H. et al. (Eds.) (2013). Proceedings of the 21st International Conference on Computers in Education.
Indonesia: Asia-Pacific Society for Computers in Education
and recommendation strategies. The first part of the CA constitutes the Competence Structure Editor
(CSE). The CSE embodies the interface to model the competence structure of a learning domain,
consisting of competences and their interrelations (i.e. prerequisite relations). It also supports the
learner in gaining insights into the learning domain in terms of critical reflection on learning goals and
competences and on how these competences of the learning domain relate to each other.
The second part of the GUI is to define a learner’s prior knowledge, in form of a “User Competence
Profile” (UCP) and learning goals by means of a “Target Competence Profile” (TCP). The difference
between these profiles results in the “Gap Competence Profile” (GCP) that encompasses the set of
competences a learner aims to develop by engaging with the TEL application.
Relevant and related work to the CA can be found in European research projects in the field
of TEL. APOSDLE (http://www.aposdle.tugraz.at), a work-integrated learning system, implemented
two competence related authoring tools. First, Moki (Ghidini et al., 2009) allows modelling tasks and
concepts (or skills) in a wiki and to export this model as an OWL-Ontology. The second competence
related authoring tool developed within APOSDLE, the Task Competence Tool (TACT), allows to
directly model tasks, topics and relations between tasks, topics and competences needed for a certain
task (Lindstaedt & Christl, 2011). Another description about competence gap analysis was given by
Coi et al. (2007) who presented a gap analysis tool on a conceptual level. Nkambou, Gauthier &
Frasson (1996) amongst others, describe similar authoring system, but do not work on the basis of
competences and competence relations.
In this paper we will introduce the Competence Analyser tool as a GUI component to define
CbKST supporting user and domain model construction. We will present the graphical user interface
including information about its implementation and possible application. Subsequently, we will
provide insights into the main results of the usability study and finish with a brief discussion on the
topic.
2. Functionality and implementation
A first prototype of the CA was implemented for and integrated in a Digital Educational Game in the
course of a TEL platform called TARGET (www.reachyourtarget.org ). The main component of
TARGET is a 3D virtual learning environment which consists of story dependent indoor and outdoor
locations where the learner communicates with non-playable characters to gather pieces of
information in order to master game scenarios. Stories represent learning objects that are
recommended to a user on the basis of a learner’s current competence state modelled and updated
through the CbKST service that works on the basis of the models designed in the CA (Kopeinik,
Nussbaumer, Bedek & Albert, 2012).
In TARGET each learning cycle was described through a learning plan (i.e. an ordered set of
game scenarios the learner has to engage with). We developed CA to assist in the definition of
learning plans with learning goals, prior knowledge and a learning domain. Within the two main
frames of the GUI learner and instructor can handle tasks such as: (1) the definition of the learning
domain in terms of competences and relations between competences (Competence Structure Editor)
and (2) the selection of competences to learn.
2.1 Competence Structure Editor
The Competence Structure Editor (CSE) was implemented as a prototype to support instructors and
organizations in creating competence structures. In this context, a competence structure is defined as a
set of competences ordered by means of prerequisite relations. A prerequisite relation between two
competences x and y means: “If a person wants to develop competence x, he or she has to develop y
beforehand” which is conceptually the same as “If a person has shown that he or she has competence
x, he or she has competence y too”. In this case, competence y is a prerequisite competence for
competence x.
The CSE provides not only a tool to relate competences to each other but also to visualize a
previously constructed competence structure as a modified Hasse diagram. The main difference to a
standard Hasse diagram is the visualization of the connectedness of a vertex by increasing the
Wong, L.-H. et al. (Eds.) (2013). Proceedings of the 21st International Conference on Computers in Education.
Indonesia: Asia-Pacific Society for Computers in Education
diameter of the vertex (see Figure 1). This visualization should make the information contained in the
competence structure more meaningful and insights more intuitive. The Hasse diagram is made
directly editable by mouse events. A mouse editing event is only accepted if it does not contradict the
Hasse diagram. A contradiction can e.g. occur if a cycle is introduced or by trying to add edges which
are already indirectly inferred by the Hasse diagram. For the cycle detection we used an algorithm
which is based on a dynamic topological sort algorithm as described by Pearce and Kelly (2006). For
calculating the transitive reduction of the graph as needed by the Hasse diagram visualization, we
implemented a simple algorithm which basically calculates the transitive closure of the graph (Aho,
Garey & Ullman, 1972). Consequently, for every valid editing action the transitive reduction and the
topological order of the underlying graph data structure is updated. The visualization itself was based
on the Jung Framework (Madadhain et al., 2005).
Figure 1. The Competence Structure Editor GUI.
The CSE (see Figure 1) offers three different views: (CSE1) The Prerequisite Editor which allows to
create and edit competences and to model their interrelations. (CSE2) The Closure View displays
the competence structure as a more detailed representation that also includes transitive connections.
(CSE3) The Tabular View displaying the competences and their relations in form of a table. (CSE4)
Filter and Edit Panels are part of all three views and can be used to search and edit competences.
2.2 Competence Analyser
The CA view fosters a learner’s active part in the personalization process. It is provided as an input
device to create an initial state of the learner profile and is used either by the learners and their
supervisors in collaboration, or by the learners themselves. Using the CA the learner is asked to drag
and drop competences from the list provided on the left hand side to either the UCP pane (information
about existing expertise) or the TCP pane (learning goals). When the learner is finished, the UCP pane
should include all competences the learner is competent in the use of at this point in time. The TCP
should include all competences the learner would like to show after the execution of the resulting
learning cycle. The Gap Competence Profile (GCP) then shows the competences a learner has to learn
in order to achieve the defined goal state.
Figure 2 presents the visual framework of the CA: (CA3) The Competence Panel implements a
tree like representation of the competence structure that is based on the inferences available from the
Competence Structure Editor and the semantics of a Prerequisite Relation. The competence panel
provides drag and drop functionality from the panel to a visualization. (CA4) The Profiles Pane shows
predefined job profiles consisting of a set of competences. (CA5) The User Competences Profile Pane
Wong, L.-H. et al. (Eds.) (2013). Proceedings of the 21st International Conference on Computers in Education.
Indonesia: Asia-Pacific Society for Computers in Education
allows adding and editing of competences that are identified as prior knowledge of a learner. (CA6)
The Target Competence Profile Pane allows adding and editing of competences that are defined as the
learning goal. (CA7) A panel that visualizes the difference between TCP and UCP. (CA1, CA2, CA8
and CA9) Menu Items to save, load and update data in the Panes.
Some options offered by the interface are: (a) Drag and drop of competences from the list view
to the profile tabs. (b) Selection and multiple selection of competences to move/delete within the
profile tabs; (c) The right mouse click on a competence-vertex displays a context menu; (d) Scrolling
is supported using the mouse wheel.
Figure 2. The Competence Analyser GUI.
3. Usability Studies
During the development process of the Competence Analyser we carried out several iterative
evaluation cycles for both, formative and summative evaluations, respectively. For the formative
evaluations we applied a thinking-aloud method with a small set of experts (n = 4 6) in the field of
cognitive psychology and TEL. Participants were asked to verbalize their thoughts as they interacted
with the tool. For each major release of the tool, we applied a more standardized procedure for
carrying out summative evaluations. The rationale behind a more standardized procedure was to
compare the evaluation results over time (at least on a descriptive level). For each of the studies, a
minimum of eight participants who were unfamiliar with the tool took part. The participants have
been provided with the background and purpose of the evaluation study as well as with a brief
introduction to the tool´s interface and its main features. Afterwards they had to structure a learning
domain (in the field of elementary algebra) and define a current and target competence profile for an
imaginary 2nd grade pupil. Finally, they were asked to fill out a short version of IsoMetrics inventory
for the testing the usability of the GUI (Willumeit, Gediga & Hamborg, 1996). IsoMetrics is based on
the ISO 9241/10 standard. The short version of the IsoMetrics inventory consists of the following 7
factors:
Suitability (15 items), e.g.
“In a given screen, I find all of the information I need in that situation.”
Self Descriptiveness (12 items), e.g.
“The terms and concepts used in the software are clear and unambiguous.”
Controllability (11 items), e.g.
“In order to perform my tasks, the software requires me to perform a fixed sequence of steps.”
(formulated conversely)
Wong, L.-H. et al. (Eds.) (2013). Proceedings of the 21st International Conference on Computers in Education.
Indonesia: Asia-Pacific Society for Computers in Education
Conformity with User Expectations (8 items), e.g.
“When executing functions, I have the feeling that the results are predictable.”
Error Tolerance (15 items), e.g.
I have never made an entry that caused a software error (e.g. a system/program crash or an
undefined dialog state).”
Suitability for Individualization (6 items), e.g.
The software lets me adapt forms, screens and menus to suit my individual preferences.”
Suitability for Learning (8 items), e.g.
“In order to use the software properly, I must remember a great many details.”
(formulated conversely)
The rating scale for the IsoMetrics inventory has a range from 1 (predominantly disagree) to 5
(predominantly agree) as well as the opportunity to rate 0 (no opinion). The following figure shows
the results (means and standard deviations for all sub-scales) for the final summative evaluation study
with a sample size of 9 participants.
Figure 3. IsoMetrics results of the summative evaluation study
As shown in the figure above, the results of the formative evaluation indicate, that the Competence
Analyser Tool has potential for improvement. However, the quite high value for the dimension
suitability for learning is very satisfying. In any case, we consider the results in general and the
answers / responses to the open questions of the short version of the IsoMetrics inventory, as
important insights that will be taken into account during further iteration cycles in the development of
the Competence Analyser Tool.
4. Discussion
The Competence Analyser was developed, integrated and repeatedly evaluated in the course of the
European research project TARGET (7th framework program), where resulting learning profiles
served as a basis for adaptive learning object recommendation (Kopeinik, Nussbaumer, Bedek &
Albert, 2012). Additionally, it aims to facilitate the learner´s reflection upon his or her own
performance, strengths and weaknesses: The CA-GUI is the application of an open learner model
(Bull, 2004), where the learner´s interactions with the GUI and his or her reflection should lead to an
enhanced learning process (Mitrovic & Martin, 2002). We will extend the reflection support of the
Wong, L.-H. et al. (Eds.) (2013). Proceedings of the 21st International Conference on Computers in Education.
Indonesia: Asia-Pacific Society for Computers in Education
tool by editing the appearance of nodes according to the learning performance of completed learning
tasks. Although the tool worked well from a technical perspective, usability studies have shown that
some end-users found it difficult to handle the bubble visualization in the user profile tab and would
have preferred a simpler table like view. Unfortunately, the summative evaluation studies were
restricted to one time use (e.g. different participants were in the samples) and therefore do not
consider potential improvements after familiarization. Problems have also been identified with the
visualization of the knowledge domain, since readability tends to decrease with the growth of the
graph. The main issue is the increasing number of edge crossings which hinder visual clarity. Future
work will therefore inspect algorithms to optimize the positioning of nodes in order to minimize edge
crossings.
Acknowledgements
This paper is part of the EC-Project TARGET funded by the 7th Framework Programme of the
European Commission’s IST-Programme (contract no. 231717). The authors are solely responsible
for the content of this paper. It does not represent the opinion of the European Community, and the
European Community is not responsible for any use that might be made of data appearing therein.
References
Aho, A.V., Garey, M.R., & Ullman, J.D. (1972). The Transitive Reduction of a Directed Graph,
SIAM Journal on Computing, 1, 131-137.
Brusilovsky, P., & Millán, E. (2007). User models for adaptive hypermedia and adaptive educational
systems. In P. Brusilovsky, A. Kobsa and W. Nejdl (Eds), The Adaptive Web (pp. 3-53). Berlin,
Heidelberg: Springer-Verlag..
Bull, S. (2004). Supporting learning with open learner models. In: Proceedings of the 4th Hellenic
Conference on Information and Communication Technologies in Education, Athens, Greece,
September 29 - October 3, pp. 47-61.
Coi, J.L., Herder, E., Koesling, A., Lofi, C., Olmedilla, D., Papapetrou, O., & Siberski, W. (2007) A
Model For Competence Gap Analysis, In Proceedings of 3rd International Conference on Web
Information Systems and Technologies (pp. 4-6), INSTICC Press.
Doignon, J.-P., & Falmagne, J.-C. (2011). Learning Spaces. Berlin, Heidelberg: Springer-Verlag.
Ghidini, C., Kump, B., Lindstaedt, S., Mahbub, N., Pammer, V., Rospocher, M. & Serafini, L.,
"MoKi: The Enterprise Modelling Wiki, In Demo Proceedings of the 6th European Semantic Web
Conference (ESWC 2009), 831-835.
Lindstaedt, S. & Christl, C (2011).APOSDLE learn@work: Firsthand Experiences and Lessons Learned.
Work-Integrated Learning in Engineering, Built Environment and Technology: Diversity of Practice in
Practice. (2011), S. 52 - 71
Nkambou, R., Gauthier, G., & Frasson, C. (1996) CREAM-Tools : An Authoring Environment for
Curriculum and Course Building in an intelligent Tutoring System, 3rd International Conference
on Computer Aided Learning and Instruction in Science Engineering , Lecture Notes in
Computer Science, no. 1108, pp. 186-194, Springer Verlag, Donostia, San Sebastian, Spain.
Heller, J., Steiner, C., Hockemeyer, C., & Albert, D. (2006). Competence-Based Knowledge
Structures for Personalised Learning. International Journal on E-Learning, 5(1), 75-88.
Kay, J., & Kummerfield, B. (2012). Lifelong learner modelling. In P.J. Durlach and A.M. Lesgold
(Eds), Adaptive Technologies for Training and Education (pp. 140-165). New York: Cambridge
University Press.
Klašnja-Milićević, A., Vesin, B., Ivanović, M., & Budimac, Z. (2011). E-Learning personalization
based on hybrid recommendation strategy and learning style identification, Computers &
Education, 56 (3), 885-899.
Kopeinik, S., Nussbaumer, A., Bedek, M., & Albert, D. (2012). Using CbKST for Learning Path
Recommendation in Game-based Learning. In Proceedings of the 20th International Conference
on Computers in Education ICCE 2012( pp. 86 91).
Wong, L.-H. et al. (Eds.) (2013). Proceedings of the 21st International Conference on Computers in Education.
Indonesia: Asia-Pacific Society for Computers in Education
Madadhain, J., Fisher, D., Smyth, P., White, S., & Boey, Y.B. (2005). Analysis and visualization of
network data using JUNG, Journal of Statistical Software, 10, 1-35.
Mitrovic, A. & Martin, B. (2002). Evaluating the Effects of Open Student Models on Learning, In. P.
De Bra, P. Brusilovsky and R. Conejo (Eds.). Adaptive Hypermedia and Adaptive Web-Based
Systems (pp. 296-305), Berlin: Springer-Verlag.
Pearce, D. J., & Kelly, P.H. (2006). A Dynamic Topological Sort Algorithm for Directed Acyclic
Graphs, Journal on Experimental Algorithmics, 11, 1-24.
Willumeit, H., Gediga, G., & Hamborg, K. (1996). IsoMetricsL: Ein Verfahren zur formativen
Evaluation von Software nach ISO 9241/10, Ergonomie & Informatik, 27, 5-12
Article
Full-text available
In recent years, various recommendation algorithms have been proposed to support learners in technology-enhanced learning environments. Such algorithms have proven to be quite effective in big-data learning settings (massive open online courses), yet successful applications in other informal and formal learning settings are rare. Common challenges include data sparsity, the lack of sufficiently flexible learner and domain models, and the difficulty of including pedagogical goals into recommendation strategies. Computational models of human cognition and learning are, in principle, well positioned to help meet these challenges, yet the effectiveness of cognitive models in educational recommender systems remains poorly understood to this date. This thesis contributes to this strand of research by investigating i) two cognitive learner models (CbKST and SUSTAIN) for resource recommendations that qualify for sparse user data by following theory-driven top down approaches, and ii) two tag recommendation strategies based on models of human cognition (BLL and MINERVA2) that support the creation of learning content meta-data. The results of four online and offline experiments in different learning contexts indicate that a recommendation approach based on the CbKST, a well-founded structural model of knowledge representation, can improve the users? perceived learning experience in formal learning settings. In informal settings, SUSTAIN, a human category learning model, is shown to succeed in representing dynamic, interest based learning interactions and to improve Collaborative Filtering for resource recommendations. The investigation of the two proposed tag recommender strategies underlined their ability to generate accurate suggestions (BLL) and in collaborative settings, their potential to promote the development of shared vocabulary (MINERVA2). This thesis shows that the application of computational models of human cognition holds promise for the design of recommender mechanisms and, at the same time, for gaining a deeper understanding of interaction dynamics in virtual learning systems.
Conference Paper
Full-text available
This paper presents a novel approach how learning paths consisting of game units can be created and adapted to learners based on their behavior during the game play. Non-invasive assessment procedures interpret the behavior and calculate information about the competences of the learners. A user model holds probabilistic information on the competence profile. Based on this competence profile game units/stories are recommended fitting to the actual competence state of the learner. This approach is part of the EC-funded TARGET project which provides the technical infrastructure regarding the 3D virtual game environment. The innovative part of this paper is the adaptive learning strategy and how it can be included in a game-based environment. The user perspective is demonstrated on a concrete scenario where the learner has to solve a task in the game-based environment.
Article
Full-text available
This paper describes a survey undertaken to discover students' wishes concerning the contents, interaction and form of open learner models in intelligent learning environments. It was found that, in general, students are receptive to the idea of using open learner models to support their learning. Several different kinds of open learner model are presented, which illustrate a range of approaches and issues relating to the use of open learner models.
Article
Full-text available
Modeling competences is an integral part of many Human Resource (HR) and e-Learning related activities. HR departments use competence descriptions to define requirements needed for performing specific tasks or jobs. The same competences are acquired by employees and applicants by e.g. experience or certifications. Typically, HR departments need to match such required and acquired competences in order to find suitable candidates. In e-Learning a similar situation arises. Curricula or training programmes need to describe pre-requisites that must be fulfilled before joining and the competences that will be acquired after successful completion. This paper analyses the limitations and extends existing approaches for modeling competences in order to allow (semi-)automatic competence matching.
Chapter
Full-text available
The main goal of the third stage of the SAFARI project is the delivery by an intelligent tutoring system (ITS) for a complete course. To achieve this goal, a subject-matter model, called CREAM, that can support course generation and delivery has been proposed. The acquisition of such a knowledge model requires to enable the designer with dedicated tools and methods. The purpose of this paper is to present the authoring environment (CREAM-Tools) we developped to support course and curriculum construction using the CREAM approach. This authoring environment consists of a course generation kit, of building tools and methodologies. Curriculums and courses produced with this environment are directly usable by an ITS. We also show ways other modules in an intelligent tutorial system can exploit the resulting curriculums and courses.
Article
This chapter presents a domain-independent computational environment which supports work-integrated learning at the professional workplace. The Advanced Process-Oriented Self-Directed Learning Environment (APOSDLE) provides learning support during the execution of work tasks (instead of beforehand), within the work environment of the user (instead of within a separate learning system), and repurposes content which was not originally intended for learning (instead of relying on the expensive manual creation of learning material). Since this definition of work-integrated learning might differ from other definitions employed within this book, a short summary of the theoretical background is provided. Along the example of the company Innovation Service Network (ISN), a network of SME's, a rich and practical description of the deployment and usage of APOSDLE is given. The chapter provides the reader with firsthand experiences and discusses efforts and lessons learned, backed up with experiences gained in two other application settings, namely EADS in France and a Chamber of Commerce and industry in Germany.
Article
This chapter explores the potential for improving long-term learning by exploiting the large amounts of data that we can readily collect about learners. We present examples of interfaces to long-term learner models that illustrate both the challenges and potential for them. The challenges include the creation of a suitable technical framework as well as associated interfaces to support a learner in transforming arbitrary collections of learning data into a lifelong learner model. We explain how this can provide new ways to support both the learner and applications in personalizing learning over the long term, taking account of the individual's long-term development. Lifelong learning deals with the full breadth of learning, be it in a formal classroom or outside (Longworth, 2003), over long periods of time. This contrasts with the current norm for educational technologies, where the learner interacts with many computer-based tools, each operating independently of the others. Each collects its own data about the learner; this means that there are many independent silos of information about the learner and their learning progress. Commonly, one learning tool holds the data associated with learning a very specific skill, or a single subject, perhaps running over weeks or months. Such silos reduce the possibility for making use of this aggregated data over the long term.
Article
The JUNG (Java Universal Network/Graph) Framework is a free, open-source software library that provides a common and extendible language for the manipulation, analysis, and visualization of data that can be represented as a graph or network. It is written in the Java programming language, allowing JUNG-based applications to make use of the extensive built-in capabilities of the Java Application Programming Interface (API), as well as those of other existing third-party Java libraries. We describe the design, and some details of the implementation, of the JUNG architecture, and provide illustrative examples of its use.