Ontologies and Context for Educational Process Modeling in IMS Learning Design Colin Knight 1 and Dragan Gašević 2 and Griff Richards 3 Abstract. This paper discusses the role of context and design, a mechanism is needed to guide the transfer ontologies in educational process modeling. Educational of models for use in new situations. We propose the process modeling seeks to represent the complex interac- use of contexts as a solution to this problem. tions that take place in multi-actor learning environments, with the view that the sequence and types of interactions Previous work on development of ontologies for can be equally as important as the sequence and types of eLearning has focused on the authoring process content. The IMS Learning Design specification provides [6],[7] and sequencing of content [8],[9] with rela- the semantics to represent multi-actor interactions within tively little emphasis on expressing the role of multi- an educational process, and the IDLD project has resulted actor interactions in the learning process. However, in a substantial catalogue of Learning Design models these efforts have provided a useful framework for from a variety of contexts. To facilitate reuse of these us to work within. models in different contexts, ontologies have been devel- oped based on the IMS Learning Design specification and An exploration of contextual variables for learning we propose to use context to determine the relevance of environments has been completed in [2]. Since the Learning Design models when used in new situations to variability of these environments is almost as great guide the learning process. as the variability of the designs themselves, it is nec- essary to simplify to include the context elements that had the greatest influence on the structure and 1 INTRODUCTION sequence of the course structure. This simplification usually occurs after consultation with the course au- Learning design has emerged at the forefront of re- thor or instructor. search into the modeling of dynamic, learner- centered eLearning experiences. The IMS Learning 2 LOCO - an ontology compatible with IMS- Design (LD) Specification [1] provides a way to rep- LD resent complex multi-actor interactions in an educa- tional environment. This specification has been The IMS-LD Information Model and XML binding widely integrated into a number of learning man- is the specification for Learning Design [1]. The agement systems and authoring tools, and several LOCO ontology [2] is a light-weight ontology in the ontologies have been developed around the specifi- OWL language, based on the IMS-LD Information cation [2],[3],[4]. The recently-completed IDLD [5] Model. The ontology is able to represent complex project effort involved modeling the educational series and parallel interactions of actors. Each actor processes of dozens of actual on-line and face-to- is assigned one or many Roles, which are associated face courses being delivered at universities across with Activities according to the Role-part that each Canada according to the IMS LD specification. Ef- actor engages in and in Environment (resources, ser- fective reuse of Learning Design models remains a vices) in which the activities take place. challenge because each model is gathered from a diverse learning situation, meaning that some of elements of the model become irrelevant when ap- plied to new situations. Since it is a time-consuming task to model the educational process in a learning 1 Simon Fraser University, Canada. Email cjk2@sfu.ca 2 Simon Fraser University, Canada. Email: dgasevic@sfu.ca 3 Simon Fraser University, Canada. Email: griff@sfu.ca REFERENCES [1] IMS Global Learning Consortium (2003). IMS Learning Design Information Model. Version 1.0 Final Specification, revision 20. Retrieved March 25, 2005, from http://www.imsglobal.org/learningdesign/ldv1p0/imsld_infov1p0.h tml. [2] Knight, C. Gasevic, D. & Richards, G. (2005). Ontologies to inte- grate learning design and learning content. Journal of Interactive Media in Education. Special Issue on Advances in Learning De- Figure 1. The class hierarchy of the LOCO ontology sign, 2005 (7). To create the LOCO, some changes were made to [3] Psyche, V., Bourdeau, J., Nkambou, R., Mizoguchi, R. (2005). the Information Model [1] in order to conform to Making Learning Design Standards Work with an Ontology of Educational Theories. In proceedings of the 12th Annual established best-practice recommendations for on- Conference on Artificial Intelligence in Education, Amsterdam, tology design [11], and to resolve some ambiguities NL. and inconsistencies in the information model. These [4] Amorim, R., Lama, M., Sanchez, E., Riera, A., Vila, X. (2006). A changes are described in detail in [2]. To date the Learning Design Ontology Based on IMS-LD. Jounal of Education, Technology, and Society, 9 (1), 38-57. LOCO only addresses IMS-LD Level A. [5] Paquette, G., Marino, O., Lundgren-Cayrol, K., Léonard, M., I de la Teja , I. (2006) The IDLD Repository – Classification and Re- purposing of Learning Designs. .TENCompetence Workshop, 3 CONCLUSIONS Sofia, Bulgaria, March 20. [6] Mizoguchi, R., and Bourdeau, J. (2000). Using Ontological Engineering to Overcome Common AI-ED Problems. International Journal of Artificial Intelligence in Education, 11, 107-121. Treatment of context information stands as a barrier [7] Inaba, A., & Mizoguchi, R. (2004). Learning design palette: An to the reusability of Learning Design. Existing solu- ontology-aware authoring system for learning design. In proceed- tions for the use of context and ontologies in learn- ings of the International conference on computers in education, ing applications could be enhanced by incorporating Melbourne, Australia. educational process modeling into the semantic rep- [8] Jovanović, J., Gašević, D., Verbert, K., Duval, D. (2005). Ontology of learning object content structure. In Proceedings of the 12th In- resentation of the learning space. A method of using ternational Conference on Artificial Intelligence in Education, Am- context to effectively transfer these processes to new sterdam, The Netherlands, 322-329. settings would greatly benefit learners by enabling [9] Dichev, C., & Dicheva, D. (2005). Contexts as Abstractions of pedagogical methods such as collaborative learning Grouping. In proceedings of the Context & Ontologies Workshop 2005, Pittsburg, Pennsylvania, USA. to become more of reality in eLearninig. Also, re- [10] Mizoguchi, R., Vanwelkenhuysen, J., and Ikeda, M. (1995). Task usability would enhance the cost-effectiveness of Ontology for Reuse of Problem Solving Knowledge. Towards modeling learning designs. Future research will in- Very Large Knowledge Bases. Mars, N.J.I (eds), IOS Press. volve determining suitable methods for using con- [11] Noy, F. N. & McGuinness, D.L. (2001). Ontology Development 101: A Guide to Creating Your First Ontology, Technical Report text effectively to transfer learning designs to new SMI-2001-0880, Stanford Medical Informatics,Stanford University, learning situations with minimal redesign effort. Stanford, CA, USA.