Title: Roles of Shared Ontology in AI-ED Research --
Intelligence, Conceptualization, Standardization, and Reusability
Author: Riichiro Mizoguchi, Mitsuru Ikeda, and Katherine Sinitsa
Reference: Proc. of AIED-97, Kobe, Japan, pp.167-174 (1997)
A lot of research on AI-ED has been done to date. Although we have many theories and implemented systems some of which are used in practice, we still need promising directions to which effort should be devoted in order to enable further progress of AI-ED research.
This paper discusses long-term perspectives of AI-ED research aiming at giving a clear view of what we need for further promoting the research and for enjoying the bright prosperity of AI-ED community. The main topic here is how to engineer knowledge in IESs. To do this, we analyze intelligent systems and show one of the essential properties common to existing intelligent systems is "Declarative representation of what the system knows". On the basis of this observation, we discuss the importance of ontology engineering which is a innovative research area in artificial intelligence. Ontology plays several roles critical to overcoming the drawbacks which the current IESs have. (1) It makes systems smart and reflexive. (2) It explicates the conceptualization on which the system is based. (3) It contributes to standardization of vocabulary. (4) It enables them to be literate and hence to communicate with humans. (5) It makes knowledge reusable, and so forth. In this paper, we discuss how ontology contributes to IES research in general and exemplify how it makes authoring systems smarter.
Article (PDF file, 43 KB)