The SCHOLAR legacy: A new look at the affordances of semantic networks for conversational agents in intelligent tutoring systems


The time is ripe for a new look at the affordances of semantic networks as backbone structures for knowledge representation in intelligent tutoring systems (ITSs). While the semantic space approach has undeniable value, and will likely continue to be an essential part of solutions to the problem of computer-based dialogue with humans, technical advances such the automatic extraction of ontologies from text corpora, now encourage a vision in which intelligent tutoring agents have access to forms of knowledge representation that allow them to more fully "understand" something of what they are talking about with learners. These developments have important implications for key ITS components including the structure of expert domain models, learner models, instructional modules, and dialogue strategies, particularly in respect to issues of transportability across systems. As such, they in turn have important implications for the design of a general-purpose framework such as the U.S. Army's Generalized Intelligent Framework for Tutoring (GIFT).

Publication Title

CEUR Workshop Proceedings

This document is currently not available here.