Context-Centered Tools for Intelligent Assistant Systems

Abstract

Modeling experts’ experience requires an explicit consideration of the context and a uniform representation of elements of knowledge, reasoning and contexts. In a companion paper in this book (see Chap. 31) the author propose the Contextual-Graphs formalism for representing expertise with practices. The exploitation of such experience bases is a new challenge for designing and developing context-based support systems able to tackle context in the same way as knowledge and reasoning. The author present a conceptual framework for implementing management tools in intelligent assistant systems (IASs) that (1) work on experience described as practices, (2) deal with the process of decision-making and not the result only, and (3) build a context-specific model jointly with the decision-making process. Thus, an IAS can be equipped with domain-independent tools for managing the experience base, simulating practice development, explaining the rationale behind each practice, incrementally acquiring knowledge and learning practice. This chapter shows that functions like acquisition, learning and explanation, which were considered separately, become naturally integrated in IASs.