Methodological issues in the design of intelligent and cooperative information systems

摘要

The goal of the author is to examine the question "to what end do we study cooperative dialogue?" Based on an analysis of previous work, he argues that despite more than a decade of research in artificial intelligence, our understanding of the general principles governing cooperative discourse has improved very little. Furthermore, he suggests that the standard approach of modeling examples has produced technology that is too often restricted to the chosen domain or application context and consequently difficult or impossible to generalize to other settings. The author characterizes two perspectives from which principles of cooperation can be formulated and argues that if progress on understanding cooperative dialogue phenomena is to be achieved, the dominant questioner-based perspective should be abandoned and a respondent-based perspective adopted instead.