Generating context-sensitive responses to object-related misconceptions

摘要

A study of transcripts of users interacting with database or expert systems reveals that users often exhibit misconceptions about the objects modeled by the systems. Simply ignoring these misconceptions may not only mislead the user, but may seriously hamper the ability of the system. This paper discusses a method for responding to such misconceptions in a domain-independent and context-sensitive fashion. The method calls for reasoning about possible sources of the misconception using a model of the user and generating a response based on this reasoning. The process is made context-sensitive by augmenting the user model with domain perspective which provides shifting highlights on the user model.