Production of cooperative answers on the basis of partial knowledge in information-seeking dialogues

Abstract

This paper presents the component of a plan-based consultation system that selects the relevant information to be included in the answer to the user's question. The relevant information can be determined on the basis of both the recognized plan and the form of the input question. We classify the possible user requests by establishing a correspondence with the possible answers.We show how it is possible to respond in a cooperative way also in presence of partial knowledge about the constraints of the plans. Moreover, in order to limit the clarification dialogues that the system may need to carry out in the recognition of the user's plans and goals, we characterize the notion of relevance of the ambiguity among alternative hypotheses on the user's plans and goals on the basis of the characteristics of the constraints present in the plans.KeywordsNoun PhraseContext ModelNegative AnswerInformation SeekerPartial PlanThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.