"Motivation-Based Modeling of Utterances in Dialogue"
Yoichi Yamashita, Riichiro Mizoguchi
Proc. of 1996 International Symposium on Spoken Dialogue (ISSD '96), Philadelphia, 61-64 (1996).

Abstarct:
This paper proposes a new model of utterance emergence in dialogue based on the motivation chain. In general, we have motivations when we generate utterances and carry out dialogue. The motivation can be regarded as 'driving force' of dialogue which is indispensable to perform a goal-oriented dialogue. It is difficult to identify a unique motive which makes an utterance emerge. Multiple motives often exist for generating an utterance. We investigated motives for the utterances in dialogue corpus and categorized them into five different levels; communication establishment, problem solving, problem solving control, dialogue control, communication. In a proposed model, an utterance is driven by the motivation chain which is successive combination of several motivations at different levels. The patterns of the motivation chain are classified into six categories.

Ftp article (ps-file, 4 pages, 193308 bytes)