Abstract
Ambiguity often appears when humans and agents communicate. In the buying and selling behaviors in e-commerce, one of the important dialogues is the discourse of disambiguation. Namely, both selling and buying agents may often have to conduct conversation to some extent in order to disambiguate the real intents of the customers. And in the thesis, we propose four different kinds of disambiguation strategies: (1) Guessing (2) Filtering (3) Recommending and (4) Asking more hints. It also requires building four models: the world model, the mental model, the language model, and the rational model, to incorporate necessary common sense knowledge for disambiguation. We implement these strategies and models in a Buyer-Seller interaction system, and conduct experiments to measure the performance of the dialogue system to solve the referential ambiguity in the problem domain of selling computer equipments.