Abstract
During dialog with a customer for addressing his/her complaint the chatbot may pose questions or observations based on its underlying model. Sometimes the questions or observations posed may not be relevant given the nature of complaint and the current set of beliefs that the customer holds. In this paper we present a framework to build conversation system that addresses customer complaints in a meaningful manner using domain understanding, opinion analysis and epistemic reasoning 1 . Extraction of latent beliefs assists in performing epistemic reasoning to maintain a meaningful conversation with the customer.