Conversational Architectures Project


Access the Conversational Architecture home page


Here is a sample paper on the project:

A Computational Architecture for Conversation

Eric Horvitz
Microsoft Research
Redmond, Washington 98052

Tim Paek
Stanford University
Stanford, CA 94305

Author email: horvitz@microsoft.com, paek@psych.stanford.edu

Abstract:

We describe representation, inference strategies, and control procedures employed in an automated conversation system named the Bayesian Receptionist. The prototype is focused on the domain of dialog about goals typically handled by receptionists at the front desks of buildings on the Microsoft corporate campus. The system employs a set of Bayesian user models to interpret the goals of speakers given evidence gleaned from a natural language parse of their utterances. Beyond linguistic features, the domain models take into consideration contextual evidence, including visual findings. We discuss key principles of conversational actions under uncertainty and the overall architecture of the system, highlighting the use of a hierarchy of Bayesian models at different levels of detail, the use of value of information to control question asking, and application of expected utility to control progression and backtracking in conversation.

Click here to access postscript or pdf file.

Keywords: Bayesian user modeling, common ground, joint activity, conversational systems, dialog systems, computational linguistics.

In: Proceedings of the Seventh International Conference on User Modeling, Banff, Canada, June 1999. New York: Springer Wien, pp. 201-210.

Related Papers

T. Paek and E. Horvitz, Uncertainty, Utility, and Misunderstanding: A Decision-Theoretic Perspective on Grounding in Conversational Systems, AAAI Fall Symposium on Psychological Models of Communication in Collaborative Systems, Cape Cod, MA. November 5-7, 1999.