Paper: Learning Lexical Alignment Policies for Generating Referring Expressions for Spoken Dialogue Systems

ACL ID W09-0611
Title Learning Lexical Alignment Policies for Generating Referring Expressions for Spoken Dialogue Systems
Venue European Workshop on Natural Language Generation
Session
Year 2009
Authors

We address the problem that different users have different lexical knowledge about problem domains, so that automated dialogue systems need to adapt their gen- eration choices online to the users’ domain knowledge as it encounters them. We ap- proach this problem using policy learning in Markov Decision Processes (MDP). In contrast to related work we propose a new statistical user model which incorporates the lexical knowledge of different users. We evaluate this user model by showing that it allows us to learn dialogue poli- cies that automatically adapt their choice of referring expressions online to differ- ent users, and that these policies are sig- nificantly better than adaptive hand-coded policies for this problem. The learned policies are consistently between 2 and 8 turns shorte...