CGF: Constrained Generation Framework for Query Rewriting in Conversational AI

Anonymous

Keywords: query rewriting, constrained generation

Abstract: In conversational AI agents, Query Rewriting (QR) plays a crucial role in reducing users frictions and satisfying their daily demands. Users frictions are caused by various reasons, such as errors in the spoken dialogue system, users’ accent or their abridged language. In this work, we present a novel Constrained Generation Framework (CGF) for query rewriting at both global and personalized level. The proposed framework is based on the encoder-decoder framework and consists of a context-enhanced encoding and constrained generation decoding phrases. The model takes the query and its previous dialogue context information as the encoder input, then the decoder relies on the pre-defined global or personalized constrained decoding space to generate the rewrites. Extensive offline and online experimental results show that the proposed CGF significantly boosts the query rewriting performance.