A Multi-view Discourse Framework for Integrating Semantic and Syntactic Features in Dialog Agents
This work addresses the challenge of generating more human-like responses in dialogue agents, but it is incremental as it builds on existing methods for discourse understanding.
The paper tackled the problem of response selection in retrieval-based dialogue systems by introducing a discourse-aware framework that integrates semantic and syntactic features, achieving significant improvements in automatic evaluation metrics on the Ubuntu Dialogue Corpus.
Multiturn dialogue models aim to generate human-like responses by leveraging conversational context, consisting of utterances from previous exchanges. Existing methods often neglect the interactions between these utterances or treat all of them as equally significant. This paper introduces a discourse-aware framework for response selection in retrieval-based dialogue systems. The proposed model first encodes each utterance and response with contextual, positional, and syntactic features using Multi-view Canonical Correlation Analysis (MCCA). It then learns discourse tokens that capture relationships between an utterance and its surrounding turns in a shared subspace via Canonical Correlation Analysis (CCA). This two-step approach effectively integrates semantic and syntactic features to build discourse-level understanding. Experiments on the Ubuntu Dialogue Corpus demonstrate that our model achieves significant improvements in automatic evaluation metrics, highlighting its effectiveness in response selection.