A Controllable Model of Grounded Response Generation

Current end-to-end neural conversation models inherently lack the flexibility to impose semantic control in the response generation process, often resulting in uninteresting responses. Attempts to boost informativeness alone come at the expense of factual accuracy, as attested by pretrained language models’ propensity to “hallucinate” facts. While this may be mitigated by access to background knowledge, there is scant guarantee of relevance and informativeness in generated responses. We propose a framework that we call controllable grounded response generation (CGRG), in which lexical control phrases are either provided by a user or automatically extracted by a control phrase predictor from dialogue context and grounding knowledge. Quantitative and qualitative results show that, using this framework, a transformer based model with a novel inductive attention mechanism, trained on a conversation-like Reddit dataset, outperforms strong generation baselines.

Authors

• Zeqiu Wu
• Michel Galley
• Chris Brockett
• Yizhe Zhang
• Xiang Gao Intuit
• Chris Quirk
• Rik Koncel-Kedziorski
• Jianfeng Gao
• Hannaneh Hajishirzi
• Mari Ostendorf
• Bill Dolan

Keywords

No keywords are indexed for this paper.