Prediction, Selection, and Generation: Exploration of Knowledge-Driven Conversation System

Abstract

In open-domain conversational systems, it is important but challenging to leverage background knowledge. We can use the incorporation of knowledge to make the generation of dialogue controllable, and can generate more diverse sentences that contain real knowledge. In this paper, we combine the knowledge bases and pre-training model to propose a knowledge-driven conversation system. The system includes modules such as dialogue topic prediction, knowledge matching and dialogue generation. Based on this system, we study the performance factors that maybe affect the generation of knowledge-driven dialogue: topic coarse recall algorithm, number of knowledge choices, generation model choices, etc., and finally made the system reach state-of-the-art. These experimental results will provide some guiding significance for the future research of this task. As far as we know, this is the first work to study and analyze the effects of the related factors.

References in corpus (7)

• A Neural Conversational Model
• Neural Natural Language Inference Models Enhanced with External Knowledge
• A Simple, Fast Diverse Decoding Algorithm for Neural Generation
• Multi-Stage Document Ranking with BERT
• Explicit State Tracking with Semi-Supervision for Neural Dialogue Generation
• GLGE: A New General Language Generation Evaluation Benchmark
• Diverse, Controllable, and Keyphrase-Aware: A Corpus and Method for News Multi-Headline Generation