Abstract:Virtual assistants aim to build a human-like conversational agent. However, current human-machine conversations still cannot make users feel intelligent enough to build a continued dialog over time. Some responses from agents are usually inconsistent, uninformative, less-engaging and even memoryless. In recent years, most researchers have tried to employ conversation context and external knowledge, e.g. wiki pages and knowledge graphs, into the model which only focuses on solving some special conversation problems in local perspectives. Few researchers are dedicated to the whole capability of the conversational agent which is endowed with abilities of not only passively reacting the conversation but also proactively leading the conversation. In this paper, we first explore the essence of conversations among humans by analyzing real dialog records. We find that some conversations revolve around the same context and topic, and some require additional information or even move on to a new topic. Base on that, we conclude three conversation modes shown in Figure 1 and try to solve how to adapt to them for a continuous conversation. To this end, we define "Adaptive Knowledge-Grounded Conversations" (AKGCs) where the knowledge is to ground the conversation within a multi-turn context by adapting to three modes. To achieve AKGC, a model called MNDB is proposed to model natural dialog behaviors for multi-turn response selection. To ensure a consistent response, MNDB constructs a multi-turn context flow. Then, to mimic user behaviors of incorporating knowledge in natural conversations, we design a ternary-grounding network along with the context flow. In this network, to gain the ability to adapt to diversified conversation modes, we exploit multi-view semantical relations among response candidates, context and knowledge. Thus, three adaptive matching signals are extracted for final response selection. Evaluation results on two benchmarks indicate that MNDB can significantly outperform state-of-the-art models.

Knowledge-Enhanced Interactive Matching Network for Multi-turn Response Selection in Medical Dialogue Systems

A Multi Domain Knowledge Enhanced Matching Network for Response Selection in Retrieval-Based Dialogue Systems

Response Ranking with Deep Matching Networks and External Knowledge in Information-seeking Conversation Systems

Learning to Detect Relevant Contexts and Knowledge for Response Selection in Retrieval-based Dialogue Systems

Hierarchical Interactive Matching Network for Multi-turn Response Selection in Retrieval-Based Chatbots

Learning Multi-turn Response Selection in Grounded Dialogues with Reinforced Knowledge and Context Distillation

MedKP: Medical Dialogue with Knowledge Enhancement and Clinical Pathway Encoding

Multi-level Spatio-temporal Matching Network for Multi-turn Response Selection in Retrieval-based Dialogue Systems

Adapting to Context-Aware Knowledge in Natural Conversation for Multi-Turn Response Selection

Research on Medical Dialogue Generation of External Knowledge

Improving knowledge-based dialogue generation through two-stage knowledge selection and knowledge selection-guided pointer network

A Knowledge-Enhanced Hierarchical Reinforcement Learning-Based Dialogue System for Automatic Disease Diagnosis

Deep Hybrid Networks Based Response Selection for Multi-turn Dialogue Systems

Dialogue History Matters! Personalized Response Selection in Multi-Turn Retrieval-Based Chatbots

Dialogue History Matters! Personalized Response Selectionin Multi-turn Retrieval-based Chatbots

Improve the Response Diversity of Multi-turn Dialogue System by Combining Knowledge

Learning to Select Knowledge for Response Generation in Dialog Systems

A novel DL-based algorithm integrating medical knowledge graph and doctor modeling for Q&A pair matching in OHP

Response Ranking with Multi-types of Deep Interactive Representations in Retrieval-based Dialogues

ChatDoctor: A Medical Chat Model Fine-tuned on LLaMA Model using Medical Domain Knowledge

Knowledge-Grounded Dialogue with Reward-Driven Knowledge Selection