Abstract:Building an intelligent dialogue system with the ability to select a proper response according to a multi-turn context is challenging in three aspects: (1) the meaning of a context–response pair is built upon language units from multiple granularities (e.g., words, phrases, and sub-sentences, etc.); (2) local (e.g., a small window around a word) and long-range (e.g., words across the context and the response) dependencies may exist in dialogue data; and (3) the relationship between the context and the response candidate lies in multiple relevant semantic clues or relatively implicit semantic clues in some real cases. However, existing approaches usually encode the dialogue with mono-type representation and the interaction processes between the context and the response candidate are executed in a rather shallow manner, which may lead to an inadequate understanding of dialogue content and hinder the recognition of the semantic relevance between the context and response. To tackle these challenges, we propose a representation [ K ] -interaction [ L ] -matching framework that explores multiple types of deep interactive representations to build context-response matching models for response selection. Particularly, we construct different types of representations for utterance–response pairs and deepen them via alternate encoding and interaction. By this means, the model can handle the relation of neighboring elements, phrasal pattern, and long-range dependencies during the representation and make a more accurate prediction through multiple layers of interactions between the context–response pair. Experiment results on three public benchmarks indicate that the proposed model significantly outperforms previous conventional context-response matching models and achieve slightly better results than the BERT model for multi-turn response selection in retrieval-based dialogue systems.

Building an Efficient Retrieval-based Dialogue System with Contrastive Learning

Building an Efficient and Effective Retrieval-based Dialogue System via Mutual Learning

Exploring Dense Retrieval for Dialogue Response Selection

Towards Efficient Coarse-grained Dialogue Response Selection

Improving Contextual Language Models for Response Retrieval in Multi-Turn Conversation

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

A Survey on Response Selection for Retrieval-based Dialogues.

Post-encoding and Contrastive Learning Method for Response Selection Task

Improving BERT with local context comprehension for multi-turn response selection in retrieval-based dialogue systems

Self-attention Comparison Module for Boosting Performance on Retrieval-based Open-Domain Dialog Systems

CORE: Cooperative Training of Retriever-Reranker for Effective Dialogue Response Selection

Deep context modeling for multi-turn response selection in dialogue systems

Speaker-Aware BERT for Multi-Turn Response Selection in Retrieval-Based Chatbots

Fine-grained Post-training for Improving Retrieval-based Dialogue Systems

Small Changes Make Big Differences: Improving Multi-turn Response Selection in Dialogue Systems Via Fine-Grained Contrastive Learning

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

Learning an Effective Context-Response Matching Model with Self-Supervised Tasks for Retrieval-based Dialogues

BERT-BC: A Unified Alignment and Interaction Model over Hierarchical BERT for Response Selection

Enhancing Dialogue Generation Via Multi-Level Contrastive Learning.

DialogBERT: Discourse-Aware Response Generation Via Learning to Recover and Rank Utterances

Learning to Match Representations is Better for End-to-End Task-Oriented Dialog System