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

Abstract

Recently, knowledge-grounded conversations in the open domain gain great attention from researchers. Existing works on retrievalbased dialogue systems have paid tremendous efforts to utilize neural networks to build a matching model, where all of the context and knowledge contents are used to match the response candidate with various representation methods. Actually, different parts of the context and knowledge are differentially important for recognizing the proper response candidate, as many utterances are useless due to the topic shift. Those excessive useless information in the context and knowledge can influence the matching process and leads to inferior performance. To address this problem, we propose a multi-turn Response Selection Model that can Detect the relevant parts of the Context and Knowledge collection (RSM-DCK). Our model first uses the recent context as a query to pre-select relevant parts of the context and knowledge collection at the word-level and utterance-level semantics. Further, the response candidate interacts with the selected context and knowledge collection respectively. In the end, the fused representation of the context and response candidate is utilized to post-select the relevant parts of the knowledge collection for matching with more confidence. We test our proposed model on two benchmark datasets. Evaluation results indicate that our model achieves better performance than the existing methods, and can effectively detect the relevant context and knowledge for response selection.

Initiative-Aware Self-Supervised Learning for Knowledge-Grounded Conversations

“…(1) improving knowledge selection (KS) [29]; (2) improving knowledge-aware response generation [59] or response selection [11];…”

Learning Implicit User Profile for Personalized Retrieval-Based Chatbot

“…Early studies encoded user information by using user ID embeddings which are learned during training [1,4,16], and generate personalized responses with the help of the ID embeddings. Recently, some works focused on building personalized chatbots using manually created user profiles to maintain the consistency of personality for chatbots [11,13,17,23,36,37,48]. Such predefined user profiles are explicit user profiles which usually contain several persona descriptions or key-value-based personal information.…”

“…CSN performs document selection over the persona sentences [61]. RSM-DCK conducts selection over both dialogue context and user profiles and lets the response candidates interact with the selected results [13].…”

“…There are multiple ways to model U. Previous works either learn user ID embeddings [1,3,16] or build explicit user profiles [13,47,52,61]. For example, Zhang et al [47] used explicit persona descriptions and Qian et al [31] used key-value profile to model U.…”

OpenViDial 2.0: A Larger-Scale, Open-Domain Dialogue Generation Dataset with Visual Contexts

“…And whether a model can generate diverse (Xu et al, 2018;Baheti et al, 2018), coherent (Li et al, 2016bTian et al, 2017;Bosselut et al, 2018;Adiwardana et al, 2020), informative (Shao et al, 2017;Lewis et al, 2017;Ghazvininejad et al, 2017;Young et al, 2017;Zhao et al, 2019) and knowledge-fused (Hua et al, 2020;Zhao et al, 2020;He et al, 2020) responses or not has become metrics to evaluate a dialog generation model. However, the mainly researches described above are developed on textual only and the development of multimodal dialog generation is relatively slow since the lack of large-scale datasets.…”