UniConv: A Unified Conversational Neural Architecture for Multi-domain Task-oriented Dialogues

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 1860–1877, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 1860 UniConv: A Unified Conversational Neural Architecture for Multi-domain Task-oriented Dialogues Hung Le†§, Doyen Sahoo‡, Chenghao Liu†, Nancy F. Chen§, Steven C.H. Hoi†‡ † Singapore Management University {hungle.2018,chliu}@smu.edu.sg ‡ Salesforce Research Asia {dsahoo,shoi}@salesforce.com §Institute for Infocomm Research, A*STAR nfychen@i2r.a-star.edu.sg Abstract Building an end-to-end conversational agent for multi-domain task-oriented dialogues has been an open challenge for two main reasons. First, tracking dialogue states of multiple do- mains is non-trivial as the dialogue agent must obtain complete states from all relevant do- mains, some of which might have shared slots among domains as well as unique slots specif- ically for one domain only. Second, the dia- logue agent must also process various types of information across domains, including di- alogue context, dialogue states, and database, to generate natural responses to users. Unlike the existing approaches that are often designed to train each module separately, we propose “UniConv" — a novel unified neural architec- ture for end-to-end conversational systems in multi-domain task-oriented dialogues, which is designed to jointly train (i) a Bi-level State Tracker which tracks dialogue states by learn- ing signals at both slot and domain level inde- pendently, and (ii) a Joint Dialogue Act and Response Generator which incorporates infor- mation from various input components and models dialogue acts and target responses si- multaneously. We conduct comprehensive ex- periments in dialogue state tracking, context- to-text, and end-to-end settings on the Multi- WOZ2.1 benchmark, achieving superior per- formance over competitive baselines. 1 Introduction A conventional approach to task-oriented dialogues is to solve four distinct tasks: (1) natural language understanding (NLU) which parses user utterance into a semantic frame, (2) dialogue state tracking (DST) which updates the slots and values from se- mantic frames to the latest values for knowledge base retrieval, (3) dialogue policy which determines an appropriate dialogue act for the next system re- sponse, and (4) response generation which gener- ates a natural language sequence conditioned on the dialogue act. This traditional pipeline modu- lar framework has achieved remarkable successes in task-oriented dialogues (Wen et al., 2017; Liu and Lane, 2017; Williams et al., 2017; Zhao et al., 2017). However, such kind of dialogue system is not fully optimized as the modules are loosely inte- grated and often not trained jointly in an end-to-end manner, and thus may suffer from increasing error propagation between the modules as the complexity of the dialogues evolves. A typical case of a complex dialogue setting is when the dialogue extends over multiple domains. A dialogue state in a multi-domain dialogue should include slots of all applicable domains up to the current turn (See Table 1). Each domain can have shared slots that are common among domains or unique slots that are not shared with any. Directly applying single-domain DST to multi-domain dia- logues is not straightforward because the dialogue states extend to multiple domains. A possible ap- proach is to process a …

Content truncated to 500 words