Dynamic Fusion Network for Multi-Domain End-to-end Task-Oriented Dialog

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 6344–6354 July 5 - 10, 2020. c⃝2020 Association for Computational Linguistics 6344 Dynamic Fusion Network for Multi-Domain End-to-end Task-Oriented Dialog Libo Qin†, Xiao Xu†, Wanxiang Che†∗, Yue Zhang‡, Ting Liu† †Research Center for Social Computing and Information Retrieval †Harbin Institute of Technology, China ‡School of Engineering, Westlake University, China ‡Institute of Advanced Technology, Westlake Institute for Advanced Study †{lbqin, xxu, car,tliu}@ir.hit.edu.cn, yue.zhang@wias.org.cn Abstract Recent studies have shown remarkable suc- cess in end-to-end task-oriented dialog system. However, most neural models rely on large training data, which are only available for a certain number of task domains, such as nav- igation and scheduling. This makes it difficult to scalable for a new domain with limited la- beled data. However, there has been relatively little research on how to effectively use data from all domains to improve the performance of each domain and also unseen domains. To this end, we investigate methods that can make explicit use of domain knowledge and intro- duce a shared-private network to learn shared and specific knowledge. In addition, we pro- pose a novel Dynamic Fusion Network (DF- Net) which automatically exploit the relevance between the target domain and each domain. Results show that our model outperforms exist- ing methods on multi-domain dialogue, giving the state-of-the-art in the literature. Besides, with little training data, we show its transfer- ability by outperforming prior best model by 13.9% on average. 1 Introduction Task-oriented dialogue systems (Young et al., 2013) help users to achieve specific goals such as restau- rant reservation or navigation inquiry. In recent years, end-to-end methods in the literature usually take the sequence-to-sequence (Seq2Seq) model to generate a response from a dialogue history (Eric and Manning, 2017; Eric et al., 2017; Madotto et al., 2018; Wen et al., 2018; Gangi Reddy et al., 2019; Qin et al., 2019b; Wu et al., 2019a). Taking the dialogue in Figure 1 as an example, to answer the driver’s query about the “gas station”, the end-to-end dialogue system directly generates sys- tem response given the query and a corresponding knowledge base (KB). ∗Email corresponding. Address Distance POI type POI Traffic info 5672 barringer street 5 miles certain address 5672 barringer street no traffic 200 Alester Ave 2 miles gas station Valero road block nearby 899 Ames Ct 5 miles hospital Stanford Childrens Health moderate traffic 481 Amaranta Ave 1 miles parking garage Palo Alto Garage R moderate traffic Driver Address to the gas station. Dialogue Knowledge Base (KB) Car Valero is located at 200 Alester Ave. Car Since there is a road block nearby, I found another route for you and I sent it on your screen. Driver OK , please give me directions via a route that avoids all heavy traffic. Figure 1: Example of a task-oriented dialogue that incorporates a knowledge base (KB) from the SMD dataset (Eric et al., 2017). Words with the same color refers queried entity from the KB. Better viewed in color. Though …

Content truncated to 500 words