Implicit Discourse Relation Identification for Open-domain Dialogues

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 666–672 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 666 Implicit Discourse Relation Identification for Open-domain Dialogues Mingyu Derek Ma1, Kevin K. Bowden2, Jiaqi Wu2, Wen Cui2 and Marilyn Walker2 1Human-Computer Communications Laboratory & Stanley Ho Big Data Decision Analytics Research Centre The Chinese University of Hong Kong derekma@cuhk.edu.hk 2Natural Language and Dialogue Systems Lab University of California, Santa Cruz {kkbowden, jwu64, wcui7, mawalker}@ucsc.edu Abstract Discourse relation identification has been an active area of research for many years, and the challenge of identifying implicit relations re- mains largely an unsolved task, especially in the context of an open-domain dialogue sys- tem. Previous work primarily relies on a cor- pora of formal text which is inherently non- dialogic, i.e., news and journals. This data however is not suitable to handle the nuances of informal dialogue nor is it capable of nav- igating the plethora of valid topics present in open-domain dialogue. In this paper, we designed a novel discourse relation identifi- cation pipeline specifically tuned for open- domain dialogue systems. We firstly propose a method to automatically extract the implicit discourse relation argument pairs and labels from a dataset of dialogic turns, resulting in a novel corpus of discourse relation pairs; the first of its kind to attempt to identify the dis- course relations connecting the dialogic turns in open-domain discourse. Moreover, we have taken the first steps to leverage the dialogue features unique to our task to further improve the identification of such relations by per- forming feature ablation and incorporating di- alogue features to enhance the state-of-the-art model. 1 Introduction Discourse analysis considering relations between clauses has received increasing attention from the field, and implicit discourse relation identification is one of the most challenging problems in dis- course parsing since it is purely based on textual features. Previous work has defined four widely accepted major classes of discourse relation - “Comparison”, “Expansion”, “Contingency” and “Temporal” (Miltsakaki et al., 2008; Prasad et al., 2008). These four relations can either be explicitly or implicitly realized. When explicitly realized, there are often clear connective words between clauses which result in an associated discourse re- lation, while implicit realizations are often much harder to detect. For example, people can imply there is a “Comparison” relation between the fol- lowing two sentences by understanding the mean- ing. Without clear keywords like “but” however, it is hard for machines to recognize such implicit relations. Arg 1: it’s a great album. Arg 2: it’s probably not their best. Since the development of the Penn Discourse Treebank (PDTB)1, discourse relation identifica- tion has been treated as a supervised learning problem. For explicit discourse relation pairs, simple classification methods based on connective cues achieve more than 90% accuracy (Pitler et al., 2008). For implicit discourse relations however, where there is no discourse clue, relations needs to be inferred on the basis of textual features, making this a challenging problem in …

Content truncated to 500 words