A Knowledge-Driven Approach to Classifying Object and Attribute Coreferences in Opinion Mining

Findings of the Association for Computational Linguistics: EMNLP 2020, pages 1616–1626 November 16 - 20, 2020. c⃝2020 Association for Computational Linguistics 1616 A Knowledge-Driven Approach to Classifying Object and Attribute Coreferences in Opinion Mining Jiahua Chen, Shuai Wang, Sahisnu Mazumder, Bing Liu Department of Computer Science, University of Illinois at Chicago, USA jiahuaqy@gmail.com, shuaiwanghk@gmail.com sahisnumazumder@gmail.com, liub@uic.edu Abstract Classifying and resolving coreferences of ob- jects (e.g., product names) and attributes (e.g., product aspects) in opinionated reviews is cru- cial for improving the opinion mining perfor- mance. However, the task is challenging as one often needs to consider domain-specific knowledge (e.g., iPad is a tablet and has as- pect resolution) to identify coreferences in opinionated reviews. Also, compiling a hand- crafted and curated domain-specific knowl- edge base for each domain is very time con- suming and arduous. This paper proposes an approach to automatically mine and leverage domain-specific knowledge for classifying ob- jects and attribute coreferences. The approach extracts domain-specific knowledge from un- labeled review data and trains a knowledge- aware neural coreference classification model to leverage (useful) domain knowledge to- gether with general commonsense knowledge for the task. Experimental evaluation on real- world datasets involving five domains (product types) shows the effectiveness of the approach. 1 Introduction Coreference resolution (CR) aims to determine whether two mentions (linguistic referring expres- sions) corefer or not, i.e., they refer to the same entity in the discourse model (Jurafsky, 2000; Ding and Liu, 2010; Atkinson et al., 2015; Lee et al., 2017, 2018; Joshi et al., 2019; Zhang et al., 2019b). The set of coreferring expressions forms a corefer- ence chain or a cluster. Let’s have an example: [S1] I bought a green Moonbeam for myself. [S2] I like its voice because it is loud and long. Here all colored and/or underlined phrases are mentions. Considering S1 (sentence-1) and S2 (sentence-2), the three mentions “I”, “myself” in S1 and “I” in S2 all refer to the same person and form a cluster. Similarly, “its” in S2 refers to the object “a green Moonbeam” in S1 and the cluster is {“its” (S2), “a green Moonbeam” (S1) }. The mentions “its voice” and “it” in S2 refer to the same attribute of the object “a green Moonbeam” in S1 and form cluster {“its voice” (S2), “it” (S2)}. CR is beneficial for improving many down- stream NLP tasks such as question answer- ing (Dasigi et al., 2019), dialog systems (Quan et al., 2019), entity linking (Kundu et al.), and opinion mining (Nicolov et al., 2008). Particularly, in opinion mining tasks (Liu, 2012; Wang et al., 2016; Zhang et al., 2018; Ma et al., 2020), Nicolov et al. (2008) reported performance improves by 10% when CR is used. The study by Ding and Liu (2010) also supports this finding. Considering the aforementioned example, without resolving “it” in S2, it is difficult to infer the opinion about the attribute “voice” (i.e., the voice, which “it” refers to, is “loud and long”). Although CR plays such a crucial role in opinion …

Content truncated to 500 words