A Unified Span-Based Approach for Opinion Mining with Syntactic Constituents

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1795–1804 June 6–11, 2021. ©2021 Association for Computational Linguistics 1795 A Unified Span-Based Approach for Opinion Mining with Syntactic Constituents Qingrong Xia1, Bo Zhang2, Rui Wang3∗, Zhenghua Li1†, Yue Zhang2, Fei Huang2, Luo Si2, Min Zhang1 1Institute of Artificial Intelligence, School of Computer Science and Technology, Soochow University, China 2Alibaba Group 3Vipshop (China) Co., Ltd. 1kirosummer.nlp@gmail.com,{zhli13,minzhang}@suda.edu.cn 2{klayzhang.zb,shiyu.zy,f.huang,luo.si}@alibaba-inc.com 3mars198356@hotmail.com Abstract Fine-grained opinion mining (OM) has achieved increasing attraction in the natural language processing (NLP) community, which aims to find the opinion structures of “Who expressed what opinions towards what” in one sentence. In this work, motivated by its span- based representations of opinion expressions and roles, we propose a unified span-based approach for the end-to-end OM setting. Fur- thermore, inspired by the unified span-based formalism of OM and constituent parsing, we explore two different methods (multi-task learning and graph convolutional neural network) to integrate syntactic constituents into the proposed model to help OM. We conduct experiments on the commonly used MPQA 2.0 dataset. The experimental results show that our proposed unified span-based approach achieves significant improvements over previous works in the exact F1 score and reduces the number of wrongly-predicted opinion expressions and roles, showing the effectiveness of our method. In addition, incorporating the syntactic constituents achieves promising improvements over the strong baseline enhanced by contextualized word representations. 1 Introduction Opinion mining (OM), which aims to find the opin- ion structures of “Who expressed what opinions towards what.” in one sentence, has achieved much attention in recent years (Katiyar and Cardie, 2016; Marasovi´c and Frank, 2018; Zhang et al., 2019b, 2020). The opinion analysis has many NLP appli- cations, such as social media monitoring (Bollen et al., 2011) and e-commerce applications (Cui et al., 2017). The commonly used benchmark ∗ Rui Wang’s contributions were carried out while at Alibaba Group. †Corresponding author. John is happy because he loves being Enderly Park . Holder Target Holder Target Figure 1: An example of OM, where the blue, yellow, and green blocks denote the opinion expressions, hold- ers, and targets, respectively. MPQA (Wiebe et al., 2005) uses span-based anno- tations to represent opinion expressions and roles. Figure 1 gives an example of its opinion structures with two opinion expressions and related roles. Previous OM works (Yang and Cardie, 2013; Katiyar and Cardie, 2016; Quan et al., 2019; Zhang et al., 2020) mainly treat it as a BMESO-style tag- ging problem, which converts opinion expressions and opinion roles (holder/target) into BMESO- based labels and uses a linking module to connect the predicted expressions and roles. The B, M, and E represent the beginning, middle, and ending word of a role, S denotes a single-word role, and O de- notes other words. However, this kind of method is not perfect for the end-to-end OM setting, because one word can only belong to one opinion role (one word has only one label), while there exist overlap- ping …

Content truncated to 500 words