Aspect Sentiment Quad Prediction as Paraphrase Generation

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 9209–9219 November 7–11, 2021. c⃝2021 Association for Computational Linguistics 9209 Aspect Sentiment Quad Prediction as Paraphrase Generation ∗ Wenxuan Zhang1, Yang Deng1, Xin Li2, Yifei Yuan1, Lidong Bing2 and Wai Lam1 1The Chinese University of Hong Kong 2DAMO Academy, Alibaba Group {wxzhang,ydeng,yfyuan,wlam}@se.cuhk.edu.hk {xinting.lx,l.bing}@alibaba-inc.com Abstract Aspect-based sentiment analysis (ABSA) has been extensively studied in recent years, which typically involves four fundamental sentiment elements, including the aspect category, aspect term, opinion term, and sentiment polarity. Ex- isting studies usually consider the detection of partial sentiment elements, instead of predict- ing the four elements in one shot. In this work, we introduce the Aspect Sentiment Quad Pre- diction (ASQP) task, aiming to jointly detect all sentiment elements in quads for a given opinionated sentence, which can reveal a more comprehensive and complete aspect-level sen- timent structure. We further propose a novel PARAPHRASE modeling paradigm to cast the ASQP task to a paraphrase generation process. On one hand, the generation formulation al- lows solving ASQP in an end-to-end manner, alleviating the potential error propagation in the pipeline solution. On the other hand, the se- mantics of the sentiment elements can be fully exploited by learning to generate them in the natural language form. Extensive experiments on benchmark datasets show the superiority of our proposed method and the capacity of cross- task transfer with the proposed unified PARA- PHRASE modeling framework. 1 Introduction As a fine-grained opinion mining problem, aspect- based sentiment analysis (ABSA) aims to anal- yse sentiment information at the aspect level (Liu, 2012; Pontiki et al., 2014). Typically, four funda- mental sentiment elements are involved in ABSA, including 1) aspect category denoting the type of the concerned aspect; 2) aspect term which can be either explicitly or implicitly mentioned in the given text; 3) opinion term which describes the ∗Work done when Wenxuan Zhang was an intern at Al- ibaba. This work was supported by Alibaba Group through Alibaba Research Intern Program, and a grant from the Re- search Grant Council of the Hong Kong Special Administra- tive Region, China (Project Codes: 14204418). opinion towards the aspect; and 4) sentiment polar- ity denoting the sentiment class. Given an example sentence “The pasta is over-cooked!”, the senti- ment elements are “food quality”, “pasta”, “over- cooked”, and “negative”, respectively. Due to its broad application scenarios, many re- search efforts have been made on ABSA to predict or extract those sentiment elements (Pontiki et al., 2014, 2015, 2016). Early studies focus on the pre- diction of a single element such as aspect term extraction (Liu et al., 2015; Xu et al., 2018), as- pect category detection (Zhou et al., 2015), aspect sentiment classification based on either an aspect category (Ruder et al., 2016; Hu et al., 2019a) or an aspect term (Huang and Carley, 2018). More recent works propose to extract multiple associated sentiment elements at the same time (Zhang et al., 2021). For example, Chen et al. (2020) consider the aspect …

Content truncated to 500 words