Recognizing Conflict Opinions in Aspect-level Sentiment Classification with Dual Attention Networks

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, pages 3426–3431, Hong Kong, China, November 3–7, 2019. c⃝2019 Association for Computational Linguistics 3426 Recognizing Conflict Opinions in Aspect-level Sentiment Classification with Dual Attention Networks Xingwei Tan, Yi Cai∗and Changxi Zhu School of Software Engineering, South China University of Technology Guangzhou, China setanxw5@mail.scut.edu.cn ∗ycai@scut.edu.cn Abstract Aspect-level sentiment classification, which is a fine-grained sentiment analysis task, has re- ceived lots of attention these years. There is a phenomenon that people express both positive and negative sentiments towards an aspect at the same time. Such opinions with conflict- ing sentiments, however, are ignored by ex- isting studies, which design models based on the absence of them. We argue that the exclu- sion of conflict opinions is problematic, for the reason that it represents an important style of human thinking – dialectic thinking. If a real- world sentiment classification system ignores the existence of conflict opinions when it is de- signed, it will incorrectly mixed conflict opin- ions into other sentiment polarity categories in action. Existing models have problems when recognizing conflicting opinions, such as data sparsity. In this paper, we propose a multi- label classification model with dual attention mechanism to address these problems. 1 Introduction Aspect-level sentiment classification is a fine- grained sentiment analysis task (Pang et al., 2008). Given aspect categories or target entities in the text, this task aims at inferring the sentiment po- larity (e.g., positive, negative, or neutral) of the aspects. A recent study (Kenyon-Dean et al., 2018) on sentiment analysis dataset construction brings attention to sentiments beyond the scope of positive, negative, and neutral categories, indicat- ing that they will benefit the real-world sentiment analysis systems. We find that there is a similar type of sentiment which is labeled as conflict in SemEval dataset for aspect-level sentiment classi- fication (Pontiki et al., 2014). Conflict label ap- plies when both positive and negative sentiment is expressed about an aspect. For example, the ∗Corresponding author sentence “The atmosphere is attractive, but a lit- tle uncomfortable.” expresses conflict sentiment towards ambience aspect. However, most of ex- isting studies ignore conflict opinions, for the rea- son that they are sparse in the datasets (Tang et al., 2016b; He et al., 2018). In this work, we argue that the exclusion of conflict opinions is problematic. The conflict opinions are the production of an important style of human thinking – dialectic thinking. Dialectic thinking, unlike the thinking constrained by the laws of formal logic, shows a degree of tolerance of contradiction (Peng and Nisbett, 1999). If a real-world sentiment classi- fication system ignores the existence of conflict opinions when it is designed, its predicted positive opinions may contain negative expressions and its predicted negative opinions may contain positive expressions. Considering the existence of senti- ment expressions of both sites, it is unreasonable to put conflict opinions into either positive or neg- ative sentiment classes. There are problems when including conflict …

Content truncated to 500 words