Neural Aspect and Opinion Term Extraction with Mined Rules as Weak Supervision

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 5268–5277 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 5268 Neural Aspect and Opinion Term Extraction with Mined Rules as Weak Supervision Hongliang Dai Department of CSE, HKUST hdai@cse.ust.hk Yangqiu Song Department of CSE, HKUST yqsong@cse.ust.hk Abstract Lack of labeled training data is a major bottle- neck for neural network based aspect and opin- ion term extraction on product reviews. To alleviate this problem, we first propose an al- gorithm to automatically mine extraction rules from existing training examples based on de- pendency parsing results. The mined rules are then applied to label a large amount of aux- iliary data. Finally, we study training proce- dures to train a neural model which can learn from both the data automatically labeled by the rules and a small amount of data accurately an- notated by human. Experimental results show that although the mined rules themselves do not perform well due to their limited flexibil- ity, the combination of human annotated data and rule labeled auxiliary data can improve the neural model and allow it to achieve perfor- mance better than or comparable with the cur- rent state-of-the-art. 1 Introduction There are two types of words or phrases in product reviews (or reviews for services, restaurants, etc., we use “product reviews” throughout the paper for convenience) that are of particular importance for opinion mining: those that describe a product’s properties or attributes; and those that correspond to the reviewer’s sentiments towards the product or an aspect of the product (Hu and Liu, 2004; Liu, 2012; Qiu et al., 2011; Vivekanandan and Aravin- dan, 2014). The former are called aspect terms, and the latter are called opinion terms. For ex- ample, in the sentence “The speed of this laptop is incredible,” “speed” is an aspect term, and “in- credible” is an opinion term. The task of aspect and opinion term extraction is to extract the above two types of terms from product reviews. Rule based approaches (Qiu et al., 2011; Liu et al., 2016) and learning based approaches (Jakob and Gurevych, 2010; Wang et al., 2016) are two major approaches to this task. Rule based ap- proaches usually use manually designed rules based on the result of dependency parsing to ex- tract the terms. An advantage of these approaches is that the aspect or opinion terms whose usage in a sentence follows some certain patterns can al- ways be extracted. However, it is labor-intensive to design rules manually. It is also hard for them to achieve high performance due to the variability and ambiguity of natural language. Learning based approaches model aspect and opinion term extraction as a sequence labeling problem. While they are able to obtain better per- formance, they also suffer from the problem that significant amounts of labeled data must be used to train such models to reach their full potential, especially when the input features are not manu- ally designed. Otherwise, they …

Content truncated to 500 words