Improving aspect-based sentiment analysis via aligning aspect embedding

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journal

View graph of relations

Author(s)

  • Xingwei Tan
  • Yi Cai
  • Jingyun Xu
  • Ho-Fung Leung
  • Qing Li

Related Research Unit(s)

Detail(s)

Original languageEnglish
Pages (from-to)336-347
Journal / PublicationNeurocomputing
Volume383
Online published12 Dec 2019
Publication statusPublished - 28 Mar 2020

Abstract

Aspect-Based Sentiment Analysis (ABSA) is a fine-grained sentiment analysis task, which aims to predict sentiment polarities of given aspects or target terms in text. ABSA contains two subtasks: Aspect-Category Sentiment Analysis (ACSA) and Aspect-Term Sentiment Analysis (ATSA). Aspect embeddings have been extensively used for representing aspect-categories on ACSA task. Based on our observations, existing aspect embeddings cannot properly represent the relation between aspect-categories and aspect-terms. To address this limitation, this paper presents a learning method which trains aspect embeddings according to the relation between aspect-categories and aspect-terms. According to the cosine measure metric we proposed in this paper, the limitation is successfully alleviated in the aspect embeddings which are trained by our method. The trained aspect embeddings can be used as initialization in existing models to solve ACSA task. We conduct experiments on SemEval datasets for ACSA task, and the results indicate that our pre-trained aspect embeddings are capable of improving the performance of sentiment analysis.

Research Area(s)

  • Aspect embedding, Representation learning, Sentiment analysis

Citation Format(s)

Improving aspect-based sentiment analysis via aligning aspect embedding. / Tan, Xingwei; Cai, Yi; Xu, Jingyun; Leung, Ho-Fung; Chen, Wenhao; Li, Qing.

In: Neurocomputing, Vol. 383, 28.03.2020, p. 336-347.

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journal