Generating syntactic tree templates for feature-based opinion mining

  • Authors:

  • Liang Wu;Yuanchun Zhou;Fei Tan;Fenglei Yang;Jianhui Li

  • Affiliations:

  • Computer Network Information Center, Chinese Academy of Sciences, China;Computer Network Information Center, Chinese Academy of Sciences, China;Computer Network Information Center, Chinese Academy of Sciences, China;Computer Network Information Center, Chinese Academy of Sciences, China;Computer Network Information Center, Chinese Academy of Sciences, China

  • Venue:
  • ADMA'11 Proceedings of the 7th international conference on Advanced Data Mining and Applications - Volume Part II
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Feature-based sentiment analysis aims to recognize appraisal expressions and identify the targets and the corresponding semantic polarity. State-of-the-art syntactic-based approaches mainly focused on designing effective features for machine learning algorithms and/or predefine some rules to extract opinion words, target words and other opinion-related information. In this paper, we present a novel approach for identifying the relation between target words and opinion words. The proposed algorithm generates tree templates by mining syntactic structures of the annotated corpus. The proposed dependency tree templates cover not only the nodes directly linked with sentiment words and target words, but also subtrees of the nodes on syntactic path, which proved to be effective features for link relation extraction between opinions and targets. Experiment results show that the proposed approach achieves the best performance on the benchmark data set and can work well when syntactic tree templates are applied to different domains.