Semantically Oriented Sentiment Mining in Location-Based Social Network Spaces

Carlone D., Ortiz-Arroyo D.

9th International Conference on Flexible Query Answering Systems (FQAS 2011), Ghent, Belgium, 26 - 28 October 2011, vol.7022, pp.234-245 identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 7022
  • City: Ghent
  • Country: Belgium
  • Page Numbers: pp.234-245
  • Middle East Technical University Affiliated: No

Abstract

In this paper we describe a system that performs sentiment classification of reviews from social network sites using natural language techniques. The pattern-based method used in our system, applies classification rules for positive or negative sentiments depending on its overall score, calculated with the aid of SentiWordNet. We investigate several classifier models created from a combination of different methods applied at word and review levels. Our experimental results show that using part-of-speech helps to achieve better accuracy.