Movie Review Mining: a Comparison between Supervised and Unsupervised Classification Approaches

  • Authors:
  • Pimwadee Chaovalit;Lina Zhou

  • Affiliations:
  • University of Maryland, Baltimore County;University of Maryland, Baltimore County

  • Venue:
  • HICSS '05 Proceedings of the Proceedings of the 38th Annual Hawaii International Conference on System Sciences (HICSS'05) - Track 4 - Volume 04
  • Year:
  • 2005

Quantified Score

Hi-index 0.01

Visualization

Abstract

Web content mining is intended to help people discover valuable information from large amount of unstructured data on the web. Movie review mining classifies movie reviews into two polarities: positive and negative. As a type of sentiment-based classification, movie review mining is different from other topic-based classifications. Few empirical studies have been conducted in this domain. This paper investigates movie review mining using two approaches: machine learning and semantic orientation. The approaches are adapted to movie review domain for comparison. The results show that our results are comparable to or even better than previous findings. We also find that movie review mining is a more challenging application than many other types of review mining. The challenges of movie review mining lie in that factual information is always mixed with real-life review data and ironic words are used in writing movie reviews. Future work for improving existing approaches is also suggested.