An approach to sentiment analysis of movie reviews: Lexicon based vs. classification

Abstract

The paper examines two approaches to sentiment analysis: lexicon-based vs. supervised learning in the domain of movie reviews. In evaluation, the methods were compared using a standard movie review test collection. The results show that lexicon-based approach is easily outperformed by classification approach. © 2014 Springer International Publishing.

Publication
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Related