SENTIMENT CLASSIFICATION OF MOVIE REVIEWS BY SUPERVISED MACHINE LEARNING APPROACHES
Journal Title: Indian Journal of Computer Science and Engineering - Year 2013, Vol 4, Issue 4
Abstract
Large volumes of data are available in the web. The discussion forum, review sites, blogs and news corpora are some of the opinion rich resources. The opinions obtained from those can be classified and used for gathering online customer’s preferences. Techniques are being applied to design a system that identifies and classify opinions spread largely in the internet. Few different problems such as sentiment classification, feature based classification and handling negotiations are predominating this research community. This paper studies online movie reviews using sentiment analysising approaches. In this study, sentiment classification techniques were applied to movie reviews. Specifically, we compared three supervised machine learning approaches SVM, Navie Bayes and kNN for Sentiment Classification of Reviews. Empirical results states that SVM approach outperformed the Navie Bayes and kNN approaches, and the training dataset had a large number of reviews, SVM approach reached accuracies of atleast 80%.
Authors and Affiliations
P. Kalaivani , Dr. K. L. Shunmuganathan
Keywords
Related Articles
- EP ID EP98799
- DOI -
- Views 164
- Downloads 0
PERFORMANCE TESTING CONCURRENT ACCESS ISSUE AND POSSIBLE SOLUTIONS – A CLASSIC CASE OF PRODUCER-CONSUMER
Concurrent access issue is one of the most frequently encountered issues in performance testing. The issue is a classic case of producer consumer problem where multiple clients try to access same code on the server at th...
The Anatomy of Web Search Result Clustering and Search Engines
World Wide Web is a very large distributed digital information space. The ability to search and retrieve information from the Web efficiently and effectively is an enabling technology for realizing its full potential. Cu...
Different Views of Software Agent Paradigm
Agent modelling in software engineering is a relatively young area, and there are, as yet, no standard methodologies, development tools, or software architectures. Although our work is emphasized on to represent the agen...
REGULATED RR-MAC PROTOCOL FOR WIRELESS SENSOR NETWORKS
This paper proposes RR-MAC, a medium-access control (MAC) protocol used for sensor networks. Wireless sensor networks uses battery power to sense the devices. It consists of spatially distributed autonomous sensors to mo...
0-EDGE MAGIC LABELING FOR SOME CLASS OF GRAPHS
In this paper we introduce new labeling called 0-Edge Magic Labeling, and also shown the existence of this Labeling for Some Class of graphs.