Paper: OMS-J: An Opinion Mining System for Japanese Weblog Reviews Using a Combination of Supervised and Unsupervised Approaches

ACL ID N07-4010
Title OMS-J: An Opinion Mining System for Japanese Weblog Reviews Using a Combination of Supervised and Unsupervised Approaches
Venue Human Language Technologies
Session Main Conference
Year 2007
Authors

We introduce a simple opinion mining system for analyzing Japanese Weblog re- views called OMS-J. OMS-J is designed to provide an intuitive visual GUI of opin- ion mining graphs for a comparison of different products of the same type to help a user make a quick purchase de- cision. We first use an opinion mining method using a combination of supervised (a Naive Bayes Classifier) and unsuper- vised (an improved SO-PMI: Semantic Orientation Using Pointwise Mutual In- formation) learning.