International Sentiment Analysis for News and Blogs

Mikhail Bautin, Lohit Vijayarenu, Steven Skiena

There is a growing interest in mining opinions using sentiment analysis methods from sources such as news, blogs and product reviews. Most of these methods have been developed for English and are difficult to generalize to other languages. We explore an approach utilizing state-of-the-art machine translation technology and perform sentiment analysis on the English translation of a foreign language text. Our experiments indicate that (a) entity sentiment scores obtained by our method are statistically significantly correlated across nine languages of news sources and five languages of a parallel corpus; (b) the quality of our sentiment analysis method is largely translator independent; (c) after applying certain normalization techniques, our entity sentiment scores can be used to perform meaningful cross-cultural comparisons.

Subjects: 13. Natural Language Processing; 13.2 Machine Translation

Submitted: Feb 6, 2008

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.