Proceedings of the Twenty-Second AAAI Conference on Artificial Intelligence
BlogVox: Learning Sentiment Classifiers
July 22, 2007
Performing sentiment analysis upon a topic, specified by key words, without prior knowledge about the key words is a difficult task. With the growth of the blogosphere researchers, corporations, and politicians, among others are very interested in applying sentiment detection to blogs. To accommodate the demands from myriad users, with similarly diverse desires, a sentiment analysis engine for blogs must discover domain specific features relevant to queries in order to accurately assess the sentiment of blogs. Using meta-learning upon the results of web searches, as BlogVox does, can accomplish this goal
InProceedings
Menlo Park, CA; Cambridge, MA; London; AAAI Press; MIT Press; 1999
22
1888-1889
Downloads: 2052 downloads
Google Scholar Citations: 2 citations