Information Retrieval based on Sentiment Analysis

Bloggers have been making a huge impact on society by representing and influencing the people. Topics like politics sports business and entertainment being written in blogosphere generally represent the sentiments of the society. Also blogging by nature is about expressing and listening to opinion. Hence a good sentiment detection tools for blogs and other social media can be a useful tool for today's society. Currently no sentiment analysis framework for information retrieval based on domain independent exist. In this book I present Blogvox2 an information retrieval based domain independent sentiment analysis framework that uses customized pattern matching techniques such as nave bayesian filter bag of words and part of speech tagging are used for opinion extraction in blogs. The modular framework provides a platform where new modules for different domains can be easily plugged in. Also we developed different applications to ease the use of our framework like the two and four panel approach identifying the hot topics and also obtaining the current top topics.