a.k.a. opinion mining, natural language processing, computational linguistics, text analytics
Sentiment analysis is a research method that aims to determine the attitude of a speaker or a writer with respect to a particular newsgroup topic, blog post, website comment, or overall tonality of a document or Web page. The attitude may be his or her judgment (or evaluation), the emotional state of the author, or the intended emotional communication.
The basic task of sentiment analysis is to classify the emotional degree of a given word in a document or sentence--whether the expressed opinion is positive, negative, or neutral. Beyond the basic emotional degree of a statement, sentiment classification looks, for instance, at emotional states such as "angry," "sad," and "happy."
However, senitment analysis can be tricky when performed by computers (or algorithms) because they often lack the "sentiment" of humans; in other words, computers do not always understand the semantics associated with feeling. For example, if a user describes a particular game as being "sick" in a gamer blog, it means something good, but the same word "sick" is not associated as a positive sentiment within health-oriented websites.
The rise of social media, such as blogs and social networking, has fueled interest in sentiment analysis. With the proliferation of reviews, ratings, recommendations and other forms of online expression, online opinion has turned into a kind of virtual currency for businesses looking to market their products, identify new opportunities and manage their reputations. As businesses look to automate the process of filtering out the noise, listening to the conversations, and identifying the relevant content so as to act appropriately, many are now moving away from market research and looking to the field of sentiment analysis. According to Wikipedia, if Web 2.0 was all about democratizing publishing, then the next stage of the Web may well be based on democratizing data mining of all the content that is getting published.