SenseNews (from hakia): Latest innovation in stock analysis

This might be the first commercial service based on such an analysis, but it's hardly new. See, for instance:

Financial Analysis Using News Data:

http://www2.cs.sunysb.edu/~wbzhang/papers/RPE_final.pdf

Trading strategies to exploit news sentiment:

http://ftp.cs.sunysb.edu/~wbzhang/papers/stock.pdf

Text Mining Systems for Predicting Market Response to News:

http://www.iadis.net/dl/final_uploads/200708C025.pdf

Newscats: A news categorization and trading system

http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4053143

Trading Strategies To Exploit Blog and News Sentiment

http://www.aaai.org/ocs/index.php/ICWSM/ICWSM10/paper/downlo...

SOPS: Stock Prediction Using Web Sentiment

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.114...

News and Trading Rules

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.72....

Modelling the Stock Market using Twitter

http://homepages.inf.ed.ac.uk/miles/msc-projects/wolfram.pdf

etc, etc, etc.. the research papers in this field just go on and on and on.

Everything from mainstream news, to blogs, to stock message boards, and even stock-themed spam has been analyzed and used to devise trading rules and predict market conditions.