Clarvoynt's text classification module is named PORFIAU. Porfiau is an automated text classification engine, able to quickly scan millions of documents and identify those with a certain style of content.  
Porfiau has scored highly in two global, academic contests.  It has been proven to detect similarities in content and can identify an author's writing style.
Porfiau goes far beyond traditional sentiment analysis.  It works in any language and does not use key words. It detects even the most subtle content signal and in a test run by a major consulting firm was shown to outperform humans in eDiscovery legal review odf documents.

Porfiau is trained easily by submitting several documents with the desired content/style, along with several sample documents that do not possess the desired signal.  Porfiau then trains automatically and creates a model for scoring any new document against the selected signal.

Once created, the Porfiau model can be implemented for analysis of streaming text content, or it can be run against existing archives.  The model is extremely fast at performing the classification.
Porfiau can be configured to run against text columns in any database loaded into Clarvoynt for automated classification according to the models you have created.