Modern text mining methods and tools are based on simple techniques that are oblivious to the semantics of the domains they operate in. This leads to a lack of functionality, poor results and extensive expenditure of human expert's time to deliver such systems. For instance, tagging systems and ontology management systems cannot recommend improvements and reorganizations of user's ontologies, document classification systems operate with a set of predefined categories and require extensive training sets to function, and fact extraction tool require labor of linguistic and domain area experts.
Texterra takes a novel approach to text mining. It utilizes semantics extracted from user-generated ontologies (such as Wikipedia, etc.). This allows us to understand the meaning of ontologies and find obvious opportunities for improvement, classify documents without using training sets and with high accuracy, extract facts automatically.