Exploiting text mining in publishing and education

The paper describes an experience on using Text mining methods for enhancing a customer application in the area of publishing and education. We described the whole process of identifying customers with potentially interesting problems and narrowing down to the one of them followed by the five text mining phases (adopted from CRISP data mining methodology). From the application and research point of view, the project involved in particular:(1) taxonomy/ontology building from a plain set of documents,(2) searching the document database, and (3) addressing (non-English) language specific issues.