In this paper we discuss our plans and progress on analysing and integrating various methods of text mining in biomedical literature for a Ph.D project. The framework is a clientbased search engine that integrates different machine learning and text mining techniques. For convenience all text mining procedures have been split into five steps, so the different steps can be individually optimised while the rest of the procedures stay untouched. By the use of common interfaces the steps are interchangeable, thus simplifying the objective comparison between them.