This thesis evaluates the benefits of supervised methods applied in the task of text information retrieval (IR), commonly dominated by unsupervised methods. This thesis will deliver the following: 1. Overview and comparison of publicly-available IR datasets with an emphasis on the applicability of supervised approaches. 2. Implementation and evaluation of one or more supervised approaches on a selected set of datasets and qualitative comparison of the implemented method(s) to one or more strong unsupervised methods. 3. A selection of datasets where a potential of supervised method(s) has been shown, and integration of these datasets into a common evaluation framework. In the following years, the extended framework will serve as a convenient tool for evaluating the term competitions of the PV211 course.