• This GitHub project contains the Java code (based on [Lucene](, [Sesame](, and [RDFpro]( implementing a simple evaluation system that allow configuring and evaluating KE4IR on arbitrary document collections and queries for which relevance judgments are known. You can use this code, together with the data available on KE4IR [webpage](, to replicate the evaluation results reported in the KE4IR paper. You can also use this code as a basis for experimenting with a variation of KE4IR, or even with a different approach that can be casted in the framework of KE4IR (augmentation of term vectors with semantic terms obtained via knowledge extraction).