In this chapter we present a flexible and configurable application for mining large XML document collections. This work is centered on the process of extracting document features related to structure and content. From this process, an attribute frequency matrix is generated and, depending on the cluster algorithm, it is transformed and/or used to obtain similarity measures.