Automatic Document Annotation with Data Mining Algorithms

By combining both semantically annotated documents and semantically annotated services, it is possible for digital solutions to automatically retrieve and assign documents not only to their own services but also to those provided by others, thus improving and optimizing the experience of its users. Most of the information exchanged in and between services is still either in paper form or over email and is mostly unstructured and in lack of any form of annotation. Manual and semi-automatic approaches are not suitable to deal with the huge amounts of heterogeneous and constantly flowing data existent in this scenario, thus raising the issue of automatic annotation. In this paper, three data mining algorithms are used to annotate a set of documents and their results compared to manually provided annotations.