Automatic Extraction of ICT Competences from Unstructured Sources

Competence management is an important research object in the more general area of human resources management and knowledge management. This paper describes the process of enhancing the individual ICT competence models built from the self-declared knowledge of experts with information extracted automatically from documents using custom Unstructured Information Management Architecture (UIMA) text analysis components. The application aims to identify and extract main expertise areas of ICT experts from two types of documents: published documents (journal / conference papers) and unpublished documents (project documentation). In addition, CV files are used to extract computer skills of different types. The identified ICT competences are exported in such a format that can automatically populate expert profiles built in OWL format. The proposed approach extends the functionalities of existing enterprise information systems and offers possibilities for development of future Internet services.