Competence mining for team formation and virtual community recommendation

Currently, one problem faced by the organizations is the inability of the institution to know what it actually knows, in other words, the competences that the organization masters. This problem especially disturbs team formation for a design project, which must be composed of people with specific knowledge to execute different activities. Simple classification or a keyword approach is not sufficient for competence identification. The expertise of an employee or researcher often differs in small but significant details, and an approach is required to allow the accurate description of people´s competences and to implement a retrieval strategy that finds a person with a specific competence. We propose a technique to identify competences by mining employee documents, especially publications. We have chose a number of criteria, which is described in this paper, to find people with specific expertise and propose than to projects. Finally, a possible indication for a person´s participation in a community is her/his published knowledge and expertise degree.