The content on this page was translated automatically.

Initial situation

Future working life will be increasingly characterized by project work, freelancing and knowledge intensity. This places high demands on dynamic corporate competence and knowledge management. According to Deloitte (2016), HR departments are therefore increasingly opening up to cross-company cloud offerings in the field of people analytics, and the market is set to grow exponentially worldwide by billions. Companies are increasingly turning to technologies that enable them to generate precise expert and project profiles from unstructured text data, e.g., from social networks, internal wikis, or project management systems, which record professional competencies, facilitate the staffing of projects and the identification of contacts, and reveal operational development potential. Up to now, however, skills have been recognized as atomic, technical skills, e.g., a programming language or development methodology. However, important contextual information is lost, e.g. in which industry someone has used a certain technology. 


To address this problem, an algorithm is to be developed within the "SkillExtract" project that is capable of identifying and classifying possible relationships between skills. This should make it possible to recognize the application of specialized skills, for example in certain industries, or the combination of skills, for example "development of an API for Amazon Cloud Services".

Project participants

  • smarTransfer GmbH, Dr. René Wegener
  • University of Kassel, Department of Information Systems 


This project (HA-Project-No.: 628/18-51) is funded within the framework of Hessen ModellProjekte with funds from LOEWE - Landes-Offensive zur Entwicklung Wissenschaftlich-ökonomischer Exzellenz, Förderlinie 3: KMU-Verbundvorhaben.