DEMOIR: a hybrid architecture for expertise modeling and recommender systems

Abstract

Although employees' expertise has long been regarded as an important asset in organizations at least on a par with capital, goods and documented information, it is only recently that automated systems to enhance the visibility and traceability of employees with particular expertise started to receive research attention. In this paper, we introduce a systematic classification of alternative architectures for a core functionality of expert finding systems, namely expertise modeling. We also analyze the advantages and disadvantages of each approach and present our hybrid solution. We discuss how this solution addresses the requirements of (I) distributed and heterogeneous organizational and personal expertise data sources and, (2) centralized access to extracted expertise information.