Recommending knowledge in a knowledge based social network1

The organizations aim to increase its competitiveness. In this context, they have been searching for new ways to improve their productivity, the quality of their products, and cost reduction. To achieve these goals, it is essential to use the collaborators’ potentials and the relationship among them to find and share tacit knowledge. Since tacit knowledge is stored in people’s mind, it is hard to be formalized and documented. Facing this difficulty, identifying and recommending persons who retain the needed knowledge might be a good option. This work presents the Specialist Recommender System (SWEETS) and its application into the a.m.i.g.o.s. environment, a social network platform for knowledge management. The SWEETS system uses folksonomy to extract a lightweight ontology, which is essential to effectively identify people’s skills. This lightweight ontology is based by tags (concepts) relating them to items (instances), and its co-occurrences. In addition, such ontology is domain independent, which is a contribution of this work. Applying the SWEETS system into the a.m.i.g.o.s. environment we are looking for minimizing the communication problem in the corporation, providing an improvement on knowledge sharing. Therefore, a better usage of the collaborators knowledge may be expected.