Abstract (EN)

Nowadays, more and more collaborative tools are available to support users’ remote collaborations. Its increasing amount makes users struggle in managing and retrieving information about their collaborators during collaboration. To solve this problem, many decision support systems have been developed quickly, such as recommender systems and context-aware recommender systems. However, the performances of different algorithms in such systems are relatively unexplored. Based on our three proposed context-aware collaborator recommendation algorithms (i.e., PreF1, PoF1, and PoF2), we are interested in analyzing and evaluating their performances in terms of accuracy and time efficiency. The three algorithms all process the context of collaboration by means of ontology-based semantic similarity, but employ the similarity following two approaches respectively, to generate context-aware collaborator recommendations. In this paper, we present how to test, analyze, and evaluate the performances of the three context-aware collaborator algorithms in terms of accuracy and time efficiency.