Utilização de semântica das relações para recomendar colaborações em redes sociais acadêmicas

Abstract

Social network analysis (SNA) has been explored in many contexts with different goals. Here, we use concepts from SNA for recommending collaborations in academic networks. As a recent work shows that research groups with well connected academic networks tend to be more prolific, recommending collaborations is essential for increasing a group's connections, then boosting the group research as a collateral advantage.Therefore, we propose two metrics for recommending new collaborations or intensification of existing ones. Each metric considers a social principle (homophily and proximity) that is relevant within the academic context. Another relevant problem is how to analyze the quality of the resulting recommendations. Hence, we also propose new algorithms for evaluating the recommendations based on social concepts (novelty, diversity and coverage) that have never been used for such a goal. Overall, our experimental evaluation on real datasets shows that using our new metrics improves the quality of the recommendations when compared to the state-of-the-art. Finally, we analyze the properties of the academic networks used in the experimentation. It contributes to understand the results of the recommendations metrics.