Prediction of Co-authoring in Academic Social Networks: An Exploratory Study in Computer Science

• Luciano A. Digiampietri University of São Paulo
• Caio R. N. Santiago University of São Paulo
• Caio M. Alves University of São Paulo

Abstract

The identification of future relationships in a social network is a re-levant subject which is receiving, in the last years, attention from both, the scientific community and the industry, because it allows the prediction of the networking behavior or the recommendation of new relationships in order to optimize the network evolution. This paper presents a methodology and the results of the use of attribute selectors, filters, classifiers and functions to predict coauthorships in academic social networks in order to evaluate the performance of different algorithms and to identify some of the most relevant attributes for this prediction.

Keywords: Co-authoring Prediction, Link Prediction, Academic Social Networks