WMPLP: A model for link prediction in heterogeneous social networks

• Published in: 2014 4th International Symposium ISKO-Maghreb: Concepts and Tools for knowledge Management (ISKO-Maghreb) pp. 1 - 4
• Main Authors: Mohdeb, Djamila, Boubetra, Abdelhak, Charikhi, Mourad
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.11.2014
• Subjects: Computational modeling; Data mining; Educational institutions; Feature extraction; Heterogeneous social networks; Link prediction; Link strength; Predictive models; Random walks; Social network services; Sun
• DOI: 10.1109/ISKO-Maghreb.2014.7033447

Heterogeneous social networks present a great challenge for link mining tasks and especially for link prediction. Recent work insists on the importance of information heterogeneity intelligent exploitation in order to achieve best results. That will be more possible only with the development of new techniques that may represent heterogeneous network in an effective way making it possible to simplify the task of prediction. In this paper, we propose a new method for link prediction WMPLP (Weighted Meta Path-based Link Prediction) which follows a hybrid approach between meta-path prediction methodology and supervised random walks relational methodology.