Deim Seminar
Title
Ranking users in social networks with endorsements
Conferenciant
Hebert Pérez-Rosés
Professor/a organitzador/a
Maria Bras-Amorós
Institution
URV
Date
13-10-2016 12:00
Summary
Online social networks are becoming pervasive in modern society. In particular, professional social networks are a showcase for people's skills, and some sort of job marketplace. We show how to use the tools of graph theory and linear algebra for ranking users in a professional social network with endorsements, such as LinkedIn or ResearchGate. In particular, we show how to deduce new endorsements from existing ones, so as to complete a user's profile, and then apply a centrality or relevance scoring algorithm, such as PageRank, in order to rank users according to some predetermined skill. This ranking can then be used by companies for recruiting personnel, by online associations for weighted voting schemes, etc.
Place
Laboratori 231
Language
Anglès
