Modelling Trust-based Evolutionary Dynamics in Signed Social Networks

Projektbeschreibung

Signed Social Networks unter der Lupe

Die rasche Zunahme der Zahl der Nutzer auf den Seiten der sozialen Netzwerke hat ein Wissenschaftlerteam dazu veranlasst, die Vertrauenswürdigkeit von Signed Social Networks (SSN, mit positiven und negativen Zeichen gekennzeichnete soziale Netzwerke) zu prüfen. Die Forschergruppe untersucht Wege, über die ein unmarkiertes Netzwerk von Bekanntschaften in ein als vertrauenswürdig/nicht vertrauenswürdig markiertes Netzwerk umgewandelt werden kann. Eine Reihe vorhandener Netzwerke wie etwa Epinions haben begonnen, Links explizit entweder als Freund/Feind oder mit dem Hinweis Vertrauen/Misstrauen zu kennzeichnen. Das EU-finanzierte Projekt TEAMS entwickelt Mechanismen zur Erkennung der vertrauenswürdigen Communitys und Erfassung ihrer Entwicklungsdynamiken. Es erforscht die Darstellungsmechanismen für Signed Social Networks und nutzt dabei die formale Begriffsanalyse zur Entwicklung eines Modells.

Ziel

Users’ experience with real-world social systems (e.g. Epinions and eBay) witnesses the importance of Signed Social Networks (SSNs) that have wide practical and valuable applications in social media such as opinion guidance, personalized recommendation, and topic identification. However, the diversity of massive social interactions complicates the trust and distrust relations among users in SSNs. In particular, the complexity of distrust relations leads to significant challenges in detecting the trusted communities and capturing their evolutionary patterns.

This research aims to pioneer the innovative mechanisms for detecting the trusted communities and learning the evolutionary dynamics. To this end, we will explore the representation mechanism for SSNs by using the Formal Concept Analysis (FCA) and develop a FCA-based representation model. Next, the mechanisms and corresponding algorithms for detecting trusted communities and identifying their dynamic evolutions will be investigated. This research will provide both theoretical fundamentals and practical techniques for detection and dynamic evolution of trusted communities in SSNs. Moreover, this project can stimulate new research directions and the collaborative opportunities across multiple disciplines, such as social computing, soft computing and networking.

To broaden the fellow’s knowledge horizon, a series of research, training, and knowledge transfer activities are planned. The new knowledge and skills imparted in these activities will further promote the applicant’s research portfolio and significantly enhance his career prosperity. The research will also lay a solid foundation for the long-term and wide-range collaborations between the fellow and the host university, and eventually lead to more extensive and higher impact of research results, from which both EU and China will benefit.