Blind chance: on potential trust friends query in mobile social networks

Abstract

Potential-Trust-Friends-Query is an important query in mobile social network, as it enables users to discover and interact with others happen to be in their physical vicinity. In our context, we attempt to find top-k mobile users for such query. We propose a novel trust scoring model that encompasses profile similarity, social closeness and interest similarity. Moveover, we devise a current-user-history-record (CUHR) index structure to support dynamic updates and efficient query processing. Based on CUHR index, we propose a query processing algorithm that exploits candidate generation-and-verification framework to answer queries. Extensive experiments was conducted on the real data set to illustrate the efficiency of our methods. © 2013 Springer-Verlag Berlin Heidelberg.