الفهرس | Only 14 pages are availabe for public view |
Abstract In recent years, social networks have been the gate for the explosion of thousands of data sets. Interrelations and in{uFB02}uence between social network participants ex- pand dramatically due to the success of online social networks such as Facebook, Foursquare, Twitter and others. The challenge of identifying in{uFB02}uential nodes has been playing a signi{uFB01}cant part in the scienti{uFB01}c community revealing new research opportunities in many application domains such as recommendation systems, viral marketing, and others. Nevertheless, most of the research done considered only the network structure with its geospatial information ignoring the importance of semantics underlying these networks. In this research, we propose a semantic- in{uFB02}uence measurement based algorithm (SIMBA) that ef{uFB01}ciently detects in{uFB02}uen- tial nodes on social networks, as well as estimates the in{uFB02}uence each node has on other connected nodes. SIMBA is based on both the geospatial information as well as semantic information associated with each user. The proposed algorithm is practically used for location recommendation purposes and to examine the power of in{uFB02}uential nodes on the controllability of the recommendation process. More interestingly, we further present a Friend-of-Friend recommendation algorithm to detect the in{uFB02}uence of social ties in the recommendation process |