Influential Incremental Learning-Based Privacy Preservation for Social Network Information

Social network influence dissemination focuses on employing a small number of seed sets to generate the most significant possible in social networks and considers forwarding be only technique information transmission, ignoring all other ways. Users, for example, can post message via this mode distribution (called para), which is difficult trace, posing danger privacy leakage. This research tries address aforementioned issues by developing transmission model that supports paranormal relationship. It suggests way disseminating called Local Greedy, aids protection user privacy. Its effect helps reconcile conflict between distribution. Aiming at enumeration problem set selection, an incremental strategy proposed construct reduce time overhead; local subgraph method computing nodes given estimate propagation quickly; group satisfies constraints protection, plan deduce upper limit probability node leakage state, avoiding cost using Monte Carlo crawled Sina Weibo dataset. Experimental verification example analysis are carried out, results show effectiveness method.