A Novel Approach for Identification of Users with Access Restrictions

Information in social networks can be useful in different domains. Target Oriented Network Intelligence Collection (TONIC) problem is the problem of finding profiles that have information about given target in social networks. This problem is formalized as a heuristic search problem and is represented by different heuristics in order to decide about selection of profiles for searching in graph. We introduce new heuristic based on topology of graph in known parts of the graph that appears in progress. This heuristic is composed by Bayesian promising factor and distance of given target from other profiles. The efficiency of this heuristic is evaluated on Facebook and Enron. The new heuristic outperforms all other previous heuristics.