On the effectiveness of random jumps in an influence maximization algorithm for unknown graphs

In our previous work, we have proposed a problem called influence maximization for unknown graphs, and a heuristic algorithm for solving the problem. Influence maximization for unknown graphs is the problem of detecting a set of influential nodes in a social network using only the limited knowledge about its topological structure that is obtained from a limited amount of probing. In this paper, we extend the method of probing used in our previous algorithm, and propose an influence maximization algorithm for unknown graphs, which we call IMUG-RF (Influence Maximization for Unknown Graphs with Random Flight). IMUG-RF greedily probes the node with the highest expected degree and also periodically probes a randomly selected node in the network. Through simulation, we show that IMUG-RF can successfully identify influential seed nodes in real social networks.