Multi-location Influence Maximization in Location-Based Social Networks

With the development of location-based social networks (LBSNs), location property has been gradually integrated into the influence maximization problem, the key point of which is to bring the users in social networks (online phase) to the product locations for consuming in the real world (offline phase). However, the existing studies considered that a company dependent on the viral marketing only has a product location in the real world and could not suit the situation that there is more than one product location. In this paper, first, we propose a new propagation model, called multiple factors propagation (MFP) model which can work in the situation that there are multiple product locations in the real world. Meanwhile, the definition of multi-location influence maximization (MLIM) problem is presented. Then, we design a hybrid index structure to improve the search efficiency of offline phase, called hybrid inverted R-tree (HIR-tree). Furthermore, we propose the enhanced greedy algorithm for solving MLIM problem. Finally, we conduct a set of experiments to demonstrate the effectiveness and efficiency of enhanced greedy algorithm.