Information Network Cascading and Network Re-construction with Bounded Rational User Behaviors

Social media platforms have become increasingly used for both socialization and information diffusion. For example, commercial users can improve their profits by expanding their social media connections to new users. In order to optimize an information provider’s network connections, this paper establishes a mathematical model to simulate the behaviours of other users to build connections within the information provider’s network. The behaviours include information reposting and following/unfollowing other users. We apply the linear threshold propagation model to determine the reposting actions. In addition, the following or unfollowing actions are modeled by the boundedly rational user equilibrium (BRUE). A three-level optimization model is proposed to maximize total number of connections, which is the goal of the top level. The second level is to simulate user behaviours under BRUE. The third or bottom level is to maximize the other users’ utility used in the second level. This paper solves this problem by using exact algorithms for a small-scale synthetic network.