Method for providing dynamic decision for social network influence maximization problem

Abstract

The invention discloses a method for providing a dynamic decision for a social network influence maximization problem, and is suitable for the fields of commercial promotion, public opinion control and the like. According to the method disclosed by the invention, dynamic modeling of the social network influence maximization problem is realized. An intelligent dynamic decision meeting a limiting condition k is provided on a dynamic time level. Meanwhile, the possibility of falling into a local optimal result is reduced. The method is not only suitable for the problem of influence maximization in a basic situation, but also capable of realizing dynamic decision in a social network with competitors and maximizing our influence. The invention discloses a method for providing a dynamic decisionfor a social network influence maximization problem based on a reinforcement learning algorithm, which is divided into the following two situations: in the first situation, the existence of other competitors on a social network is not considered, that is, only one product or information needs to be popularized to realize influence maximization; and the second situation is that when a competitor exists in the social network, the influence of the competitor is considered, and a dynamic decision for maximizing the influence of the competitor is given.

Description

technical field [0001] The invention is based on a reinforcement learning algorithm, can provide dynamic decision-making for the social network influence maximization problem, and is applicable to fields such as commercial promotion and public opinion control. Background technique [0002] Acronyms and key term definitions: [0003] Markov decision process: Markov Decision Processes (MDP). [0004] Influence maximization problem: Given a budget that can only satisfy k users, select k seed nodes in the social network to spread, so as to maximize the final influence range. [0005] Infected: A node in a social network is said to be infected if it forwards a message or purchases a product. [0006] Environmental state: For the influence maximization problem, the social network is the execution environment of strategic behavior, so the social network state is the environmental state. The state of all nodes in the social network (whether infected and which information or produ...

AI Technical Summary

Problems solved by technology

Although a series of improved algorithms based on the greedy algorithm has a good range of influence, it is easy to fall into a local optimum, and the time complexity is very high, which is not well applicable when the scale of the social network becomes larger

Although the heuristic algorithm based on centrality can have a low time complexity, the propagation accuracy is not ideal, and its experimental results are usually weaker than the greedy algorithm

Neither the improved algorithm based on the greedy algorithm nor the heuristic algorithm based on node influence ranking has considered the dynamic modeling problem of the social network influence maximization problem.

This makes it impossible to give a dynamic decision to maximize influence based on the dynamically changing social network status, and to give a dynamic optimal response strategy to maximize our influence in a social network with competitors.

Images