LAUSR

Abstract In recent years, with the development of social app, the Internet has created a different interactive mode from the previous daily life, which makes the development of society gradually presents the trend of “gamification”. “Social gamification” makes the life systematically integrated into the competition activities. Therefore, we propose a new idea for the realization of a sustainable society under the trend of “social gamification” by combining strategy game ecology mining for the first time. Specifically, we use big data mining to analyze the attributes of the smurf (subsidiary account), and then extract discriminative features and compute a score for each attribute using machine learning, which could accurately locate all the smurf in the game. By banning the smurf, the game ecology is optimized to improve the player’s game experience. Meanwhile, this paper proposes appropriate operation strategies for players at different stages based on multi-level clustering. In the real society, smurf represents the citizens who disturb social order, and different types of players represent the pyramid structure of social classes. The measures in the game provide references for the construction of a sustainable society.