Apply social network analysis and data mining to dynamic task synthesis for persistent MMORPG virtual world

Abstract

This paper describes a new scheme for designing MMORPG virtualworld that assigns dynamically synthesized tasks and scripts to avatars in a MMORPG. Different from most current MMORPGs, the new scheme tries to bridge the gap between a real social world with a virtual RPG social world by introducing social relationship, social structure, and avatar personality into the game play. Adaptation of social network analysis to MMORPG virtual world design is also discussed. The objective is to increase a game's long term appeal to players. Moreover, the paper also proposes a method to use data mining technique for intelligently selecting dynamically synthesized tasks so that the tasks will be most likely interesting to the players. The paper represents an endeavor to forge a common research ground for virtual world game design, social network analysis, artificial intelligence, and traditional data mining research. © IFIP International Federation for Information Processing 2004.