Visualization of Online-Game Players Based on Their Action Behaviors

摘要

We propose a visualization approach for analyzingplayers' action behaviors. The proposed approachconsists of two visualization techniques: classicalmultidimensional scaling (CMDS) and KeyGraph.CMDS is for discovering clusters of players who behavesimilarly. KeyGraph is for interpreting actionbehaviors of players in a cluster of interest. In orderto reduce the dimension of matrices used in computationof the CMDS input, we exploit a time-series reductiontechnique recently proposed by us. Our visualizationapproach is evaluated using log of an onlinegame where three-player types according to Bartle'staxonomy are found, that is, achievers, explorers, and socializers.