MMOG Player Classification Using Hidden Markov Models

Abstract

Abstract. In this paper, we describe our work on classification of players in Massively Multiplayer Online Games using Hidden Markov Models based on player action sequences. In our previous work, we have discussed a classification approach using a variant of Memory Based Reasoning based on player action frequencies. That approach, however, does not exploit time structures hidden in action sequences of the players. The experimental results given in this paper show that Hidden Markov Models have higher recognition performance than our previous approach, especially for classification of players of different types but having similar action frequencies.

Video game personalisation techniques: A comprehensive survey

“…Other examples, using different video game genres and analysis techniques, are otherwise similar in nature by way of also classifying players into various groups based on their playing style. Matsumoto and Thawonmas [67] classify massively multiplayer online role-playing game (MMORPG) players using Hidden Markov Models (HMM). Tan and Cheng instead present the IMPLANT [68] architecture, which extracts both a POMDP (Partially Observable Markov Decision Process) modelling a player in a tennis game and a fully-observable MDP (Markov Decision Process) of the underlying game environment.…”

Going out of business: Auction house behavior in the Massively Multi-player Online Game Glitch

“…These kinds of analysis can be performed using a great variety of techniques, from basic aggregation of key behavioral features; to complex machine learning approaches [see e.g. 12,13,23,25,39,43].…”

“…purchasing of in-game currency using real-world currency, or vise-versa, has formed the basis for research interest from industry and academia, for the latter notably from the perspective of behavioral economics and sociology. This partly because of the sheer size of these virtual markets, the unique challenges imposed by virtual property rights [7,9,22], social and societal aspects [16] even subversive criminal activity within these worlds, notably gold farming [1,19,25]. However, perhaps more importantly because MMOGs form semi-controlled/contained environments for economic and behavioral research [24,26,43,65].…”

Player Classification Using a Meta-Clustering Approach

“…Matsumoto and Thawonmas [7] explore the use of HMM to classify player action sequences. They exploit the time structures which can be extracted from the action sequences.…”