Time Series Mining in a Psychological Domain

Abstract

Analysis of time series became an inevitable tool in many application areas, such as stock market analysis, process and quality control, observation of natural phenomena, medical treatments, etc. The domain of psychology also offers some interesting applications for time-series analysis, while it is not so frequently applied recently. In this paper, we utilize the system for time-series analysis FAP (developed at Department of Mathematics and Informatics, University of Novi Sad) on the data gained from a specific experimental lab system (so called Socially Augmented Microworld – SAM) developed by informatics and psychologists for Human Factors Research at Humboldt University Berlin. On the basis of experiment log files we extracted three types of time series and generated distance matrices using three kinds of time series similarity measures. Finally, we performed clustering on generated distance matrices and produced dendrograms which serve as the basis for deeper analysis. The outcome of this analysis is two-folded: (a) the most suitable similarity measure can be selected for this domain and (b) these results can serve as a basis for the development of artificial agents which may replace human participants in the experiment.