Dresden 2011 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

DY: Fachverband Dynamik und Statistische Physik

DY 17: Nonlinear Dynamics I

DY 17.6: Talk

Wednesday, March 16, 2011, 12:00–12:15, ZEU 255

Burst Event and Return Interval Statistics in Wikipedia Access and Edit Data — •Mirko Kämpf¹, Sebastian Tismer¹, Jan W. Kantelhardt¹, and Lev Muchnik² — ¹Institut für Physik, Martin-Luther-Universität Halle-Wittenberg, Halle/Saale, Germany — ²Leonard N. Stern School of Business, New York University, USA

Internet-based social networks often reflect extreme events in nature and society by drastic increases in user activity. We study the dynamics of hourly user access and edit time series for articles in the online encyclopaedia Wikipedia. Daily and weekly activity patterns occur in addition to fluctuations and bursting activity. The bursts are characterised by a power-law distribution of durations of increases and decreases; both sides seem rather uncorrelated. For describing the reoccurrence and clustering of bursts exceeding certain thresholds we investigate the statistics of the return intervals between them. We find stretched exponential distributions of return intervals with identical parameters for all thresholds in access time series, while edit time series yield a simple exponential distribution. We apply detrended fluctuation analysis, finding that most article access time series are characterized by long-term correlations with fluctuation exponents α ≈ 0.9. Finally, we also analyse the cross-correlations between edits and access rates as well as multiple articles inside a subnet. The results help in understanding the complex process of collecting, processing, validating, and distributing information in self-organized social systems.