Understanding the heavy-tailed dynamics in human behavior

Area of Science:

• Complex systems
• Human behavior analysis
• Computational social science

Background:

• Electronic data sets enable large-scale human behavior studies.
• Interevent times in communication data exhibit heavy-tailed power law dynamics.
• The origin of these power laws is debated: fundamental mechanisms vs. statistical artifacts.

Purpose of the Study:

• To test hypotheses explaining heavy-tailed power law dynamics in human communication.
• To investigate the role of task scheduling, circadian rhythms, and burstiness.
• To develop an improved quantitative model of human interaction mechanisms.

Main Methods:

• Analysis of a large-scale social media communication data set.
• Comparison of observed heavy-tail behavior with models incorporating circadian rhythms and burstiness.
• Development of a novel quantitative model for human behavior.

Main Results:

• Circadian rhythms and burstiness partially explain heavy-tailed behavior.
• A residual, unexplained heavy-tail component suggests a more fundamental cause.
• The developed quantitative model offers improved insights into human interaction mechanisms.

Conclusions:

• Human communication interevent times likely arise from fundamental mechanisms beyond simple statistical aggregation.
• Task scheduling and other intrinsic human factors play a significant role.
• The study provides a new model for understanding the dynamics of human interactions.