Understanding Spatiotemporal Patterns of Biking Behavior by Analyzing Massive Bike Sharing Data in Chicago

Area of Science:

• Urban Planning
• Transportation Science
• Data Science

Background:

• Bike sharing systems (BSS) are increasingly popular for urban transportation and recreation.
• BSS generate valuable spatiotemporal data for analyzing travel behavior, but large-scale data analysis remains limited.
• Existing studies have not fully explored spatiotemporal patterns or station functionality using flow clustering and over-demand analysis.

Purpose of the Study:

• To investigate spatiotemporal biking patterns in Chicago using massive BSS data.
• To identify distinct travel behaviors of customers and subscribers on weekdays and weekends.
• To examine temporal bike and dock demands and model over-demand patterns.

Main Methods:

• Analysis of Chicago BSS data from July to December 2013 and 2014.
• Construction of a bike flow similarity graph.
• Application of the fastgreedy algorithm for spatial community detection.
• Utilizing hierarchical clustering to analyze temporal demands and over-demand patterns.

Main Results:

• Discovery of unique weekday and weekend biking patterns.
• Identification of different travel trends between customers and subscribers.
• Successful modeling of over-demand patterns for bikes and docks in Chicago.
• Demonstration of increased BSS popularity in Chicago with a 15.9% subscriber increase.

Conclusions:

• The study provides novel insights into biking flow patterns and BSS functionality, overcoming limitations of traditional methods.
• The employed methodologies can be extended to analyze biking patterns and system performance in other cities.
• Understanding these patterns is crucial for optimizing urban mobility and BSS operations.