Understanding Collective Human Mobility Spatiotemporal Patterns on Weekdays from Taxi Origin-Destination Point Data

Area of Science:

• Geographic Information Science
• Urban Informatics
• Data Science

Background:

• Large-scale geospatial data enables new analyses of human mobility spatiotemporal patterns.
• Understanding urban spatial environments through resident movement is crucial for planning.

Purpose of the Study:

• To develop a classification model for taxi Origin-Destination (OD) point data to identify human mobility patterns.
• To correlate identified mobility patterns with regional functional characteristics.

Main Methods:

• Designed a novel aggregate unit using road intersections for taxi OD data analysis.
• Improved time series similarity measurement with normalization and time windows.
• Applied DBSCAN clustering and random forest modeling to classify patterns and analyze correlations.

Main Results:

• Successfully classified taxi OD data into seven distinct mobility patterns in Nanjing.
• Demonstrated a significant driving effect of regional functions on observed mobility patterns.

Conclusions:

• The developed model effectively classifies human mobility patterns from taxi OD data.
• Findings provide valuable insights for urban planning, traffic management, and land use analysis.