Towards big data behavioral analysis: rethinking GPS trajectory mining approaches from geographic, semantic, and...

Area of Science:

• Architecture
• Urban Planning
• Behavioral Science

Background:

• Understanding built environment usage is crucial for behavioral research.
• Traditional data collection methods for human movement lack accuracy and granularity.
• Emerging GPS trajectory mining offers potential for more detailed analysis.

Purpose of the Study:

• To review and summarize the applicability of GPS trajectory mining in architecture.
• To examine the usefulness and limitations of geographic, semantic, and quantitative GPS approaches.
• To investigate human behavioral patterns in built environments using real-world data.

Main Methods:

• Review of GPS trajectory mining approaches (geographic, semantic, quantitative).
• Case study using GPS trajectory data from visitors at the Palace Museum, China.
• Three experiments to assess the utility and weaknesses of different trajectory mining dimensions.

Main Results:

• All three dimensions of GPS trajectory mining offer valuable insights for architectural and urban design.
• Increased data dimensionality in trajectory analysis enhances the discovery of generalizable human behavioral patterns.
• GPS data alone has limitations for understanding typological human behaviors; integration with surveys is recommended.

Conclusions:

• GPS trajectory mining is a promising tool for architectural and urban design, especially with higher data dimensionality.
• Combining GPS data with traditional methods like surveys and questionnaires provides a more comprehensive understanding of human behavior in built environments.
• Future research should focus on integrating diverse data sources for richer behavioral insights.