Measuring spatio-textual affinities in twitter between two urban metropolises

Area of Science:

• Urban Studies
• Computational Social Science
• Geospatial Analysis

Background:

• Urbanization and social media usage are rapidly increasing.
• Understanding human social embeddedness in urban environments is crucial.
• Social media data offers a novel lens for studying urban life.

Purpose of the Study:

• To define and empirically investigate "spatio-textual affinity" using social media data.
• To develop a framework for analyzing the relationship between location and topical similarity in urban dwellers' social media activity.
• To compare and contrast urban social dynamics in New York City and Los Angeles.

Main Methods:

• Collection of millions of geotagged tweets from New York City and Los Angeles.
• Development of workable definitions for spatial and textual affinity.
• Implementation of experimental designs, visualizations, and measurements for spatio-textual affinity analysis.
• Utilizing Twitter data for empirical study.

Main Results:

• A viable framework for conducting spatio-textual affinity experiments on Twitter data was established.
• Detailed results illustrate the application of the framework for comparing urban areas.
• Demonstrated how location influences the topical content of social media posts in diverse metropolises.

Conclusions:

• Spatio-textual affinity provides a valuable metric for understanding urban social embeddedness.
• The developed framework enables nuanced comparisons of metropolitan areas based on social media activity.
• Leveraging geotagged social media data offers significant opportunities for urban research.