2019–2026

Projects.

Mobility networks, dynamic segregation, occupational text analysis, and spatial measurement.

01Mobility networks · racial segregation

Do racial patterns correspond with how people move?

Problem: Residential segregation tells us where groups live, but not whether the same racial-demographic divisions also organize everyday travel.
Result: Composite alignment was statistically positive in 54 of 100 regions; Black-gradient alignment was positive in 51.

02Time + exposure

Problem: Residential measures fix people at home, even though daily movement changes who shares urban space.
Result: Segregation was lower during weekday working hours; in 13 of 18 regions, even the highest estimate remained below residential segregation.

03Natural language processing · machine learning

Problem: Surveys must turn brief, uneven descriptions of people's work into standard occupation codes, but the text may not distinguish occupations reliably.
Result: The model ranked the human-assigned code first 59.7% of the time and among its first two choices 74.7% of the time, with large differences across occupations.

04Natural language processing · measurement

Problem: A one-word job title and a detailed account do not provide the same evidence, even when both describe similar work.
Result: In 60 synthetic pairs, detailed aligned descriptions scored above both similarly detailed descriptions of different work and short aligned descriptions.

05Computer-vision logic · spatial measurement

Problem: Citywide segregation indices summarize unevenness but do not locate where two specific racial populations meet. A steep change in only one group can also resemble a boundary.
Result: The Sharp Boundary Index locates a pairwise racial interface, conveying different information than a citywide segregation measure.

Chicago map

The Chicago map places a local, pairwise racial-boundary measure beside communities derived from observed travel.