Third Place Urban Analysis
During Harvard’s J-Term, I participated in Namju “NJ” Lee’s course on computational urban analytics. As a case study , we performed an urban analysis of the City of Boston, Massachusetts to understand “third place” distribution. A Logistic Regression model was trained on data collected through the Google Places API to determine the probability that an urban feature would be a “third place”. The model of Boston was then compared with models trained on data of Los Angeles, California and Honolulu, Hawai‘i to contrast each model’s interpretations of Third Place-ness in different urban contexts.
![](https://freight.cargo.site/t/original/i/95aa2ae5692f5dddcc2e5a97591f30f341cd475bde3ce7bae5c8aee092365482/output-boston-color.png)
Plot of Boston Third Place data collected from the Google Places API.
![]()
![](https://freight.cargo.site/t/original/i/08d76c1767662a0d32e28210efcde0e5613ae4097b20ab8a483bcf90e205c46a/Screenshot-2023-12-29-at-01.15.04.png)
Map visualization of Google Places urban data for City of Boston.
Third Place predictions by Place Types in Boston.
![]()
Dominant Third Place by Area in Boston.
![](https://freight.cargo.site/t/original/i/ffadf86b4f3be179ee29f6418a105cdba047f74ea678c919af214cb8837d2dfc/prediction_result_output.png)
![](https://freight.cargo.site/t/original/i/3618e595ac50c4af40906b128bff69de4623c09b499c610e7362603d096c454d/dominant_third_place_in_area.png)
Dominant Third Place by Area in Boston.
(This project documentation is still in progress)