Not All Places Are The Same: Computationally Surfacing Regional Experience Differences from Location-Based Services

1 Overview

Working with my advisor Haoqi, I investigated how place imprints experience differently across regions. Drawing on travel guidebooks and HCI theories of "space versus place," I operationalized narrative claims, such as King Cake being routine in Louisiana but "imported" in Pennsylvania, into testable hypotheses. By comparing how concretely and frequently specific experience phrases appear in reviews for the same Yelp category across states, I could hold categories constant to reveal geographic meaning shifts, interpreting results through a framework of environmental, cultural, social, and institutional drivers.

To validate this hypothesis, I conducted a comparative analysis using the Yelp Open Dataset across thirteen U.S. states. I aggregated reviews by (state, category) pairs, extracted n-gram phrases, and computed TF–IDF salience to quantify regional distinctiveness. I iteratively refined extraction thresholds to filter noise while preserving meaningful practice-related language. Crucially, I paired every statistical contrast with ranked review snippets to audit whether detected differences reflected intended experiences rather than incidental word usage.

The analysis demonstrates that review language reliably encodes regional experiential signatures, validating that these differences are recoverable and interpretable from user-generated text. This finding opens pathways for future design: interfaces that expose experiential contrasts alongside ratings, context-aware defaults informed by regional context, and transparent evidence grounding recommendations in actual user descriptions.

2 Selected visuals

Each row shows how, by holding the Yelp category fixed and contrasting review phrases across states, we recover distinct regional experiential signatures interpreted through our four-driver vocabulary.