Projects
Contact
Identifying Analogous and Unique City Districts Using Wikipedia Based Semantic Representations
Cities across the world often have districts that serve similar purposes or hold similar identities, for example, areas known for shopping, nightlife, pop culture, or historic character. While previous studies have compared cities using structured data such as points of interest or mobility patterns, these approaches often fail to capture the atmosphere or cultural meaning that people associate with a place.
This project explores a different approach by analyzing how city districts are described in Wikipedia articles. Using English-language Wikipedia as a common text source, we aim to extract and compare the semantic content of descriptions across different cities. Through this, we hope to identify districts that are semantically similar such as asking, “What is the Akihabara of Seoul?” as well as neighborhoods that appear truly unique, with no close counterpart elsewhere.
This work is intended to serve as the basis for a future conference paper or publication, depending on progress and interest.
2-3 students
6 months
The project applies basic natural language processing techniques using only open and non-sensitive data. It is designed to be accessible to students from a variety of backgrounds and suitable for international collaboration.