Mapping Activity-Based Segregation of Names in Dublin Using Google Point of Interest Data

The current generation of cities with vast cultures and heritage is influenced by various factors like immigrants from different countries, religious heritage, tourism, and many more factors. Segregation in geographical regions is one of the ways to find patterns in cities influenced by gender, religion, age, income, and many more. In this study, an HDBSCAN-based activity segregation model using Google POI (Point of Interest) is proposed to study the multi-density patterns of reviewers, with possible Indian names, and activities in the Dublin metropolitan area. In this work, the POI dataset is used to study the activity segregation of Indian names in Dublin. This research uses the username to identify the possible gender and nationality of the reviewer using the NamSor app (a machine learning model for prediction of gender and nationality) with an accuracy of 92%. The result shows the proposed HDBSCAN models identify 16 unique segregations which is just nine clusters using the traditional DBSCAN classification model.