This project was completed during the Lviv Data Science Summer School 2016. The project supervisor – Dmytro Karamshuk.
Geo-tagged social media datasets have recently emerged as an unprecedentedly rich source of information on the dynamics of modern cities. By providing fine-grain details about movements of million urban dwellers they can help us to understand the patterns of urban human mobility, reveal hidden patterns of spatial interactions between business activities, assist in understanding retail quality of locations and even help to plan better cities. In this project, students exploit a dataset of check-ins from Foursquare – a location-based social network – to analyze the real estate market in large cities. The goal was to use the information about geographic location of points of interests (POIs), the self-reported locations of Foursquare users and other open source datasets to devise predictors of real estate prices in a city. Several machine learning approaches to tackle this problem and exploit the land registry data for the city of London as a case study were considered during the project work.
If you are interested in more details, please contact email@example.com