Neighborhood activity

data-science / open-data

Understanding how a neighborhood is influenced by human activity can be helpful when it comes to urban design. This study was conducted for building project in Paris.


Based on anonymous phone logs provided by one of our clients and on the surrounding context of Place d’Italie, we made an analysis of the neighborhood’s activity, depending on several types of points of interest, like shops, or restaurants.


Using machine learning, we drafted a prediction engine to visualize what points are attractive and which are repulsive depending on the hour. While this engine gives promising results, it cannot be entirely trusted yet, and would need some refining.