With modern-day cities come vast quantities of data from online social activity. These data reflect how people interact with the urban environment and with each other. The more knowledge we can derive from social urban data, the better we can address real-life urban and regional challenges. The Social Urban Data Lab (SUDL) employs complex datasets from diverse sources to discover knowledge about urban interactions. In paving the way to more intelligent cities, SUDL develops state-of-the-art technology for the acquisition, integration, enrichment, analysis, and visualization of urban big data, making their monitoring and interpretation easier for data-driven urban analytics, planning, decision-making, transport and social science research.
In the context of contemporary cities and metropolitan regions, Social Data comprise a precious and untapped source of information about social, temporal, and social aspects of the activities, movement, and social connectivity of people.
Social Data are produced in large amounts by emerging sources such as social media, sensors, GPS devices, LoRa networks, mobile phones, and open data portals. Together, they provide an exciting reflection of the human landscape.
Besides the various opportunities of big urban social data, their large volume and “human-oriented” nature also bring unprecedented challenges, relating to their inherent heterogeneity, lack of structure, ambiguity and bias.
In SUDL we develop novel data science methods and technologies for knowledge discovery about urban interactions through big geo-social data.
Urban interactions concern interactions between people in and across cities, and between people and element of the urban or regional contexts. Think of human mobility and migration, the flow of goods, energy or information, and social relationships.
Our goal is to create powerful digital reflections of urban interactions at scale from dynamic, sparse, and ambiguous data sources, by combining expertise from Data Science, GeoComputation, Machine Learning, Applied Spatial Analysis and Human Computation.
Enabling Technology — The SocialGlass system
In achieving the above-mentioned goals, we have developed SocialGlass – an enabling technology that implements the various SUDL methods.
SocialGlass is an online system for real-time urban big data analytics and forecasting. It capitalizes on advanced Semantic Web, Machine Learning, Spatial Data Mining, and Human Computation technologies. Its pioneering toolbox helps governments, planners, entrepreneurs, and wider communities plan the future of cities in a truly data-driven fashion.
SocialGlass facilitates the combination of insights derived from sources as diverse as social media, sensors, GPS devices, LoRa networks, and open data portals. It further provides mechanisms for the effortless incoroporation of custom data sources, including crowdsourcing campaings. SocialGlass harnesses the power of urban big data to help city stakeholders better understand and forecast the dynamics of cities.
Our system has been taken up by the Municipality of Amsterdam during SAIL 2015 for enhanced crowd-management solutions. We are currently working on real-life applications relating to the prediction of appropriate retail locations, the understanding of energy consumption behavior, and the policy-making for vulnerable societal groups.
With SocialGlass, we envision a new generation of real-time Urban Computing systems that are able to scale over large amounts of geo-social data, while maintaining the quality of human understanding in the loop.
SocialGlass can be a game-changer solution for Urban Data Science and GeoComputation. Our ambition is to offer SocialGlass as a service and, thereby, extend its use to a range of municipalities, businesses, and wider communities.
Delft University of Technology (TU Delft)
City of Amsterdam
Centre for BOLD Cities
Fondazione Bruno Kessler
La Fabrique de la Cité
Dr. Alessandro Bozzon, Assistant Professor, Delft University of Technology | AMS Institute, contact
Dr. Achilleas Psyllidis, Postdoctoral Researcher, Delft University of Technology | AMS Institute, contact
Related information For more information, please visit the SocialGlass project website.
In the news
Article in De Ingenieur on SocialGlass August, 2017 (In Dutch)
Interview with Alessandro Bozon in Het Parool, “Met mobile data zien of de stemming omslaat“, August 20 2017 (In Dutch).