Systematic and reliable evaluation of social web data for urban studies

As of today, more than 50% of the world’s population lives in cities. By 2050, about 70% of the people living on the planet will identify themselves as city dwellers. Yet, in Europe we have already reached this percentage.

But, what do we really know about our cities, their problems, their trends, and their citizens? Can social media offer a better understanding of our urban environment? How can citizens become active participant in the improvement of their city’s well-being?

SocialGlass aims at enhancing our ability to sense, interpret, and understand city environments.

The project involves researchers from TUDelft, including:

  • Dr. Alessandro Bozzon
  • Dr. Birna van Riemsdijk
  • Dr. Stefano Bocconi
  • Dr. Achilleas Psyllidis
  • Christiaan Titos Bolivar


Social Data generated by users in online social networks recently attracted considerable attention as a potential source of knowledge to better understand urban environments, citizens, and related (natural) phenomena. Many examples hinted at a successful usage of social data to gain knowledge about specific cities, or for specific purposes. However, as of today, there is no general theory able to claim a comprehensive effectiveness of social data for predictive or descriptive purposes: intuitively, solutions that might work for a given city might not be as valuable for another one.

Before adoption and consideration by scientists and decision makers, social data must be subject to meticulous study that qualifies their ability to represent or describe the targeted reality. Typical questions are: do data possess enough (temporal and spatial) resolution to represent the studied phenomenon? Are there biases (of cultural or technological nature) that might influence the conclusions drawn from such data? How does the world depicted by social data compare with common or consolidate knowledge (e.g. from cadaster or municipal databases)?

Such an evaluation requires methods and tools capable of enabling consistent and comparable assessments of social data across usage scenarios (e.g. city, goal), and according to state-of-the-art qualitative metrics.


SocialGlass is a framework aimed at simplifying the process of social data investigation by facilitating the analysis, visualization, and value assessment of social data with respect to a reference reality. An important part of this process consists in integrating and cross-correlating social data with multiple data sources, such as sensor data and mobile phone data.

Thanks to a powerful and extensible palette of easy-to-use and easy-to-configure tools, SocialGlass aims at smoothening the learning curve for the adoption of advanced social data evaluation. The goal of this project is to foster the adoption of such a valuable source of information by guaranteeing the extraction of correct and actionable knowledge.

We aim at developing SocialGlass further by creating a demonstrator that showcases the need for accurate social data evaluation. The demonstrator will analyze data drawn by multiple social networks, and for many cities (e.g. Amsterdam, Rotterdam, Paris, London).

External website: