Data related to the paper "Studying social unrest through the lens of social media"

DOI:10.4121/649e8f5d-8e40-4ab7-9d07-b5ef53d810f0.v2
The DOI displayed above is for this specific version of this dataset, which is currently the latest. Newer versions may be published in the future. For a link that will always point to the latest version, please use
DOI: 10.4121/649e8f5d-8e40-4ab7-9d07-b5ef53d810f0

Datacite citation style

Spierenburg, Lucas; Cats, O. (Oded); van Cranenburgh, Sander (2025): Data related to the paper "Studying social unrest through the lens of social media". Version 2. 4TU.ResearchData. dataset. https://doi.org/10.4121/649e8f5d-8e40-4ab7-9d07-b5ef53d810f0.v2
Other citation styles (APA, Harvard, MLA, Vancouver, Chicago, IEEE) available at Datacite

Dataset

Version 2 - 2025-06-06 (latest)
Version 1 - 2024-12-12
Delft University of Technology logo

Geolocation

France
lat (N): 47
lon (E): 2.5
view on openstreetmap

Time coverage

2023-06-23 19:17:18 to 2023-07-16 02:49:31.353

Licence

CC BY 4.0

Interoperability


Dataset corresponding to the paper "Studying social unrest through the lens of social media".


107,674 geolocated visual posts from a social media were collected during and after the 'Nahel Merzouk' riots in the summer 2023 in 7 French cities. These posts were fed to a computer vision model with the objective of identifying riot-related posts. This dataset contains the metadata (date, time, and location) of those posts along with the label of the posts (according to the model). Riot-related posts are then clustered into "events", based on their spatiotemporal proximity (see paper for more details).


Columns:

"timestamp" (TIMESTAMP): Date and time of the posts

"latitude" (REAL): Latitude at which the post was published

"longitude" (REAL): Longitude at which the post was published

"pred_class" (INTEGER): Binary variable with value 1 if it represents a riot, 0 otherwise

"event" (TEXT): Event associated to the post, structured as follows:

"No event" if the post is not marked as riot-related

"day_city_id" with "day" being the day of the month associated to the event, such as "2", "city" being the city in which the event happened, such as "Paris", "id" being an integer. "29_Marseille_0" corresponds to event "0" happening in Marseille on June 29th 2023. If the value of the id is "-1", the post could not be associated to any event.

History

  • 2024-12-12 first online
  • 2025-06-06 published, posted

Publisher

4TU.ResearchData

Format

sqlite3

Organizations

TU Delft, Faculty of Civil Engineering and Geosciences, Department of Transport and Planning

DATA

Files (2)