CrossOver Monitoring maps the algorithmic recommender systems of online platforms, search engines and marketplaces. It helps users track how these recommender systems contribute to disinformation and other forms of information manipulation. This data is visualised via easy-to-use dashboards on a dedicated website. CrossOver is currently active in about 15 countries throughout Europe and Africa.
CrossOver emulates non logged-in users browsing several social media platforms and search engines, when searching for specific keywords. Users can be emulated from across different locations in the world (cities, regions, countries), enabling geographical granularity on the data collection. We currently emulate user experiences’ on Tiktok, Youtube, Bluesky, Amazon, Google search and Bing. The overall aim of the project is to understand what type of content they are shown by VLOP’s and VLOSE’s recommender systems.
The dataset comes from platforms’ websites, not apps, and emulates a user who is not logged in. The dashboard provides a visualization of data that is retrieved through the data collection using Crossover in order for users to analyse them -with the guidance of CheckFirst if needed.
This is done by using a blend of user-simulation and data via official APIs or programmatic endpoints.
CheckFirst empowers its partners to navigate today’s complex online information ecosystem and equips decisionmakers with customisable solutions to combat online harms.
Our story