Our project seeks to expose and reduce the prevalence of fake online activities in Europe, such as influencer marketing fraud, phishing scams, and counterfeit product sales. By combining advanced technologies, including large language models and cybersecurity tools, the project will map and monitor fake activity shops operating across European countries. This interdisciplinary effort involves collaboration among social media platforms, regulatory bodies, academics, and the general public. The goal is to offer new tools, methodologies, and policy recommendations to protect online spaces and improve trust in digital ecosystems.
We aim to deliver a platform featuring one year of collected Fake Activity Stores (FAS) data across multiple EU languages, interactive visualization maps, APIs, and analytical benchmarks. It will also provide actionable policy recommendations and educational tools to empower stakeholders. With this, researchers can analyse patterns and trends in fake online activities, policymakers can develop effective regulations using data-driven insights, and developers can build innovative tools to detect and mitigate fraudulent behaviors:
Project Kick-Off (October 2024): A cross-disciplinary team was formed, including experts in technology, law, social sciences, and policymaking.
Building Crawling Infrastructure (February 2025): Multilingual web-crawling tools will be deployed to collect data on FAS, forming the core dataset for the project.
First Review and Initial Data Analysis (June 2025): Early data analysis to reveal the scale, pricing models, and strategies employed by FAS.
Platform Launch and Policy Development (September 2025): A platform with accessible datasets, benchmarks, and APIs will be launched. Drafting of policy recommendations began based on these insights.
Visualisation and Second Review (December 2025): The first interactive visualization map will be developed, offering a clear depiction of FAS activities across Europe to support policymakers and public awareness efforts.
Final Platform and Recommendations (February 2026): A fully functional platform with one year of FAS data, an interactive map, APIs, and policy recommendations.
We developing a data crawling infrastructure based on LLMs:
With Google search engine and LLM filters we discovered around 1800 FAS.
We downloaded FAS websites and currently extracting metrics with LLMs to analyse prices, quality, and other attributes.
We organising the collected data into a OA dataset.
FAMOUS project
6 Month report.
[report - PDF]
April 2-3 Workshop in Luxembourg for the Advisory Board and stakeholders.
[link]
[report - PDF]
FAMOUS workshop on EMIF Winter Conference, December 2024 in Florence, Italy.
[PDF presentation]