Howdy, this is the organization page of Debunker-Assistant: our AI-enhanced tool for fake news detection
- Our Mission
- Debunker-Assistant: Community Edition
- De-Escalate
- Debunker Ecosystem
- Contribute
- Follow us
In aequa-tech we believe that fighting misinformation is a crucial step to build better societies. Therefore we developed a tool that supports citizens in the identification and understanding of news that conveys false or otherwise problematic contents. Debunker-Assistant is developed for and with the communities of professionals and organizations who fight misinformation,
Debunker-Assistant is an Open Source and it always will be. We and weLaika maintain 3 repositories that represents the building blocks of our system
- Debunker-Assistant APIs: a set of Natural Language Processing and Network Analysis methods for the multidimensional analysis of the quality of news (that will soon become a library)
- The Web Application: an application that allows users to create an account and have access to our system through a graphical interface
- The Annotation Platform: a module that allows users to provide feedback and annotations about misinformative contents.
Non-experts can also access to our technology through our graphical interface
Debunker-Assistant is now in beta but our journey is just begun. Our next objective is to develop De-Escalate a technology that is able to (i) identify news that provokes polarization among citizens and (ii) help them understand opposite perspectives in order to foster mutual dialogue between polarized communities.
Debunker-Assistant is part of a network of partners and stakeholders that are essential to fulfill our mission. We would never been able to develop the first version of our technology without the NGI-Search Initiative who funded our project with 150,000 euros and supported our work with its group of experts in Open Source, Business Modelling, Open Data, etc. We take care of the AI stuff, weLaika takes care of the rest of the service. Futura is our legal partner: they are not only the ones that wrote our terms and conditions but also a crucial support to design a product that embraces the principles of the EU market. Our product would not exist without a community of activists and NGOs. During our journey we met activists from Amnesty International and Hate Trackers who help us trace the path of our techology.
Defining the contribution rules for Debunker-Assistant is something that we are taking seriously. That's why they are not ready, yet. In the meantime please meet the people who collaborate and collaborated with us in these months:
The following master students had an internship in our company and contributed to the development of Debunker-Assistant
- Arianna Longo collaborated to the creation of a corpus annotated for the detection of vulnerable identities in news headlines;
- Vittoria Tonini worked on a corpus of counternarratives against hate speech in viral news shared on social media.
Friends of aequa-tech is a program aimed at PhD students and post-docs that grants a scholarships to carry on Research & Development activities related to Debunker-Assistant
- Silvia Casola implemented a module for the summarization of news contents.
- Marta Marchiori Manerba developed an explanation module of our prediction.
- Soda Marem Lo is developing a computational resource to detect different moral stances about the same news.
From December you can follow Debunker-Assistant's adventures by subscribing to aequa-tech's Substack