Goals and Objectives
The main aim of the MuseAI project is to increase the efficiency of fact-checking operations and scaleup fact-checking tasks associated with monitoring the web and social networks to detect fact-checkable claims, as well as repetitions and spreading, even with certain variations, of a claim.
To this end, the proposed project will develop a multilingual claim detection and matching solution for fact-checkers and journalists that will allow them:
- to increase the volume of claims they can detect online;
- to detect similar and identical claims circulating online, also in different languages, within a given content pool;
- to operate with different information modalities (text, audio, video) through automated transcription using deep learning models.
This overarching aim can be broken down to several measurable objectives:
- Develop a new method to identify factual claims in a text;
- Develop a new multilingual AI-based method to assess the check-worthiness of a claim, based on a set of indicators for check-worthiness;
- Develop a new method to identify similar claims or different phrasings of the same claim;
- Make the developed method accessible through an intuitive and user friendly interface;
- Assess and evaluate the developed solution in terms of accuracy, performance, and usability.