Skip to Main content Skip to Navigation

Contribution aux mécanismes d'automatisation du fact-checking pour un journalisme augmenté

Abstract : This thesis deals with the questions related to automation of the fact-checking in the journalistic web context. It consists, for the media and the professionals of information, of checking the veracity of facts in the media sphere. In other words, this is to authenticate the accuracy of the statements. An investigation of the state of the art shows that it was more conerned with manual practice which is restricted regarding the velocity, the volume and the variety of processed data on internet. Hence, thenecessity to have new methods and automated tools in order to verify the facts at due time. This research analyses how the fact-checking, as a journalistic genre, is automatable. To put it in another, we seek to focus on the algorithms and the agents of automated fact-checking but also, to propose new orientations for the evolution of the journalistic profession. Five contributions are presented in the thesis. The first contribution that is presented is an automatic collection (web scraping) of articles and factuals claims from the online press (FactExtract). The second contribution is an automatic tagger of facts in French-language speeches (FactTagger). A synthetic data set generator, customizable and close to the real journalistic environment (JFact) and two algorithms of factchecking (SnVera and SenFact) are available.
Complete list of metadatas

Cited literature [182 references]  Display  Hide  Download
Contributor : Edouard Ngor Sarr <>
Submitted on : Monday, September 14, 2020 - 8:42:28 PM
Last modification on : Wednesday, September 16, 2020 - 8:28:07 AM


Files produced by the author(s)


  • HAL Id : tel-02938383, version 1


Edouard Ngor Sarr. Contribution aux mécanismes d'automatisation du fact-checking pour un journalisme augmenté. Web. Université de Thiès (Sénégal), 2019. Français. ⟨tel-02938383⟩



Record views


Files downloads