Yaqinn

Simone Lorent, Fake News Detection Using Machine Learning,Liege Universite,

Inspiration

Misinformation about Muslims (including refugees, immigrants and ordinary citizens) has been used to construct harmful narratives, reinforce existing Islamophobia and, arguably, create a fertile environment for enacting profound social and political change. Although the role of Islamophobia in the media has been explored in depth, less research has been done into social media, especially the role of social media. These informations are easily manipulated when taken out of context. They can be used to help certain political agendas, heighten divisions in society, and cause actual harm to vulnerable groups. Research suggests that negative attitudes towards Islam and Muslims remain frequent in the social media context. Moreover, media coverage of Muslims seemed to have ‘gained its own momentum’ over time, starting with 9/11 and growing since then.where journalists do not question the dominant narratives of Muslims but simply perpetuate them. In recent years the general public trying to express their opinions about Islam and Muslims on social media has increased. People have also become more likely to share images associated with Muslims without stopping to check their veracity. This has led to the spread a number of ‘fake news’ stories, where Muslims have been linked with certain news events, taken out of context, or used to create entirely fake news. In these cases, it is not merely the mainstream media to blame. Members of the public can easily create and share fake news, contributing to reinforcing negative perceptions of Muslims and perpetuating an ongoing narrative that has the potential to do harm.

What it does

This is what led us to think about an automated way using AI to prevent the spread of this harmful misinformations, which not only affects muslims but contributes to the tremendous rise of crimes and hatred in the world. The Idea mainly is a fake news detection mechanism using machine learning, to help expose false stories on social media and can be extended to textual informations on the internet, In this idea we’ll be exploring the possibility to detect fake news on datasets extracted from the internet by applying traditional machine learning methods and Natural language processing tools (NLP). Also users will be able to upvote and downvote posts (informations) through the mobile app to help the model detect false info.

the following system will be as a first time based on Fake News Detection on Social Media where we can extract data using data mining. After extracting data from social media, we can define a variety of models to check the truthfulness of the content, some of the models are : Expert-oriented: relies on experts, such as journalists or scientists, to assess the news content. Computational-oriented: relies on automatic fact checking, that could be based on external resources such as DBpedia. Propagation-based approaches use features related to sharing such as the number of retweet on twitter.

Stance-based: given a news article and news headline pair. Depending on how similar the news article content and headlines are, the stances between them can be defined as ‘agree’, ‘disagree’, ‘discuss’ or ‘unrelated’.