ISSN 2330-717X

Identifying Fake News In Social Media: AI And ML – OpEd

By

In addition to the growth of online usage, the advent and universal acceptance of the idea of social media has transformed the way content is generated and published. News has become quicker, less expensive, modified, and readily available via social media. This motion has also come together with certain drawbacks.

The fake news concerns have become an important research focus due to the high traffic on social media. It is convenient for people to write false comments and misinterpreted information on social media with eye-catching highlights for entertainment purposes without realizing the consequences. The main challenge is to determine the difference between true and false news. However, it is necessary to understand how the techniques in the fields of machine learning will help us to detect fake news.

 Misinformation often involves information that is manipulative or otherwise damaging. Misinformation is becoming a global problem as fake news not only damage democracy but also public trust.  False social media news identification has particular features and poses new problems. Firstly, fake news is deliberately published to mislead viewers to accept misleading facts, making it impossible to detect news-based content.

We also need to use auxiliary information such as social media user commitments, to help separate it from real news. Secondly, the manipulation of this auxiliary information is non-trivial in itself since users’ social contributions to false news generate data that is massive, unreliable, and unstructured. Fake social media news has its own special features. For example, malicious accounts can be easily and rapidly generated to spread fake news, such as social bots, cyber users, or trolls. In comparison, people are selectively introduced to some forms of news as a result of the way news feeds appear on their social media. 

Spreading of fake news is not a new phenomenon but because of the technological advancement false news occurs more and more regularly. Propagation of fake news is now easier than ever to produce thanks to a wide variety of free mediums. The platform is so advanced that even the slightest or minor information spreads like fire in the bushes. For decades, deception has existed, but social media proved to be the new medium of contact to be exploited and spread disinformation or even rumors. False news can become inescapable on social media. Fake news also makes usage of fake identities of writers and false outlets. If the website has a habit of producing deceptive statements, or specifics may not appear reliable in the author’s profile (or even a description is non-existent), the material should be viewed with intense scrutiny.

Over the pandemic situation, propaganda about the latest coronavirus vaccinations has accelerated on social media, including those of viral anti-vaccine messages spread across numerous channels and also by ideological parties. The conspiracy news related to Covid-19 vaccines soon to be eradicated by social media website. Authorities which involve in spreading the fake news related to vaccinations being handled. The widespread of this fake news induced fear among individuals. Not just this but other fake news that promotes racism, cultural demolition, gender biases; more or less give rise to harassment and abuse not just sabotaging social media platforms but also ruining the image of being the best source of information. 

Machine learning may have achieved immense popularity, but it is only one way to achieve artificial intelligence. At the birth of the AI, AI was described as any machine capable of performing a task that would usually require human intelligence. In today’s world, machine language (ML) is being used to counter fake news.  Fabula AI is a London-based startup that leverages ML to detect the spread of fake news online. Twitter has acquired Fabula AI. Fabula has developed a patented AI system called “geometric deep learning” an algorithm that learns from large and complex data gathered from social networks. Fake story spreads faster than the real story itself.

Fabula focuses on detecting the patterns of differences in how content is spreading on social media and allocating an authenticity score. Fabula said “As this technology detects the spread pattern, it is language and locale independent; in fact, it can be used even when the content is encrypted.”

If we talk about the most used social media app, Facebook is already using machine language algorithm to ensure that images and videos that contain brutality and pornographic content are flagged and deleted as soon as possible. With over 2 billion users it is impossible for any individual to check and monitor content on a daily basis but, the company is all set to use the same technology to detect fake news on a massive level.  Before that, Facebook from the help of an agency used to check and label content which lessened the fake news  by 80%.

Tessa Lyons, the Director of product Management at Facebook said in a blog post, “Facebook has already been tackling the spread of false news across its network for the last year and a half, and now they are upping the ante, introducing a series of updates designed to fight fake news. This comprises a combination of technology and human review, including removal of fake accounts, partnership with fact-checkers, and promoting news literacy.” 

Researcher identifies linguistic features for the identification of false news using machine language and natural language processing technologies. Fake news articles use more terms that are used in hate speech, as well as words relating to sex, death and fear. On the other side, genuine news comprises a greater proportion of terms related to work (business) and money (economy). This indicates that a stylistic approach combined with machine learning could be useful for detecting suspicious news. 

Now it depends to what extent governments, corporations, and customers are now realizing the fact that how extreme the distortion of public perception can indeed be when articulated to “fake news.” Government of Pakistan also starting to believe that false news are one that needs to be vigorously battled against. Respective federal departments are already establishing programs that several believe to be fraudulent to discredit news. They still are currently applying rules and prosecuting websites that release disinformation. Government has launched a Twitter account namely “Fake News Buster” to expose misleading and bogus content on social media. This website is lunged to timely catch the “false propaganda” or “Misinformation” on social media sites, to eradicate its spread as well as to avoid further misfortunes. 

However, in thought logically, one can prevent further progression by having a good sense of interest in what you read on your page, consider about all viewers do gets curate by social media outlets, and also utilize investigation approaches. Once done properly, social media can be turned into a valuable platform for both corporations and people.

*Daniyal Talat has a Masters in Defense & Strategic Studies from Quaid-i-Azam University, Islamabad and is currently working as a Program Coordinator for the School of Inclusion.

Click here to have Eurasia Review's newsletter delivered via RSS, as an email newsletter, via mobile or on your personal news page.

One thought on “Identifying Fake News In Social Media: AI And ML – OpEd

  • February 2, 2021 at 11:24 am
    Permalink

    Very well summarized.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *