a/SocialMedia · 2 Subscribers
Fake news is all around - really hard to determine what is real and what is fake and remove the fakes quick before they spread.
Tackling fake news is an easy process but it requires a lot of process implementation and development.
This requires internet giants (Google, Facebook, Twitter etc.) and reputable news sources (The Times, Huffington Post etc.) to agree on implementing a synchronised 75% automated and 25% manual verification process via cloud server checking.
Below is a process I created of how this would work:
All articles/digital publications that want to be classified as news must include a 75% automated and 25% manual source verification process.
An article in an article tagges as "news" needs to contain a link (used for verification by crawlers and people) that will send a request to a centralised server where reputable and non-reputable sources will be listed. The server will then reply through API to for automated or e-mail for manual checking, with the source of the news article.
The reply will need to contain the original source of the article in a cookie format for automated verification of a reputable source. This can be a BBC interview with date of airing or a mention that it is available on request from a reputable offline publication.
The manual verification process will go a step further, and check that the source of the article and the article are both real. This will need to happen as soon as possible for articles on publications that are not marked as "reputable sources" that gain popularity.
Please do not confuse this with articles copied from publications. If you copy an article from The Times on your site and mark their site as a source - this does not apply to you - only to the original article. However this should cause downraking in visibility anyway as it's not original content.
The internet giants (Google, Facebook) will need to remove (or not even list in the first place) any articles without a source verification link or coming from an automated un-reputable source or a source that has been marked as "fake" through a manual process. This forms part of ranking algorithms.
Some scenario examples of this:
1 An article gets posted on PropagandaWebsite.com about "War being declared". Does not have a source verification link, does not pass verification via cloud API. Instantly does not register in searches and does not allow sharing on Facebook/Twitter. Instant warning and removal placed by aggregators as news is not trustworthy.
2 An article gets posted on PropagandaWebsite.com about "War being declared". Has a source verification link with a valid reputation mark, and passes verification via cloud API. A "moderator" person (or artificial intelligence algorithm) gets notified of the article gaining momentum from number of shares and manually accesses the verification link and uses the referenced material coming back as a response to check for fact accuracies. The article has a source link that spoofed the BBC, but when accessed via manual verification, comes back with a document scanned by someone in their basement. This article gets removed and marked as fake news with a warning in place when accessed by links already floating online.
3 An article gets posted on PropagandaWebsite.com about "War being declared". Has a source verification link, and passes verification via cloud API. Verification comes back with a mention of a BBC interview with three world leaders declaring war in a press statement - considered a reputable source. The article passes manual verification and gains reputation points. As soon as enough points are gained - the publication will be eligible for API verification with privileges removed if articles are marked as fake by random manual checking.
4 An article gets posted on The Times about "War being declared" - a reputable source. The article has a source link that, when accessed via API or manual verification, comes back with a mention of a BBC interview with 3 world leaders declaring war in a press statement - also considered a reputable source - this article's API verification is sufficient.
5 The article is created as a research report to be used as a source. (Like the article you are reading now). There will be a source link pointing to the report being uploaded to the cloud server. This will have it's own category and be marked for manual review as a research report.
There will always be cases and scenarios where the actual source of news provided to reputable sources can be fake - however - news sources classified as reputable should check their facts before publication.
Each reported incident of news being fake, or attempts to bypass the system, will remove reputation points depending on the size of the publication and the impact of shares. An article with 100 hits and 20 shares will loose 120 points if the article is fake, while an article with 1 million hits and 2000 shares can loose 20 million points if it's determined fake as it is deemed that the news has had more impact.
The above is just a possible process of removing fake news from the internet that I developed.
Let me know what you think of my method in the comments,
PS. Original post https://www.linkedin.com/pulse/method-remove-all-fake-news-from-internet-narcis-radoi dated 26'th Jan 2017