Idea: Skhumbuzo Tuswa
Brief: Community Morale
In many ways, fake news spreads the same way the virus does. There are a couple of South African Twitter accounts with a reputation of spreading fake or harmful news. If these accounts engage with a piece of information, we can confidently assume that the information is false, or at least likely to be false. Also, these accounts are collectively followed by hundreds of thousands of South Africans. This means that if any of those people are engaging with information out there, it's likely to be false or inaccurate.
Verifying with 100% accuracy whether news real of fake is time-consuming for humans, but a Twitter bot needs only a few simple rules to quickly estimate the likelihood that a tweet may contain fake news.
Step 1 (Flag The Culprits)
We feed a Twitter bot the above-mentioned usernames (possibly more) and tell it to add them to 'red flag' list.
Step 2 (Flag The Accomplices)
We ask it to load all public profiles that follow those accounts into another list i.e. 'yellow flag'. Although these accounts aren't the main culprits, they need to be flagged as they're at risk of spreading false information.
Engaging The Bot: Step 1: Anytime users see a tweet with information they'd like to verify, they simply @ mention the bot in that same tweet e.g. @FakeNewsBot_ZA.
The bot would then run through all the users engaging with the tweet, and if anyone on the 'red flag' and 'yellow flag' list shows up as having retweeted or liked it, the bot shoots back a message the user's mentions, saying that the tweet is likely fake news and additional verifications are necessary.
To bring this idea to life email us.