The majority of online media companies have revenue driven by user traffic and as such are pressured into producing sensationalist headlines that grab user attention. The resultant market has been saturated with articles that fall short on the sensationalism advertised, leaving users dissatisfied or misinformed.
Facebook has issued 40 million warnings that posts on its site may be misleading. The social detriment of this can be seen in the current pandemic. A BBC team tracking corona virus misinformation has found links to arson, assault and death.
What it does
Symany is a Google Chrome Extension that warnings over fake news and clickbait posts in Facebook. We make sure that any external links on Facebook have been tested against our metrics bringing you the most reliable news source - keeping you up to date on the most recent real news.
How we built it
- The Python Flask API utilises Machine Learning to categorise headlines as either clickbait or not.
- Comparisons against a well-update blacklist of known fake-news sites are made.
The classifier continually scans posts on Facebook and asks the API if they're fake or clickbait. If they are, it displays an overlay.
Challenges we ran into
Time was the biggest challenge that we ran into. We found that our initial vision was too big and so had to dial it back given that there was only 40 hours to work on it. We found that working on a simple prototype and building upon that allowed us to combat this issue as we would end up with a product that worked if we ran out of time.
"Shoot for the moon. Even if you miss it you will land among the stars."
Accomplishments that we're proud of
Throughout this tight 40 hour time crunch, we are all proud of producing a functional Fake News/Click Bait Google Chrome Extension Detector that has the capability to filter out malicious news article swarming Facebook which have been causing harm to users globally. We are all especially proud of keeping our heads down, working through the 40 hours and never giving up on building a product that will one day help not only our friends and family but the wider community.
What we learned
What's next for symany
- Introduce more features - provide a function that suggests better sources articles on the same topic
- Expand to other social media (Twitter, Instagram) and build other explorer extensions
- Add further classification metrics (scoring system)