Inspiration
https://github.com/surajjunni/twitter_misinformation_eradication
What it does
It classifies the English tweets/statements whether it is fake or legitimate
How we built it
We used Python as our programming language and made use of AssemblyAI API for youtube video to text conversion.
Challenges we ran into
Creating benchmark data was our biggest challenge, currently we used data from WHO and CDC websites via their tweets and youtube channels.
Accomplishments that we're proud of
Not just focusing on text tweets we also worked on links shared (Eg. youtube videos) and classified the information present
What we learned
Assembly AI api usage and web scarpping
What's next for Twitter misinformation prevention
Extracting text from images. Real time detection and prevention of misinformation. Scalable to another situation like GIS. Data collection improvement to have a better baseline. Application in stock markets.
Log in or sign up for Devpost to join the conversation.