Social Crypt is a utility tool that helps to solve the problem of detection of fake propagandist news and Twitter sentimental analysis to provide a detailed dashboard for users to view unusual activities and network analysis of these bodies. It is a modern UI/UX solution with a simple dashboard with multiple features to cater to the needs of users in need to factual check the information of news and social media. The models used are machine learning and deep learning models with TensorFlow support and hugging face.
We ran into a few challenges during the building process. We wanted to make realtime system for the problem statement using twitter API , but it couldn't be possible due to not having access to it. The solution was to scrape it in realtime and multithreaded to able to do it. Also the factual checking was difficult hence we had to have discussions with our mentor to get some idea on how we can approach it & we did implement it successfully.
Technologies used
Discussion