WHO (World Health Organization) has stated that this current pandemic is also creating an "infodemic". This is because of the many fake news articles being spread around on social media platforms like Whatsapp, Facebook, and many more. In fact, an Arizona man died after taking a chemical called chloroquine after hearing that it was a coronavirus treatment. (Time) After hearing about this, we knew there was something we had to do to stop misinformation. To help solve this we have created a bot that interacts with people on chatting and texting platforms to help stop the spread of misinformation in its tracks.
What it does
Our bot uses natural language processing (NLP) to help verify statements inputted by the user. Users can easily check if some of the headline claims are a myth by asking the bot. Currently, our bot lives on Slack, Discord, and SMS. We have also implemented statistics about the current pandemic allowing users to get up to date information on the current pandemic.
How we built it
We built it using Stanford's Glove model for the base Word2Vec, then we retrained using some new coronavirus hit words. This allows us to get a lot of words in our model and get more accurate results for when the user uses the bot. For the discord we used discord.py. We also made a Slack, and Twilio bot using their respective APIs. We also use online databases to help check for links and the current pandemic status.
Accomplishments that we're proud of
We are by no means NLP experts, but we were able to use the others' research in a novel approach to curbing the spread of misinformation.
What's next for Fakovid
We'd love for Fakovid to make it onto everyone's devices so that people have a second eye in determining the legitimacy of information. We are currently working on a central API so that Fakovid can be accessed from everywhere.
Try It out
axios, bolt.js, discord, gensim, glove, nestjs, node.js, python, rxjs, slack, twilio