In today's era of misinformation and data, it is more important than ever to determine which sources are credible and trustworthy, as opposed to those that are intended to spread misinformation. It is this blatant misinformation, and the clear danger that it presents, that has inspired us to create Credibot to analyze the credibility of claims and help stem the tide of fake news.
What it does
The Credibility bot determines how reliable a certain source is based on the articles Polarity and Accuracy.
How we built it
Our bot is written in python. It uses machine learning to determine the reliability of a source.
Challenges we ran into
We ran into some trouble when we tried to compare the credibility of different tweets to sources. Another problem was the training times, as some portions took a long time to train.
Accomplishments that we're proud of
We are proud of the fact we were able to implement this bot using 3 different critera.
What we learned
We learned different types of models such as seq2seq, a model that is used for language interpretation.
What's next for Credibility Bot
We hope we can add more criteria to make Credibility Bot even more reliable and robust!