In today's media driven society it can be difficult to tell whether or not an articles that may have been "tweeted" or "posted" are "fake" or not. Big social media companies such as Facebook and Twitter are constantly being pulled into lawsuits over the spread of these unreliable news sources. With amount of posts on these websites, it is almost impossible for these companies to have a person check over all these posts. To help vulnerable people combat falling into these traps of unreliable news sources, we created Checc News to let the user know whether or not a source is unreliable or not.
What it does
Our website provides a place for people to check whether or not a news article is reliable or not.
How we built it
For our backend we used Python to code a Multinomial Naive Bayes Classifier to classify articles using a labeled news article database. We also used Flask to build the Webapp's backend. For our frontend, we used HTML, CSS, and JS to create a functional and appealing website.
Challenges we ran into
Some challenges we ran into was cleaning up the dataset and reducing the dataset so it wouldn't skew the data when we created the model. Another challenge that we ran into was using Flask to attach the backend to the frontend because it was difficult passing the model into the flask code.
Accomplishments that we're proud of
We are proud that we have a functioning website and we were able to achieve everything that we planned to do especially for using Flask for the first time.
What we learned
We learned how to use Flask and how to use a model on Flask.
What's next for Checc News
For the future of Checc News, instead of using just an NLP model, we could combine it with a dataset of known unreliable sources and add it to the NLP model.