Fake News Detector (Team-57 The Decision Trees)

Logo
Libraries Used
Working Model 1
Working Model 2

Inspiration

We see that many fake news result in political, religious and social wars among people. Such fake news are a result of political agendas, unethical activities or religious extremists. Also during this pandemic time, we felt that rumors like false death reports and fake number of cases disrupted a common man's peace. Fake news also results in daily life mishaps which might cost us time, efforts and money. In order to deal with these fake news, we have developed a model which can easily detect fake news. This detection would help us know the accurate and factual news.

What it does

Once the user enters a text paragraph/sentence or headline of news/rumor, our model simply shows an accurate output if it is Real or Fake.

How we built it

We have used Python as our main programming language. Libraries used: Numpy, pandas, re, nltk, glob, pickle, matplotlib & sklearn. Platform Used: Jupyter Noteboook & IBM Z. We rummaged for an accurate dataset in order to build a predicting model.

Challenges we ran into

Our team is always eager to implement innovative ideas. We had to find accurate and valid datasets for our ML model which was a tedious task. We also managed to come up a common solution, but all of us had different implementation ideas. Finally we came up with the best and feasible solution which not only solved our problem statement but also helped us in learning IBM Z and ML concepts.

Accomplishments that we're proud of

We focused on a simple website which is easier to use and we achieved the target of accurate prediction using all provided datasets and programming platforms. Also, all of us have worked towards giving a user friendly view of this project, using IBM Z and Python in their best way possible. It was our first encounter with IBM Z cloud, and to be honest, it was amazing and we are proud to announce ourselves to be the IBM Z users.

What we learned

IBM Z platform introduced us to a whole new level of programming enthusiasm and we learned use of new libraries and ML algorithms. Also, team work spirit and keeping up with deadlines served as an added advantage to our personalities. In a summary, during this hackathon, we learnt about tensorflow(though didn't use it), scikit learn, Natural Language Processing and most importantly, how to use the IBM Z services effectively.

What's next for Fake News Detector

We plan to convert our project into a working API and an extension for web browsers, such that when someone selects the text and click on the API button, it will highlight the text as fake or not. This can also be used in social media in order to classify news as fake or real, in order to reduce the spread of fake news.