Fake and Real News Predictor

Do you know the news you read today morning was a fake one or true ? Can you find the difference between them ?

We here, try to solve this issue of fake news through the concepts of machine learning by predicting if the content of information is real or fake.

Team Members : Avishi Goyal , Medha Gulati , Shivangi Mishra , Pooja Garg

The repositories include:

1. Code - separate repositories for each algorithm
2. Data - Datasets and graph images
3. Report - Overleaf Report
4. Team Members - separate repositories for team members

Link to Kaggle Dataset:

https://www.kaggle.com/clmentbisaillon/fake-and-real-news-dataset

Number of instances: 17903 Fake News + 20826 True News

Number of attributes: 4

Whether labeled or unlabeled: labeled

Type of label information (if present): categorical

Section 3.1: Fake News Dataset

https://github.com/goyalavishi/Fake-and-Real-News-Predictor/blob/master/Data/Fake.csv

Fake News Preview :

True News Preview :

Section 3.2: Data visualisation , Preprocessing can be found in file (preprocessing.ipynb)

Word Count Comparision In Text Content of News

Word Count Comparision In Title of News

Stop Words Comparision In Title of News

Stop Words Comparision In Content of News

Tokenisation -> preprocessing.ipynb

https://github.com/goyalavishi/Fake-and-Real-News-Predictor/blob/master/Code/DataPreProcessing/preprocessing.ipynb

Sentimental Analysis -> SentimentalAnalysis.ipynb

https://github.com/goyalavishi/Fake-and-Real-News-Predictor/blob/master/Code/SentimentAnalysis/SentimentAnalysis.ipynb

Sentiments in Fake News

Sentiments in True News

Length of Characters -> news3.ipynb

https://github.com/goyalavishi/Fake-and-Real-News-Predictor/blob/master/Code/GrammarCheck/news3.ipynb

Length of Characters in a news

Proposed Methodology :

Section 4.1.1: Multinomial NB and Random Forest can be found in file (XGB, RF, NB without stematization and Lemmatization.ipynb)

https://github.com/goyalavishi/Fake-and-Real-News-Predictor/blob/master/Code/NaiveBayes/XGB%2C%20RF%2C%20NB%20without%20stematization%20and%20Lemmatization.ipynb

Final Result :

OVERLEAF DOCUMENT

https://www.overleaf.com/8941899816hbnfyxwpdzrq