Mangalmay Institute of Engineering and Technology, Greater Noida
Department of Computer Science and Engineering
Presentation on
Fake News Detection System using Machine Learning Algorithms
Presentation by: Guided by:
Ms. Srinkhala Maddhesia
Prerna Chaurasia (2007860100059) Assistant Professor
Satyam Srivastava (2007860100071) MIET - Greater Noida
Saurabh Rao (2007860100072)
Utkarsh Gupta (2007860100082)
Presentation Outline
● Introduction
● Problem Statement
● Literature Review
● Objectives
● Proposed Methodology
● Conclusion
● Future Scope
● Refrences
Introduction
● In today's digital age, the rapid spread of fake news through online platforms
undermines trust in media, shapes public opinion, and influences election outcomes,
posing a significant challenge to society and democracy.
● Our project, the Fake News Detection System (FNDS), utilizes advanced machine
learning techniques to automatically classify news articles as real or fake, aiming to
detect and combat misinformation effectively.
Problem Statement
● Spread of misinformation
○ There's a lot of wrong information and fake news on the internet.
○ This causes a big mess, makes people not trust what they read, and can hurt individuals,
groups, and society as a whole.
● Social division and Conflict
○ False stories and rumors can make people in a community fight with each other.
○ This happens when fake news spreads lies and stereotypes about different groups of people.
It can cause arguments, tension, and even riots.
Literature Review
● Fake News on Social Media: The impact on society [Paper Link]
By Femi Olan, Uchitha Jayawickrama, Emmanuel Ogiemwonyi Arakpogun, Jana Suklan &
Shaofeng Liu
o This paper concludes that it is important for social media providers to keep working
hard to understand how fake news spreads and how it affects society.
o It's important for them to use fact-checking tools to stop fake news from spreading.
This helps build trust and confidence in society.
Literature Review
● Learning Feature Engineering for Classification [Paper Link]
By Fatemeh Nargesian, Horst Samulowitz, Udayan Khurana, Elias B. Khalil, Deepak Turaga
○ This paper proposed an novel framework LFE (Learning Feature Engineering) to
perform automated feature engineering by learning patterns between feature
characteristics, class distributions, and useful transformations, from historical data.
Literature Review
● Detecting Fake News using Machine Learning: A Systematic Literature Review [Paper Link]
By Alim Al Ayub Ahmed, Ayman Aljarbouh, Praveen kumar donepudi, Myung Suh Choi
o This paper concludes the ease of spreading fake news due to widespread internet and
social media usage. Fake news can harm individuals, organizations, and influence
political opinions. Detecting fake news is crucial, and machine learning classifiers
offer a potential solution.
o The review focuses on supervised classifiers, which require labeled data for training,
but accessing such data can be challenging. Future research could explore the use of
unsupervised machine learning classifiers for fake news detection.
Objectives
● To develop an web application which helps to distinguish between the fake
news and real news.
● To develop machine learning algorithms like feature engineering, logistic
regression for the analysis of the text.
Proposed Methodology
● Frontend - React.js and Next.js with Tailwind CSS and ShadCN UI
● Backend - Django
● Datasets - Kaggle
● Data processing - Python pandas
● Machine Learning algorithms
○ Random Forest
○ Logistic Regression
Dataset from Kaggle in csv
format
Read csv files in pandas
Remove duplicate data Remove Missing value
Download cleaned
dataset
Data Cleaning Process
Data Collection
Data cleaning
Client Feature Engineering Server
Model Training
Model Evaluation
Flowchart of Web Application
Fig. Homepage of Web Application
Fig. Application page
Analysis of Average Title & Text Lengths
for Real and Fake News
Results
● Logistic Regression model results ‘Fake News’ or ‘Real News’ with an
accuracy score of 73%
● Random Forest model result 0 and 1 for ‘Fake News’ and ‘Real News’
respectively with an accuracy score of 93%.
Conclusion
● The proposed Fake News Detection System offers comprehensive solution to
address the problem of spreading fake news or misinformation around the
social media.
● By leveraging machine learning models like logistic regression and random
forest, these models will analyse the text and generate the result.
Future Scope
● Implementation of more advanced machine learning models.
● Use social media for analyzing the most trending posts in the country or
globally.
● Integrate third-party APIs to increase the amount of data source.
Refrences
• Femi Olan, Uchitha Jayawickrama, Emmanuel Ogiemwonyi Arakpogun, Jana Suklan & Shaofeng Liu Fake News on Social
Media: The impact on society , January 2022.
• Fatemeh Nargesian, Horst Samulowitz, Udayan Khurana, Elias B. Khalil, Deepak Turaga Learning Feature Engineering for
Classification , November 2017.
• Alim Al Ayub Ahmed, Ayman Aljarbouh, Praveen kumar donepudi, Myung Suh Choi Detecting Fake News using Machine
Learning: A Systematic Literature Review, January 2021.
• Iftikhar Ahmad, Muhammad Yousaf, Suhail Yousaf, Muhammad Ovais Ahmad Fake News Detection Using Machine Learning
Ensemble Methods, October 2020.
• Uma Sharma, Sidarth Saran, Shankar M. Patil , Fake News Detection Using Machine Learning Algorithms , 2021 Volume 9
Issue 3.
• Abhishek Singh, Aditya Ugale, Niraj Shah, Prof. Amruta Sankhe , Fake News Detection Using Logistic regression &
multinomial naive bayes , April 2021.
Thank
You
