This project focuses on fake news detection using machine learning techniques. The tech stack includes Python, pandas for data manipulation, scikit-learn for machine learning, and Natural Language Processing (NLP) libraries for text analysis. The primary algorithm used is Linear Support Vector Classifier (LinearSVC) for binary classification.
Challenging Parts🚀:
Data Quality: Ensuring a high-quality labeled dataset is crucial. Cleaning, balancing, and preprocessing data for effective training were challenging.
Feature Engineering: Selecting relevant features and text preprocessing required careful consideration to capture fake news patterns.
Dataset Used: IFND from Kaggle : (https://datasetsearch.research.google.com/search?query=Fake%20Content%20Detection&docid=L2cvMTFzODY4el85cA%3D%3D) Author: Sonal Garg
Results✨:
The model achieved accuracy of 96.58% in classifying real and fake news articles. However, challenges remain in handling evolving fake news tactics.