This project is a powerful News Topic Classifier that automatically categorizes news articles into one of four topics: 🌎 World , 🏅 Sports , 💼 Business and 🔬 Science/Technology
Built using Natural Language Processing (NLP) techniques and a Logistic Regression classifier (achieving 91% accuracy), the project features a modern, interactive web app built and deployed with Streamlit Community Cloud.
🤖 Automatic Topic Classification using Logistic Regression and TF-IDF
💻 Interactive Web App: User-friendly, visually appealing, and easy to use
📊 Model Confidence Visualization: See prediction probabilities for each topic
📈 Text Analytics: Get word count, unique words, top keywords, and more
🧹 Preprocessing Transparency: View how your text is cleaned before prediction
📝 Sample Inputs: Try out the app with real-world news examples
Link : https://new-article-classification-using-nlp-nrsrc3pxheagvv2vwsxqvd.streamlit.app/
🖥️ Try the App on Streamlit Cloud!
├── app.py # Streamlit web app
├── log_reg_model.pkl # Trained Logistic Regression model
├── tfidf_vectorizer.pkl # TF-IDF vectorizer
├── requirements.txt # Project dependencies
├── README.md # This file
└── News_Classification.ipynb # Jupyter notebook
Classifier: Logistic Regression
Feature Extraction: TF-IDF
Test Accuracy: 91%
Classes: World, Sports, Business, Science/Technology
🤖 Built a robust Logistic Regression classifier with 91% accuracy for news topic classification.
💻 Developed an interactive Streamlit web app for easy, real-time predictions.
📊 Included model confidence visualization and text analytics for deeper insights.
🚀 Deployed the app on Streamlit Community Cloud for public access and demonstration.