BBC News Category Classification with NLP

Project Overview

This project aims to classify BBC news articles into predefined categories using Natural Language Processing (NLP) techniques and Machine Learning algorithms.

The dataset contains news articles from five categories:

Business
Entertainment
Politics
Sport
Technology

Text preprocessing, stopword removal, TF-IDF vectorization, and machine learning models were applied to automatically predict the category of a news article.

Dataset Information

Dataset: BBC News Dataset
Number of Articles: 2,225
Number of Categories: 5

Features

Title
Content

Target

Category

Data Analysis

Several exploratory data analysis (EDA) steps were performed:

Dataset structure examination
Missing value analysis
Category distribution visualization
Text length analysis
Average word count analysis
Most frequent words analysis
WordCloud visualization

Text Preprocessing

The following NLP preprocessing techniques were applied:

Lowercase conversion
Special character removal
Number removal
Stopword removal
Text cleaning using Regular Expressions

Feature Engineering

TF-IDF (Term Frequency – Inverse Document Frequency) was used to transform text documents into numerical vectors.

TF-IDF Features

5000 Features

Resulting Matrix Size

(2225, 5000)

Machine Learning Models

Two machine learning algorithms were trained and compared.

1. Multinomial Naive Bayes

Suitable for text classification problems and commonly used with TF-IDF features.

2. Logistic Regression

A linear classification algorithm that performs extremely well in NLP classification tasks.

Model Results

Model	Accuracy
Multinomial Naive Bayes	97.08%
Logistic Regression	97.30%

The Logistic Regression model achieved the highest performance and was selected as the final model.

Final Model Files

news_category_model.pkl
tfidf_vectorizer.pkl

Streamlit Application

A Streamlit web application was developed where users can enter a news article and instantly receive a predicted category.

The application:

Accepts user news text
Performs preprocessing
Applies TF-IDF transformation
Predicts category using the trained Logistic Regression model
Displays prediction results

Live Demo

The application is deployed on Hugging Face Spaces.

Project Links

Related Resources

🚀 Launch Application

💻 GitHub Repository

Project Structure

bbc-news-category-classification
│
├── src
│   ├── streamlit_app.py
│   ├── news_category_model.pkl
│   └── tfidf_vectorizer.pkl
│
├── BBC_News_Category_Classification.ipynb
├── requirements.txt
└── README.md

Libraries Used

Pandas
NumPy
Matplotlib
Seaborn
Scikit-Learn
Joblib
Streamlit
WordCloud

Conclusion

In this project, BBC news articles were successfully classified into five categories using NLP techniques and machine learning algorithms. After preprocessing and TF-IDF vectorization, Logistic Regression achieved the best performance with approximately 97.30% accuracy. The results demonstrate that traditional NLP approaches combined with machine learning can provide highly accurate solutions for automatic news categorization tasks.

Author

M. Sefer Kaya

GitHub: https://github.com/MSK-237

Hugging Face: https://huggingface.co/MSK34

Downloads last month: -; Downloads are not tracked for this model. How to track