NuryeNigusMekonen/amharic-news-classifier-in-bbc

make your news categorized based on the news type # Amharic News Classifier - BBC Dataset ``` This project is a machine learning-based text classifier that categorizes Amharic news articles into four distinct categories: 1, Health, 2, Politics, 3, Sport, and 4, Technology. The classifier is trained using an enhanced dataset sourced from BBC Amharic news, with 250 labeled articles per categor - if we need more accurate one we can add more data by scrapping - means by adding page number to be scraped. project is requested by HILCOE school NLP subject. ``` ## Features ``` - Classifies Amharic news articles into 4 categories. - Enhanced with an expanded dataset for better generalization. - Evaluation using accuracy, precision, recall, F1-score, and confusion matrix. ``` ## Project Structure ``` NLP/ data/ bbc_amharic_dataset_300+.csv models/ amharic_news_classifier_300+.pkl tfidf_vectorizer_300+.pkl figures/ accuracy_bar.png classwise_metrics.png confusion_matrix.png scripts/ classifier.py import_requests_for_more_than_300_articles.py UI_testing_with_model.py models/ *.pkl *.pkl docs/ NLP_Projects_Guide.pdf .gitignore requirements.txt LICENSE README.md ``` --- ## Performance Results ## Overall Accuracy ## Class-wise Metrics ## Confusion Matrix ## Installation ``` 1. Clone the repository: A, git clone github.com --- https B, git clone git@github.com:NuryeNigusMekonen/amharic-news-classifier-in-bbc.git --- ssh cd amharic-news-classifier-in-bbc 3. Install dependencies: pip install -r requirements.txt ``` ## How to Run ``` 1. Train the Classifier: python classifier.py 2. Test on Sample Data: python UI_testing_with_model.py ``` ## output will look like below 1. INPUT UI 2. PREDICTION RESULT ## Author ``` 👤 Author Nurye Nigus Electrical & Software Engineer 📧 Email : nurye.nigus.me@gmail.com 🌐 LinkedIn (linkedin.com) 🐙 GitHub: @NuryeNigusMekonen 📞 Phone : +251929404324 ``` ## License This …