← Natural Language Processing learning pathNatural Language Processing · Module 4 of 10
Text classification
Sort text into categories like spam or sentiment.
📖 Key ideas
- Examples: spam detection, sentiment, topic labelling, intent detection.
- Classic: TF-IDF + logistic regression or Naive Bayes.
- Modern: fine-tune BERT-like models.
- Metrics: accuracy, precision, recall, F1.
- Imbalanced classes need careful evaluation.
💡 Real-life example: Text classification is a postman sorting letters into the right bins by reading them.
Spam filterClassifies emails as spam or not
Naive BayesSimple probabilistic classifier
Intent detectionFinding what a user wants
F1 scoreBalance of precision and recall