Natural Language Processing

Working with text: tokens, embeddings and transformers.

Natural Language Processing is taught here in 12 lessons, running from Text preprocessing and tokens through to Evaluating NLP systems. Each lesson takes one topic, shows the working code, and links onward to the next, so the course can be read straight through in order.

Lessons

1. Text preprocessing and tokensTurn raw text into clean, tokenised input: normalisation, a sparse baseline, and how subword tokenizers actually split words.
2. Word and sentence embeddingsWhy dense vectors replaced sparse counts, how cosine similarity works, and how to use sentence embeddings for real retrieval.
3. Transformers and fine-tuning basicsWhat self-attention computes, how encoder and decoder stacks differ, and when fine-tuning is worth the cost compared with prompting.
4. Classical text classificationTF-IDF, naive Bayes and linear models, and why a well-tuned non-neural baseline is the number you have to beat before reaching for a transformer.
5. Sequence labelling: POS tagging and NERBIO tagging, CRFs and BiLSTM taggers, transformer token classification, subword alignment, and entity-level evaluation that is actually correct.
6. Text similarity, clustering and topic modellingDocument similarity, near-duplicate detection, clustering without a fixed k, and using LDA or BERTopic to get topics you can actually label.
7. Semantic search and vector databasesChunking strategies, FAISS and Chroma indexes, hybrid keyword plus vector search, and reranking with a cross-encoder.
8. SummarisationExtractive TextRank, abstractive sequence-to-sequence and transformer summarisers, length control, and checking that the summary is actually faithful.
9. Question answering and reading comprehensionExtractive QA with SQuAD-style models, open-domain QA with retrieval, grounding answers in the passage, and abstaining when the answer is not present.
10. Translation and multilingual pipelinesSequence-to-sequence and multilingual models, tokenisation for non-Latin scripts, quality estimation, and the post-processing translation actually needs.
11. Working with large language models for NLP tasksZero-shot and few-shot prompting, instruction tuning, structured extraction into JSON, and a decision rule for prompting versus fine-tuning.
12. Evaluating NLP systemsBLEU, ROUGE, METEOR and BERTScore, entity-level F1, designing human evaluation, error analysis, and reporting uncertainty honestly.

More in AI & Intelligent Development

AI Basics AI Agents Math for AI Machine Learning scikit-learn TensorFlow PyTorch LangChain Ollama OpenCV Codex Claude Code OpenCode Vibe Coding Selenium Playwright

FAQ