Speech and Language Processing (3rd ed. draft)
8 days ago
- #Release
- #NLP
- #Textbook
- Preference alignment with DPO in posttraining (Chapter 9).
- New ASR (Whisper) and TTS (EnCodec and VALL-E) material in Chapters 15 and 16.
- Restructured earlier chapters to fit current teaching methods.
- Moved Naive Bayes to Appendix, using Logistic Regression for classification.
- Moved PPMI to appendix, focusing on tf-idf in Chapter 11.
- Introduced LLMs, LLM sampling, and training in Chapter 7 before Transformers in Chapter 8.
- Delayed RNN/LSTM chapter to 13, allowing flexible teaching order.
- Restructured Chapter 2 to focus on tokens, words, and Unicode.
- Fixed typos and added new slides.
- Divided dialogue and chatbot chapter into other chapters, with parts moved to Chapter 25 and Appendix J.
- Encouraged use of draft chapters and slides for feedback.
- Provided citation details for the book.
- Previous drafts (Jan 2025 and Aug 2024) are available.