Natural Language Processing (NLP)

Status: public · Confidence: medium (0.82) · Basis: verified_sources

## TL;DR

NLP enables computers to understand, interpret, and generate human language. Tasks: text classification, named entity recognition (NER), machine translation, summarization, question answering, sentiment analysis. Before Transformers (2017): rule-based → statistical → neural (RNN/LSTM). After Transformers: pre-trained language models dominate.

## Core Explanation

Pre-Transformer era: TF-IDF, Word2Vec (Mikolov 2013), GloVe (Pennington 2014) for word embeddings. The Transformer architecture uses attention mechanisms for sequence transduction without recurrent or convolutional layers. Post-Transformer: BERT (2018), GPT series, T5 — pre-train on massive text, fine-tune for specific tasks.

## Further Reading

- [Speech and Language Processing (3rd Ed, Jurafsky & Martin)](https://web.stanford.edu/~jurafsky/slp3/)

## Related Articles

- [Low-Resource NLP: Multilingual Models, Endangered Language Preservation, and Translation](../low-resource-nlp.md)
- [Text-to-SQL: Natural Language Database Querying with Large Language Models](../text-to-sql.md)
- [AI for Code Translation: Language Migration, Legacy Modernization, and Transpilation](../ai-code-translation.md)