Aurora AITell us your case

Offering

ServicesProductsCase studies

For whom

Private EquityEnterpriseSMB
ServicesProductsCase studiesAboutBlogContact

Knowledge base

Start hereWikiGlossaryGuides

AI Glossary

Natural language processing (NLP)

NLP, natural language processing, language processing

Natural language processing (NLP) is the field of AI concerned with how machines understand, analyze, and generate human language — from text classification and translation to holding a conversation. Large language models are its modern peak.

Natural language processing (NLP) is the area of artificial intelligence concerned with how a computer can understand and produce human language. It spans both simple tasks, such as recognizing whether an opinion is positive, and complex ones — translation, summarization, or holding a conversation in natural language.

For decades NLP relied on rules and statistics, and progress accelerated thanks to machine learning and then deep learning. A common preliminary step is tokenization — breaking text into tokens, the units the model actually operates on, usually fragments of words.

The most advanced approach in the field today is the large language model built on the transformer architecture, which handles with a single mechanism tasks that were once solved separately. NLP and LLMs are not the same thing: NLP is the entire area of research and application, while an LLM is a specific class of model that now dominates it. In a company, NLP underpins search, automated query handling, document analysis, and content classification — often invisible to the user, yet present in many processes.

Related terms