Home Glossary Named Entity Recognition (NER)

Named Entity Recognition (NER) - Page 3

Named entity recognition, or NER, finds mentions of specific entities in text and assigns them to categories such as people, organizations, locations, products, dates, or monetary values. It turns unstructured language into structured information that can support search, document analysis, customer service, compliance, and knowledge graphs. Systems may use rules, statistical models, neural networks, or language models. The task becomes difficult when names are ambiguous, new entities appear, categories overlap, or a domain uses specialized terminology. Reliable NER requires representative examples, clear label definitions, multilingual testing where relevant, and evaluation of both whether an entity was found and whether its boundaries and category were correct.