Named entity recognition

Also: NER, named-entity recognition, entity extraction

A sequence-labeling task that both locates entity spans and classifies them, usually by tagging each token with a BIO label (B-PER, I-PER, O, …). Context matters: "Washington" can be a person, a city or an organization.

Example

"Barack Obama was born in Hawaii" → Barack/B-PER Obama/I-PER was/O born/O in/O Hawaii/B-LOC: one PERSON span and one LOCATION span.

See it in the lesson: Named Entity Recognition →

Related terms

More in Language & LLMs