An NLP task that identifies and classifies named entities — people, organizations, locations — in text.
Named Entity Recognition (NER) is a natural language processing task that identifies and classifies named entities in text into predefined categories such as persons, organizations, locations, dates, monetary values, and domain-specific concepts. NER is a foundational step in information extraction pipelines, enabling machines to understand who and what a document is about.
NER has become a critical preprocessing step for enterprise AI systems. Before documents can be indexed in knowledge graphs, searched semantically, or processed by AI agents, the entities they mention must be identified and linked to their canonical representations. Modern NER systems go beyond generic categories to recognize domain-specific entities — financial instruments, medical conditions, legal concepts, and product names — enabling precise knowledge extraction from unstructured text.
Modern NER uses transformer-based models fine-tuned on labeled datasets. The model processes text token by token, assigning each token a label indicating whether it is the beginning, inside, or outside of a named entity span, and which entity type it belongs to. State-of-the-art NER models achieve over 90% F1 scores on standard benchmarks. Enterprise NER systems often combine pre-trained models with custom fine-tuning on domain-specific data.
A financial news processing system uses NER to extract entities from earnings call transcripts. The system identifies 'Apple Inc.' as an Organization, 'Tim Cook' as a Person, '$94.9 billion' as a Revenue figure, and 'Q1 2026' as a Time period. These entities are then linked to their knowledge graph representations, enabling automated analysis of financial performance across thousands of companies.