named entity recognition

Syllable breakdown

Dictionary meaning

the NLP task of identifying and classifying named entities — proper nouns that refer to specific real-world objects — in text into predefined categories such as person names, organisations, locations, dates, monetary amounts, and product names; NER is essential for information extraction, knowledge base population, question answering, and document summarisation; early NER systems used handcrafted rules and gazetteers; statistical systems using conditional random fields over surface features were dominant through the 2000s; neural NER systems using bidirectional LSTMs with character-level representations and contextual word embeddings dramatically improved performance; transformer-based models fine-tuned on NER benchmarks such as CoNLL-2003 now achieve near-human performance on standard English NER; cross-lingual NER remains challenging because entity boundaries and categories differ across languages; NER in informal text — tweets, forum posts — is more difficult than in news text because of inconsistent capitalisation, abbreviations, and named entities not found in training data

Examples

• The NER system correctly identified 'Apple' as an organisation rather than a fruit in the financial news article.
• Named entity recognition is used to extract drug names and disease mentions from clinical trial reports for pharmacovigilance.

Synonyms: entity extraction, entity recognition, NER