Data & retrieval
Named Entity Recognition
An NLP task that identifies and labels names of people, places, organisations, dates, and other specific entities in text.
01 ——
In plain English
Named Entity Recognition (NER) is a classic natural-language processing task: scan a passage of text and tag the spans that name real-world entities. "Sam Altman, CEO of OpenAI, met with the EU in Brussels on Tuesday" yields entities Sam Altman (person), OpenAI (organisation), EU (organisation), Brussels (location), Tuesday (date).
Why it's still useful in the LLM era:
- Indexing and search — extract entities from a document set, build filters on top
- Privacy / redaction — detect PII (names, emails, phone numbers) before storing or sharing
- Knowledge graph construction — entities become nodes, NER is the extraction step
- Analytics — what people, products, or places appear most often in your data
How it's done today:
- Specialised models — spaCy, Stanford CoreNLP, Hugging Face's transformer NER models
- LLM-based — just ask GPT/Claude/Gemini to extract entities in JSON; works well for most cases
- Hybrid — small specialised model for speed, LLM for ambiguous cases
NER tools that once needed an ML team to deploy are now a few lines of code or an API call away.
02 ——
Related terms
Natural Language Processing
The branch of AI focused on understanding, generating, and working with human language — covering everything from spell-check to ChatGPT.
Embeddings
A way of converting text (or images) into lists of numbers so an AI can measure how similar two pieces of content are.
Knowledge Graph
A structured representation of entities and the relationships between them — used to give AI systems explicit, queryable facts.
Structured Output
Forcing an LLM to return data in a specific format — usually JSON matching a schema — so downstream code can parse it reliably.
Sentiment Analysis
An NLP task that classifies text as positive, negative, or neutral — used at scale for reviews, support tickets, social media, and survey responses.
Fine-tuning
Further training a pre-trained AI model on your own data to specialise it for a specific task or style.
Back to glossaryLast reviewed May 2026