Natural Language ProcessingBeginner1:11 video5 chapters

Named Entity Recognition — lecture notes

Find people, organisations, places and dates in text and tag every token with BIO labels.

0:001. Introduction

News articles, contracts and medical notes are full of names, places and dates. Named entity recognition finds them automatically.

0:092. Tagging entities

Tagging entities — Named Entity Recognition

Here the model reads a sentence and labels each token. Ada Lovelace is a person, Charles Babbage is another person, London is a location and 1843 is a date. Each token gets a BIO tag. B marks the beginning of an entity, I marks its continuation, and O means outside any entity.

0:313. A sequence problem

A sequence problem — Named Entity Recognition

NER is a sequence labelling problem. A token’s label depends on its neighbours: New York is one place. The same word can be a company or a fruit. So models read the whole sentence, from classic conditional random fields to modern fine-tuned Transformers.

0:494. Applications

NER helps search engines understand queries, extracts parties and amounts from legal and financial documents, finds drugs and symptoms in medical notes, and builds knowledge graphs.

1:015. Recap

To recap. NER labels entities in text using BIO tags. Context decides ambiguous cases, and fine-tuned Transformers are the modern approach.

Key takeaways

NER finds and classifies entities such as people, organisations, locations and dates.
BIO tagging marks the Beginning, Inside and Outside of entities.
It is a sequence labelling task where context resolves ambiguity.
Modern systems fine-tune Transformer models; classic ones used CRFs.

Check yourself

In BIO tagging, what does “I-LOC” mean?
Show answer
A token inside (continuing) a location — I = inside, continuing an entity.
Why is “Apple” hard for NER?
Show answer
It can be a company or a fruit — The entity type depends on context.
Which tag is given to tokens that are not part of any entity?
Show answer
O — O means outside.

Go deeper

Named Entity Recognition: Finding People, Places and Organisations · The AI Lecture Hall
Part-of-Speech Tagging and Conditional Random Fields · The AI Lecture Hall