AI glossary · Fundamentals
Natural language processing (NLP)
In one sentence
Natural language processing (NLP) is the branch of artificial intelligence that lets machines work with human language: read, analyse, classify, translate and produce text or speech the way people use it.
Also known as: NLP · computational linguistics · language processing · text analytics
What is natural language processing?
Human language is ambiguous, depends on context and changes with whoever is speaking. Computers, on the other hand, work with exact data. NLP is the discipline that bridges the two. It started in the 1950s with machine translation and today sits inside your phone’s autocorrect, spam filters and every chatbot.
Put it to work
AI agents that look after your customers
From €49/mo + VAT
Voice, WhatsApp and email with the phone line included: agent, numbers and CRM in one fee.
See plans and pricingIt’s a broad term with several sub-fields. The one mentioned most is natural language understanding (NLU), which focuses on working out meaning.
Related to this term
UTIO agents
- Available 24/7 in 26 languages
- Set it up yourself for €0: built from your website
- 14-day free trial, no card needed
From €49/mo + VAT
See plans and pricingHow it works
An NLP system goes through several steps, more or less visibly:
- Prepare the text: split it into sentences and smaller units (tokens), normalise case and strip out noise.
- Analyse the structure: which words are verbs, nouns or dates, and how they relate.
- Represent the meaning: turn the text into numbers a machine can compare.
- Solve a task using that representation.
Typical NLP tasks:
| Task | What it does | Example use |
|---|---|---|
| Text classification | Assigns a category | Separating orders from complaints |
| Sentiment analysis | Detects tone | Prioritising unhappy customers |
| Entity extraction | Pulls out specific data | Names, dates, amounts |
| Translation | Converts to another language | Serving customers in several languages |
| Summarisation | Condenses text | Summing up an email thread |
| Generation | Writes new text | Drafting a reply |
Until a few years ago, each task was handled by its own model, trained specifically for it. Large language models changed that: a single one does almost all of them by following instructions.
Example
A building and renovation firm in Cardiff receives a mix of requests by email and WhatsApp: quotes, complaints, questions about timings. With NLP, each message is sorted by type, the name, address and kind of job are extracted, and the system flags whether the tone is urgent or annoyed. The team starts the day with an ordered list instead of a chaotic inbox.
What it means for your business
You don’t need to understand NLP to use it, but it helps to know what’s happening underneath. When UTIO’s email agent sorts your mail into threads, puts urgent messages at the top and quarantines suspicious ones, that’s NLP at work. When the voice agent understands natural speech, interruptions and accents, that’s NLP too. The difference from systems of a few years ago is that you no longer train a model per task: the agent works from your instructions and your business information.
Frequently asked questions
What's the difference between NLP and NLU?
NLP is the whole field: everything a machine does with language. NLU is one part of it, the part concerned with understanding meaning: what someone wants and which details they give. All NLU is NLP, but not the other way round.
Is ChatGPT NLP?
Yes. Language models like the ones behind ChatGPT are now the main tool of NLP. Each task (translating, classifying, summarising) used to need its own system; now a single model handles many of them.
Does NLP also mean neuro-linguistic programming?
The same initials are used in psychology for neuro-linguistic programming, which is unrelated. In AI, NLP always means natural language processing.
Related terms
- Natural language understanding (NLU)
Natural language understanding (NLU) is the part of NLP that deals with the meaning of a message: what the person wants (their intent) and which specific details they give (entities such as dates, names or products).
- LLM
An LLM (large language model) is an AI model trained on vast amounts of text to predict the next word. That single skill lets it draft, summarise, translate, classify and hold a conversation.
- Text classification
Text classification is the AI task of automatically assigning one or more categories to a piece of text, such as an email or a message: type of request, topic, priority or language. It's the basis for sorting, prioritising and routing work.
- Sentiment analysis
Sentiment analysis is the AI technique that detects the emotional tone of a text (positive, negative or neutral, and sometimes emotions such as anger or urgency) so you know how the writer feels without reading everything by hand.
- Tokenization
Tokenization is the process of splitting text into tokens, the pieces a language model understands, and mapping each one to a number in its vocabulary. It's the first step before a model can read or write anything.
Updated on