Data & retrieval
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.
01 ——
In plain English
Sentiment analysis labels the emotional tone of a piece of text. Classic systems output one of {positive, negative, neutral}; modern systems output fine-grained scores, aspect-level sentiment (loved the design, hated the battery), or full emotion taxonomies (joy, anger, fear, surprise).
Common applications:
- Voice of customer — sentiment over support tickets, reviews, survey responses
- Brand monitoring — how social mentions of a brand trend
- Trading signals — sentiment over earnings calls or financial news
- Product analytics — what users praise vs complain about in app reviews
How it's done today:
- General-purpose LLMs — Claude, GPT, Gemini handle most sentiment tasks well in a one-line prompt
- Fine-tuned smaller models — cheaper and faster for high-volume pipelines (DistilBERT, fine-tuned Llama)
- Specialised APIs — AWS Comprehend, Google Natural Language API, Azure Text Analytics
Where it's still hard: Sarcasm, mixed sentiment, domain-specific slang, and non-English languages outside the top 20. LLM-based pipelines handle these better than older models but can still miss.
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.
Named Entity Recognition
An NLP task that identifies and labels names of people, places, organisations, dates, and other specific entities in text.
Summarization
Compressing a longer text — a meeting transcript, an article, a chat thread — into a shorter version that keeps the key information.
Embeddings
A way of converting text (or images) into lists of numbers so an AI can measure how similar two pieces of content are.
Fine-tuning
Further training a pre-trained AI model on your own data to specialise it for a specific task or style.
LLM
Large Language Model — the type of AI behind tools like ChatGPT and Claude, trained to understand and generate text.
Back to glossaryLast reviewed May 2026