# Sentiment Analysis

> Reading text to tell if it sounds good, bad or in between.

**Field:** Artificial Intelligence  
**Category:** NLP & Language  
**Also known as:** Opinion Mining, Sentiment Classification

## Think of it like

Think of it like you read a note. Then you say if it sounds glad, sad or just fine.

## Definition

Sentiment analysis reads a piece of text and decides what feeling it shows. Most tools sort text into positive, negative or neutral, and some give a score in between. It is a kind of classification, trained on text that people have already labelled. Firms use it to read thousands of reviews or posts at once. Jokes and sarcasm are still hard for it.

## Example

Emma's shop gets 4,000 reviews a month. Sentiment analysis flags the 300 negative ones so her team reads those first.

## Why it matters

It turns piles of views into a number a team can track. One is the share of happy reviews each week.

## Common mistake

Trusting it on sarcasm. 'Great, it broke on day one' has a happy word in it and a very unhappy meaning.

## Three principles

- **Reads the feeling.** Positive, negative, neutral.
- **A kind of classification.** Learned from labelled text.
- **Works at scale.** Thousands of texts at once.

## Related terms

- [Classification (Machine Learning)](https://learn.benibix.com/term/ai/classification)
- [Natural Language Processing (NLP)](https://learn.benibix.com/term/ai/natural-language-processing)
- [Supervised Learning](https://learn.benibix.com/term/ai/supervised-learning)
- [Text Classification](https://learn.benibix.com/term/ai/text-classification)

![Three short reviews, each pointing to a label: one positive, one negative and one neutral.](https://learn.benibix.com/diagrams/nlp-sentiment.svg)

---

Source: https://learn.benibix.com/term/ai/sentiment-analysis
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-18
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
