# Annotation

> Source: https://sukruyusufkaya.com/en/glossary/anotasyon
> Updated: 2026-05-13T21:01:07.661Z
> Type: glossary
> Category: yapay-zeka-temelleri
**TLDR:** The process of adding meaningful labels, notes, or markings to data so that it becomes usable for model training.

<p>Annotation is one of the most invisible yet labor-intensive layers of many AI projects. Drawing bounding boxes on images, labeling entities in text, separating speakers in audio, or marking fields in documents are all examples of this process. Poor annotation can weaken even good data, while high-quality annotation can significantly expand what a model is able to learn. For that reason, annotation is not merely a data preparation task; it is directly tied to model quality and system reliability. Strong AI products are often built on top of strong annotation workflows.</p>