← Back to glossary
Browse letter U hub
Unsupervised Learning
Unsupervised learning discovers hidden structure in data through methods such as clustering and representation learning.
Definition
A type of machine learning in which models identify patterns in data without labelled training examples.
How It Works
Unsupervised Learning helps teams build predictable AI and translation workflows by setting clear expectations for quality, consistency, and decision-making.
In production environments, this concept is applied with process controls such as human review, terminology alignment, and repeatable quality checks across multilingual content.
It is foundational in modern AI systems and language model training, where large-scale unlabelled data drives representation quality.
Key Concepts
- core principle of unsupervised learning
- workflow-level implementation
- terminology and quality consistency
- human validation before publication
Where It Is Used
- localisation workflows
- AI translation pipelines
- multilingual content production
- cross-referencing related concepts such as Universal Language Models