Text Classification
Definition
Text classification assigns texts to one or more predefined categories.
The assignment of a text to one or more predefined classes.
Simply explained
The system sorts texts into predefined drawers, for example, service requests by topic and urgency.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
Incoming service requests are pre-sorted by topic and urgency.
Limits and pitfalls
- Ambiguous labels, class shift, and small groups can obscure performance and fairness.
Origin & history
The field evolved from rule- and feature-based methods to pretrained transformer models.
Use in mid-sized companies
- Possible application: Incoming service requests are pre-sorted by topic and urgency.
Frequently asked questions
What does text classification mean?
The assignment of a text to one or more predefined classes. What should you watch out for? Ambiguous labels, class shift, and small groups can obscure performance and fairness.
Sources
- BERT · Devlin et al., BERT
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: