Question Answering
Definition
Question answering provides suitable answers from existing information or a model.
A method that generates or extracts a suitable answer from knowledge, documents, or a model for a natural language question.
Simply explained
A system searches for or formulates an answer to your question instead of just listing possible documents.
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
An assistant system answers a question about the travel expense policy with a source reference.
Limits and pitfalls
- Confusing fluent answers with verified correctness or ignoring missing sources and access rights.
Origin & history
QA has been a research area since early AI systems and has been greatly expanded by large text corpora and transformers.
Use in mid-sized companies
- Possible application: An assistant system answers a question about the travel expense policy with a source reference.
Frequently asked questions
What does question answering mean?
A method that generates or extracts a suitable answer from knowledge, documents, or a model for a natural language question. What mistakes should be avoided? Confusing fluent answers with verified correctness or ignoring missing sources and access rights.
Sources
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: