Glossary Library

Technical GlossaryNatural Language Processing

Extractive Question Answering

A question answering approach that selects the answer as a span from a provided passage.

Extractive question answering is a strong approach when factual accuracy and grounding matter. Instead of generating an answer freely, the model identifies it within a given document or passage. This can reduce hallucination risk and make answers easier to justify. It is widely used in document-based assistants and enterprise knowledge systems.

You Might Also Like

Explore these concepts to continue your artificial intelligence journey.

Glossary Cover

dogal-dil-isleme

Text Normalization

The process of standardizing raw text at the spelling, formatting, and character levels to make it more consistent and processable.

Glossary Cover

dogal-dil-isleme

Unicode Normalization

The process of converting visually identical but differently encoded characters into a standard form.

Glossary Cover

dogal-dil-isleme

Stemming

An approach that reduces a word to a shorter root-like form by crudely stripping suffixes.