問題文
What does content extraction do as a component of the Content Understanding pipeline?
選択肢
- It splits documents into logical sections for targeted processing, so that a long contract is divided at its clause boundaries and each section can be handled by the settings that suit it rather than by one setting for the whole file
- It applies the field extraction schema to produce key-value pairs
- It converts unstructured input into normalized text and metadata, using optical character recognition, layout recognition, and transcription for audio and video
- It calculates the confidence score for each extracted value