Annotations
Annotations are feedback left by reviewers as they review the document extractions. They consist of:
- Thumbs-up/thumbs-down indications
- Text selections from documents associated with a provision
- Edits to text selections and extractions
The AI uses these types of information to train itself. For a given custom provision, you can include the annotations reviewers have made or delete them, removing them from AI training.
Note: Extractions using a prompt or key phrase are not used in continuous learning, but Discovery AI does store user feedback.
A thumbs-up sends the extraction to the AI model for training. Changes (add, update, or delete) made to an extraction without a thumbs-up are not used to train the AI, though this does not adversely affect the ongoing transaction. Thumbs-down indications are also used for training. Both thumbs-up and thumbs-down indications are recorded. If the extraction is annotated (an add or update) and you select the checkbox to the left, your annotation is saved to the agreement record even if you do not give thumbs-up feedback.Note: Feedback for image-based extractions is currently saved. However, because feedback is user-annotated text and not an image, it does not improve the accuracy of future image-based extractions.
