Currently, the NLP technology used in ABBYY FlexiCapture has the following limitations:
- Selecting multiple overlapping fields or segments leads to a decrease in extraction quality.
- Selecting a field outside of a segment when the entity corresponding to that field is extracted from that segment will lead to the field not being trained.
- Fields used in training cannot be nested more than one level deep. You can train a field inside a group of fields, but fields inside a group that is part of another group cannot be trained. The nesting level of a field into which an entity should be extracted starts at the source field.
- You cannot create multiple segmentation models within one Document Section
- Multi-lingual NLP models are not supported.