Field extraction training
Field extraction training
Overview of field extraction training in ABBYY FlexiCapture: scenarios, scope, administrator-led setup, and auto-learning from operator corrections.
ABBYY FlexiCapture can be trained to find the positions of simple fields and recurring groups on structured and semi-structured documents. During extraction, it considers the environment of each field, such as static text, separators, and checkmarks, so fields of the same type should have a similar environment. For documents that belong to the same type but look very different, use the Document Variant feature.
There are two training scenarios: an administrator trains the project before processing begins, or the system learns automatically from operator corrections during processing. For what training covers, see Scope of field extraction training.
