As a developer, using forms extractor as a data extraction method in developing document understanding processes is an excellent tool for non-variable structured documents.
Fields can be defined in the forms extractor template with anchor or custom area definitions. Anchor definitions are a great advantage when documents are skewed while scanning etc. Anchor definition are a distinctive method of defining the bounds of a custom area from which data is to be extracted is to use field-level anchors. These allow for targeting data extraction based on field-level configurations, thus allowing for more flexibility when defining your form extraction rules.
However, in my practice, I’ve noticed that table definitions are more rigid with custom area boundaries, and in cases where documents are skewed, the model is failing to recognize table field/column boundary areas. This lowers extraction accuracy as tables only accepts custom areas.
Hence, would love the UiPath document understanding team to add anchor level flexibility for tables for a more robust and accurate extraction methodology.
@Monica_Secelean @Anca_Bolca @Forum_Staff @Vibhor.Shrivastava