Ephesoft Queries - Compare Validation Patterns, Compare OCR Confidence Values

Sometimes it's easier to compare settings across index fields by looking at the values in the database instead of navigating down to each field in the Ephesoft UI.  However, changes to the values still need to be made in the Ephesoft UI.

Compare Validation Patterns
The following query can be used to view validation rules across index fields grouped by index field. This is helpful for making sure that the same validation rules are applied to the same fields across different document types. Query the batch_class table to find the batch_class_id value for your batch class (in red).

select 
a.field_type_name,
b.pattern,
c.document_type_description
from 
field_type a,
regex_validation b,
document_type c
where
a.id = b.field_type_id
and a.document_type_id = c.id
and c.batch_class_id = "12"
order by 1;

Compare OCR Confidence Threshold Values
The following query can be used to view field-level OCR Confidence Threshold values on a form. Higher values will result in more batches stopping in Validation, but overall accuracy should be improved as a result. For Ephesoft installations where multiple forms have similar fields, this query can be used to ensure that the same index field has consistent OCR confidence values across different forms. Query the batch_class table to find the batch_class_id value for your batch class (in red).

select 
a.field_type_name,
a.is_hidden,
a.ocr_confidence_threshold,
c.document_type_description
from 
field_type a,
document_type c
where
a.document_type_id = c.id
and c.batch_class_id = "12"
order by 1;

Additionally, care should be taken to ensure that hidden fields (is_hidden = 1) have OCR confidence threshold values of 0 (zero). I've seen situations where a hidden field with a high OCR confidence value can cause a batch to stop in Validation even though all of the non-hidden fields have good confidence values, causing the user to wonder why the batch stopped in Validation.