A document was was flagged for low confidence and came to Action center in UiPath. I see 2 fields checkbox was not checked, eventhough it had 100% ocr confidence and 98% extraction confidence.
Can you please help why that fields were not checked and we have to manually check the box when everything looks right.
Hi @Techbuzz
What is the field type defined for that value. Is it integer/float or string?
In Document Understanding, ‘Total’ field is mentioned as Number and another field is Table. In Taxonomy of DU, Total is mentioned as Text Type. Does Action center has separate configuration or something?
Hi @Techbuzz
The schema defined in Taxonomy is expecting text but the model is extracting it as number, that ambiguity might causing this issue, also if the similar field are found at multiple places that causes people to validate it again at action center. you can train few more documents with same datatype at both the places and can test this again for better results.