a6e5b969-be6e-4756-ae53-47d3cb899ca4.png

52f330c8-2f13-430d-b696-7f03b43d44a6.jpg

Marketplace Dev

if you can provide any demo workflow? would be apprcitaed 

I am exploring But had some quires ..if you can provide any demo workflow? Thanks Advance

Is it possible to explain how the document classification confidence score is calculated? With some documents (between 2-5%) I am getting different results with this compared to the 'test classification' function on transact, the API (/rest/ocrClassifyExtract), and the document review function. 

The documentation mentions it's a Lucene score? This seems to suggest it's linked to the search classification plugin but this is disabled in the batch class I am using and when manually tested, provides a significantly different result.

For Example: When the same document is run through the various pathways the results are:

Test Classification (search classification):   Type A (24%)

Easy to set up, easy to use - I've installed it and was using it in a matter of minutes. 


Single document extracts - I just started developing using this solution but so far it has been intuitive. One change would be great if the output was in a datatable instead of a collection.


d97b1aee-9250-482a-a0d3-1d2b8efc0b2e.nupkg

The Ephesoft package can do digitization of semi structured and unstructured data.Package consists of 7 activitiesTransact Scope&nbsp;-&nbsp;The activity opens a connection to your instance of Transact, allowing all contained activities to be authenticated. Also pulls Batch Classes to be used at design time.&nbsp;Upload Batch&nbsp;- The activity sends multiple files to your Batch Class for processing.Get All Batch Instances&nbsp;- The activity retrieves all active Batches with a specified status.Classify Document&nbsp;-&nbsp;The activity submits a (potentially multi-page document) for OCR and classification. As this file can contain multiple different types of documents (e.g. invoice, contract, transcript...), a list of all constituent document types is returned along with the fields they may comprise.Single-Page Extraction&nbsp;-&nbsp;The activity submits a single-page file for fast OCR, classification, and extraction of fields.Multi-Page Extraction&nbsp;-&nbsp;The activity will submit multi-page files covering several document types and tables for processing.&nbsp;Get Extraction Results&nbsp;-&nbsp;Retrieves extracted fields and tables using the reference code from the previous activity. <a href=https://integrations.uipath.com/docs/ephesoft-transact target=_blank>Click here for documentation on the activities</a>

Ephesoft Transact offers machine learning-powered extraction of structured or unstructured documents

Transact for Ephesoft

Transact for Ephesoft

MicroGenesis - Ephesoft Solution Workflow…

Help

How To