Issue

How do I reconcile the data displayed under "Documents Classified" and "Human in the Loop"?

Resolution

Difference between "Documents Classified" and "Human in the Loop"

1. Dashboard Documents classified includes three sections: loaded, classified and published
2. All sections share common features such as created and processed dates, identifier, raw file name, and document type
3. The created date indicates when documents are loaded into VMR via locked bag or electronically, while the processed date indicates when documents are automatically classified into the correct document type
4. 'L', 'C', and 'P' inside each menu indicate 'loaded', 'classified', and 'published', respectively, and are displayed as validation
   

   
   

   The "Human in the Loop" feature displayed on the dashboard allows users to publish documents that have been automatically classified and placed in the "Ready" menu. Users can publish files and metadata individually or in bulk to Class fund documents or SharePoint.

   
   
   
   

   
   

   Documents that have not been trained using AI for automated document type classification, and raw data that is not automatically mapped to VMR lookups, are flagged as errors. This requires human intervention to open the PDF, map the file to the relevant document type from the PDF grid, and manually map raw data to VMR lookups. Once this is done, the document and metadata can be published to Class fund documents or SharePoint.
   

   

What's Next?

Reconciling Data in the VMR Dashboard: Documents Classified vs Human in the Loop (HITL)