Preserving EA's after batch tagging

mikelauruhn

We are going to set aside a TeamSite workarea for evaluating MetaTagger results and are looking for a way to flag and lock down extended attributes after they have been reviewed by a human (whether or not they have been changed). Here is the scenario we are shooting for:

A large collection of documents is in a TeamSite workarea and a batch process is run to tag the content with an MT vocabulary. Then, a handful of Subject Matter Experts will review a fraction of the content through the MT UI. Of those they open, they will either approve the metadata that MT set (do nothing) or correct the metadata that was set (browse vocabulary and make a change to the Extended Attribute). Now, we'd like to tune and augment the vocabulary based upon the changes that were made and tag more documents in the workarea.

Any ideas on a workflow or other way to flag and preserve these Extended Attributes so that the next MetaTagger pass will ignore these and use the updated index on the content that was NOT evaluated in the first pass?

MikeL

Migrateduser

Hey Mike,

Easiest way to do this is to copy the UID's of the suggested categories to a hidden attribute facet in a final-processor. That way it's easy to do the set-difference calculations to find the categories that were added or removed.

Cheers,
Clark