We have a question on how PC (Predictive Coding) will handle the analysis if an email body is not privileged but one of the earlier/later replies or forwards is privileged...
All messages are indexed (no derived emails) so I believe each email would then be tagged on its own merit, and the prediction will be based on each email and not on the entire thread. What are the thoughts on this from the community?
Also, how would this be handled if some of the messages are derived?
Thanks for any thoughts/comments.
In Clearwell 7.1.2 where Transparent Predictive Coding was introduced Seperate Tagging of Attachements was also introduced
If I understand your query the message body will be identified on it's own merit as would any attachements.
In the case of a derived email the origional item (not derived) where the derived content was pulled from would be tagged as a derived message is not a real item and cant be tagged
What normally happen when derived items are identified the previously unknown custodian is added by the legal team as a custodian to the case and data is collected from that custodian.
Hi Liam, thanks for the response, This question is not about attachments, this is strictly about threading and derived emails and how Predictive Coding handles the forward or previous portions of threads that would be exported and produced as part of an email when asked to “train” on a tag on said email. So, in other words, is PC is going to read only the single email portion that is tagged, or will PC analyze all the forwards and replies included with the portion of the thread that is being trained.
My thought is it will only train on the email (portion) that has been tagged and is being trained and will ignore the forwards on a given email. I need confirmation on this for a client.