I thought I'd put this here rather than a customer care ticket. I can do a ticket if you want.
When we do a 'segment comparison' project in analysis workspace and don't compare two defined segments but let analysis workspace create the 'everyone else' segment which is the inverse of the segment we created a couple of odd things happen:
1) the 'everyone else' segment that is created for the workspace run gets pushed to the cloud. We can see it in the cloud in the audiences section (everyone else in cloud.jpg). and the 2 shown have been 'gathering data for over 2 weeks.(everyone else in cloud.jpg)
this means that we have multiple (one industrious user compared over 100 segments to the inverse segments so we got 100 segments called 'everyone else' in the cloud, AAM and Target.
I think this is a bug. Especially since in AAM there is an integration code saying it came from analytics so we can't even see how the trait/segment is defined. Unfortunately some of the segments being compared to 'everyone else' were duplicates created by several different users so all the names/descriptions are the same. but the segment definitions in analytics are different: https://forums.adobe.com/api/core/v3/attachments/886139/data
in any event having analysis workspace not clean up after itself is a bug that really needs to get fixed. Not only does having random traits/segments created in the cloud give us a maintenance/governance headache they can potentially cost us a lot of revenue, resources and money if things get defined incorrectly and used in AAM or Target.
Please have Analysis Workspace clean up after itself.
The ticket 160919-000201 is still in 'waiting engineering' status.
I have created a new segment comparison, gotten an 'everyone else' segment created during the analysis and it looks like that segment didn't make it into the cloud profile & audiences section and no new ones have shown up in audience manager so it looks like the cause has been fixed. However, we still haven't been told we can delete the duplicated 'everyone else' segments either the ones that are similar but have the same name and the ones that have duplicate data and have the same name.