I've got an upcoming project where there can be literally thousands of submissions at any one given time that need to be manually reviewed before publication (or rejection). So, there needs to be multiple moderators to handle the load.
Each moderator sees a list (in tabular UI format) of submissions, and all necessary info is right there on the list for a moderator to make a decision (approve/deny). Each item can be manually approved or denied one by one, or bulk-approved/bulk-denied by clicking on check boxes and then click either the approve selected or deny selected button.
If we've got multiple moderators looking at a lot of the same pending work at any one given time, there's a 100% guarantee that we'll get moderation conflicts. For example, a moderator may have already denied a particular item a mere 100 milliseconds before another moderator tries to approve it.
So, what are some usability guidelines that would apply in this situation to avoid moderation conflicts?
We can't do constant polling because that's computationally expensive, and it would update each moderator's page too fast anyway, making moderation practically physically impossible. It looks like we'll have to give them feedback on which of their approve/deny attempts went through and which ones did not (and provide a reason), but that, I would imagine, would be very annoying.
We're thinking that a moderator pulls up only unmoderated submissions at the instance he/she loads his/her list, and then immediately assign him/her as the moderator of the submissions he/she pulled? This way, nobody else can claim moderation for an item, thus greatly reducing (or even entirely getting rid of) moderation conflict?
Are there any other suggestions on how I should tackle this?