4 ms·
A potential solution to their problem is to establish a WP editing aberration detection mechanism, which would run quietly all the time and establish statistica
by saintx 14y ago
A potential solution to their problem is to establish a WP editing aberration detection mechanism, which would run quietly all the time and establish statistical baselines for editing frequency, and recognize when an unusually large number of edits from a geographical area or maybe CIDR block are coming in during a short period of time. This would help them automatically identify cases where a professor tells a couple of thousand students to make updates without supervision. Maybe correlate these with accounts or IPs that have few prior edits. This could then raise some sort of quality control flag on the edits and simplify the process of reverting the topics.
I shudder to think of what the Wikipedia-editing analogue of a global DDOS attack would be.
- rhizome 14y agoDoes any of that detection mechanism exist already?
- rohansingh 14y agoAs far as I'm aware, no. As far as I know, while IP's are logged, users with the "checkuser" privilege need to manually search and interpret the logs during each investigation.
- aaron695 14y agoMost of wiki is patrolled by robots, I'd image if it was useful it would. I'd see natural disasters possibly false positiving compared with turf wars between groups/countries after a flare up being the main reason you'd do it. I think professors asking students to edit would be rare enough not to make it worth while.