The uptime monitoring scripts used in Lux/Emory Digital Collections currently executes several canned searches to check the application state.
Over time, this has had the result of artificially inflating views of the same works and Collections significantly. We currently have no way of reasonably filtering out these automated activities.
Current searches run:
"insulin"
"atlanta"
We should adjust the searches so that they contain a unique term such as "ThousandEyes" or ensure that we have a user agent that we can filter out.