Admin email notification when percentage of open access publications changes

[EricDurante]

This depends on https://github.com/psu-stewardship/researcher-metadata/issues/123

Each time we record new stats for the total number of publications vs. the number of publications that are open access, we should compare the percentage of open access publications to some past statistics, and if the change in the percentage is greater than some threshold value, we should send an email notification to administrators so that they can investigate the reason for the change.

Questions:

1. To what past data should we compare the current percentage? Should we just compare to the last snapshot? What if the notification threshold is, for example, -5% change and the actual change has been -4% from each of the last 5 snapshots to the next? (Maybe this depends on how often snapshots are taken.)
2. What exactly is the threshold value for notification?
3. Should we notify about both positive change and negative change?
4. To what email address(es) should the notification be sent?

[anaelizabethenriquez]

Some answers if we were to do this soon:

1. The snapshots are weekly now; I'd suggest comparing to the last four snapshots.
2. Looking at the data, I'd say 1 point of change in OA percentage would be an appropriate threshold. The only changes of that magnitude that I see since weekly snapshots began in April 2021 are July 23, 2021 (likely due to reminder emails sent then) and November 12, 2021 (in the middle of development).
3. I would notify about both positive and negative change.
4. Send notifications to openaccess@psu.edu. The team monitoring that (mostly me) could determine whether it requires a larger conversation.