This report is very interesting, but vague on the details. For
instance, I don't quite understand how subscription-based
resources (whose access is limited to a closed community) are
susceptible to Internet robots, automated-processes, crawlers,
and spam-bots (RACS). What resource(s) were they testing and
whose usage accounts were they monitoring? If the intention of
bepress is to establish a new standard of counting, they have to
be a little clearer about how they filter.
Individual publisher platforms employ their own mechanisms when a
threshold of article requests per IP address per unit time has
been breached. Some systems temporarily block IP addresses; some
permanently block suspecting IP addresses waiting for a human
being to contact the publisher to appeal the block. Some
publisher platforms do absolutely nothing. I've successfully
downloaded my own article thousands of times from one reputable
publisher without setting off any alarms.
As I've argued before, COUNTER statistics are internally valid
and can be a good guide to compare articles from the same
publisher. They fail to be externally valid for comparing titles
across publisher platforms [1]. Download statistics (unlike
citations) are non-transparent -- you cannot track down and
verify whether the statistical reports you are receiving in your
library are truthful representations of what actually transpired.
Any economy that is based on non-transparency, blind trust, and
few (if any) consequences for unethical behavior is wide open to
abuse. When cancellation decisions ride solely on measures of
article downloads, librarians should approach these figures with
a healthy degree of caution.
[1] Davis, P. M., & Price, J. S. (2006). eJournal interface can
influence usage statistics: implications for libraries,
publishers, and Project COUNTER. Journal of the American Society
for Information Science and Technology, 57(9), 1243-1248.
http://arxiv.org/abs/cs.IR/0602060