From: Ken Masters <[log in to unmask]>
Date: Sat, 15 Feb 2014 15:20:46 +0400

HI Tony

I don't want to get into an argument, but your opening paragraph is
really troubling - and your use of the "rope" metaphor should be your
own warning to yourself of what my come down the line.  And you get
that extra rope because "they understand we need it as a commercial
enterprise" - justifying something because of finances is a really
rocky road.  Some of the worst human-rights abuses in the world
(including in your own country) have taken that route, so you probably
don't want to be associated with that line of reasoning.

I would strongly recommend that, for your own protection, you look
closely at what you gather, gather only what is required, keep it for
only as long as you absolutely need it, keep it secure (strongly
encrypted) and then destroy it. And justifying gathering and keeping
something purely because it fits in with a "policy", especially if you
would ever consider requesting an exception, will not stop the press
from having a field day.

Along those lines, you need to be very careful about what you define
as "personal information" (is an IP address considered personal?  or
private or not?) .  For example, you say that leaving your site and
going to Amazon is not personal data.  Really?  Just what policy at
your institution would allow you to track my purchase at, or even
visiting, another store?  By itself, it might not, but in combination
with my IP address, a pretty neat personal record can be created.  The
process of de-anonymisation (indeed, most intelligence gathering),
relies on gathering pieces of innocent information and combining them.

And even if you're not doing it, how securely is that information
stored, especially if it's not considered private?  What will your
legal defence be when a third party accesses that information,
combines it, and then puts in on the web?  I can tell you now that the
defence of "The IT guys are responsible" does not wash.  The person in
whose name that information is being gathered will have to take
ultimate responsibility.

And you're thinking that you'd like to go further, and you'd like to
know what books they purchased or searched for?  Take a gun to your
head - it will be easier.  Consider this scenario: you have a user
(id'd through his IP address), and your tracking system discovers that
this student of languages has suddenly taken an interest in books on
HIV/AIDS.  And he searches on phrases like "How to live with HIV"
None of that is personal information, so there's no need need to be
especially careful about storing that.  Mmm watch that blow up.

I know I err on the side on caution, but the alternative leads to a minefield.


Regards

Ken

Dr. Ken Masters
Asst. Professor: Medical Informatics
Medical Education Unit
College of Medicine & Health Sciences
Sultan Qaboos University
Sultanate of Oman
E-i-C: The Internet Journal of Medical Education


On 13 February 2014 23:59, LIBLICENSE <[log in to unmask]> wrote:
> From: Tony Sanfilippo <[log in to unmask]>
> Date: Thu, 13 Feb 2014 11:29:12 -0500
>
> As a data point, at the Penn State Press we do monitor some behavior,
> and use cookies, but I think the key is just how far you're willing to
> go in collecting this information, and if you've gone that far, is
> what you're collecting really that useful. Sure, we have to follow
> some University policies, but we do get a bit more rope and I think
> it's because they understand we need it as a commercial enterprise. We
> actually don't have to follow quite a few University rules, like their
> web style requirements, or some of their purchasing policies, and we
> don't have a privacy policy on our site, and have never been asked
> about it by a customer or the University.
>
> The University's privacy policy is interesting in that its only two
> rules that really impact us are: Don't collect personal information
> without permission, and don't sell any information you do collect. The
> actual policy is long (http://www.psu.edu/web-privacy-statement), but
> as for what really applies to us, that first paragraph seems most
> relevant. And thus far they've been pretty hands-off. If we wanted to
> appeal for an exemption to collecting personal information, we might
> get it, if we could define its scope and the purpose we'd have for it.
> But I'm not sure I have a good use for what we might collect, at least
> not more than what I might be able to infer from other sources, like
> the demographics of the marketing lead/mail/email lists that the
> peddlers of such information offer us, or what we might learn from
> print ad venues we've had success with, and their target demographics
> which they typically share with us, or the information that some
> scholarly societies offer in their conference marketing materials.
>
> Within the marketing and sales department we frequently ask ourselves
> about both the utility and the ethics of gathering some kinds of data.
> For example, we're already measuring traffic to our site using Google
> Analytics, and we're using Analytic tracking codes in the URLs we use
> in social media and email campaigns. And we count those who come to
> our site and then leave using our links to Amazon, or Powell's or
> other book sites, however none of that is personal data. But with a
> couple of slight modifications we could be collecting information on
> where these visitors live, their probable income, education level, or
> gender. But is it worth taking that ethical and bureaucratic leap just
> to refine the income level of a typical visitor? We already know a bit
> about their education level just by their interest in our books, and
> I'd like to think our products are for the most part gender-neutral,
> so what do we gain by collecting that data?
>
> Another issue to consider is the demographic of the typical university
> press marketer. Chances are they themselves are blocking cookies,
> and/or have JavaScript turned off, and/or use ad-blockers, so their
> interest in collecting this information is a bit colored by their own
> preferences and awareness of the issues involved. They probably find
> some of this tracking a bit creepy, and probably want to only do what
> they themselves are comfortable with, maybe a little bit more when
> pushed. But they don't like what they find on the web in terms of
> information gathering, and they don't want to be a part of that
> problem.
>
> All that said, there are some questions I'd like to know, if I could
> gather unlimited amounts of data about our web visitors and our
> direct-to-consumer customers. First, how many are librarians and from
> where? Next, what book pages do those with a low income look at and
> then leave without making a purchase? And finally, we know who left a
> book page and then went to Amazon, but we can't easily or confidently
> match that exit with a sale at Amazon, or any other number of
> associate programs we work with. We can make inferences, but I'd
> really like to know what happens after they leave. I'd like to see
> that kind of information, but I would want to seek the University's
> blessing before I collected it, and I'm not sure my educated guesses
> about those questions are so inaccurate that it's worth it, or that it
> would significantly change how we do things.
>
> Anyway, yes, I'd love more quality data, but I don't want or need to
> know everything. A lot of data isn't the same as useful data.
>
> Hope that's helpful,
> Tony Sanfilippo
> Penn State Press
>
>
>
> On Wed, Feb 12, 2014 at 10:57 AM, LIBLICENSE <[log in to unmask]> wrote:
>>
>> From: Joseph Esposito <[log in to unmask]>
>> Date: Wed, 12 Feb 2014 08:49:37 -0500
>>
>> Ken,
>>
>> No quarrel with anything you have here, but you are looking at this
>> from the point of view of the end-user, which is only part of the
>> equation.  What has caught my attention is that some organizations
>> (and I am thinking in particular of universities and university
>> presses) may be collecting data without knowing it or at least without
>> their staff knowing all the implications.  That's how I interpret Eric
>> Hellman's earlier comment.  I am still investigating this and would
>> certainly like to know if anyone can cite instances of tracking and
>> data collection by such organizations.
>>
>> Joe Esposito