[p2pu-dev] DB Dump for Datamining p2pu

॥ स्वक्ष ॥ vid at svaksha.com
Sat Jan 29 05:27:54 UTC 2011


On Fri, Jan 28, 2011 at 23:46, Philipp Schmidt <phi.schmidt at gmail.com> wrote:
> On 28 January 2011 22:53, Niels Sprong <nielssprong at gmail.com> wrote:
>> I understand the need for a sort of code of conduct to deal with data.
>>
>> I don't really understand the legal issues. ('sensitive' user data?)
>
> the legal issues are relatively minor compared to the danger of our
> community *perceiving* us sharing any data as a violation of their
> privacy. legal issues are mostly related to us not sharing user data
> with third parties for commercial reasons. and p2pu users share much
> information that would be considered personal, already on our site -
> so the database would just be offering a different lens on that
> information.

Agree. Privacy is a concern and the differentiator could be in how
data is analysed -- the difference between "how many $users took Ruby
courses?" and "was the Ruby course more popular than the php one?" as
the latter can show patterns without revealing individual preferences.

The perception implications vis-a-vis the community (even if data is
*not* shared with third parties) matters a lot. Would it help if we
had a small note mentioning this and making this a user opt-in
feature? Most times allowing users to choose makes matters easier.

At this point, it would be interesting if the educators/researchers
pitched in with the kind of patterns they seek and discussed their
thoughts. I was told that there has been a discussion, but since I'm
new, could someone point me to the archives please.

ciao,
-- 
Regards,
vid ॥ http://svaksha.com


More information about the p2pu-dev mailing list