[open-bibliography] Sharing bibliographic user data

[Rufus Pollock]

On 4 November 2010 14:39, Tim Spalding <tim at librarything.com> wrote:
> Excuse me if this surpasses the purposes of this list, but I wanted to
> start a conversation about how to accomplish the sharing of user data
> related to bibliographic data, and keep it open.

This definitely does not surpass the remit of the list (at least IMO!)
and I think this is a *really* important topic. The 'user' data you
mention is part of the crucial 'value-added' information that enhances
the original dataset and makes being 'open' attractive.

> Basically, many institutions and sites are gathering user data
> relative to books—tags, reviews, lists, etc. How can that data be
> shared?
>
> Let's say, for example, that two libraries were collecting tags or
> reviews and wanted to share that data in an open way. How would they
> do it? What should the balance be between openness, privacy, and
> keeping users in control of their data? Is there an open license that
> requires you to refresh data, so a user can release their review but
> expect to be able to update it? Should institutions sharing tags
> include primary keys relative to users, or just submit total tag
> counts, etc.?

I think there are distinct issues here:

1. What data should people make public at all (or, conversely, what
data should people *not* make public because of e.g. privacy)

2. If data is made public should it be openly licensed?

My answer for (2) would be (almost always): 'Yes'

The answer for (1) must obviously be more nuanced.

In general, my suggestion would be that we can start with aggregate
data (e.g. tags and tag counts per book, usage (views/borrowing etc)
per book), where the resulting anonymization should mitigate any
privacy concerns.

Of course, one has to be cautious about what exactly aggregate data
involves: simple tag or usage counts per work should present no
problems but if you start doing intersections e.g. the number of
people who took out book X and book Y and book Z as this may allow for
indirect identification.

An obvious other point to make here is that it would be good idea if
services started asking users (when they sign up) if they are happy
for, say, all there tags and (separately) all their usage data to be
shared (on a per-user basis ). There may be many users who are quite
happy to share their data in this way.

Regards,

Rufus