[ckan-discuss] Fwd: Some Feedback on CKAN

Tue Mar 1 12:58:19 GMT 2011

Some really useful feedback from Mark Hahnel of Science 3.0 / FigShare
with whom I had a very good chat last week. (Mark tried sending direct
but the email kept disappearing -- if anyone else encounters this
please let me know).

One of the key things I got from chatting with Mark was the importance
of providing ultra-simple mini-gateways into a CKAN instance oriented
to a specific community.

I'm wondering if we can work on groups to deliver this sort of feature
-- groups already provide a way for a given group / community of
people to have a bespoke web page and list of datasets but we could
add to this, for example by providing a default search box that just
searches a groups datasets.

Rufus

---------- Forwarded message ----------
From: Hahnel, Mark
Sent: 28 February 2011 18:55
To: ckan-discuss at lists.okfn.org
Subject: Feedback

Dear CKAN team,

I am writing to offer some feedback with regards to how I view
http://ckan.net/ -This is a follow up to a conversation I had with
Rufus and he suggested I should pass these opinions on to the whole
group.

As a practising scientist, the front end of the site does not make
uploading data easy for me. In that, I’m a little confused as to who
the site is for, can I use it? Ie. There is no mention of what types
of data are being uploaded and whether scientific data is suitable. I
only find science related stuff if I search ‘science’, so maybe
something along the lines of, “What subject are you looking for data
on?” is needed.

We spoke with reference to http://FigShare.com which is offering a
service with a lot of overlap, it’s still being developed as a micro
publishing platform for scientists to share all of their data,
including the rarely published/still important negative data. So the
target audience here is your average researcher. Hopefully we’ll
generate a good dataset which can go into CKAN.

But my point here is the average researcher, the people from my field
with the data needs to be spoon fed into uploading data. They’d need
to see something saying ‘scientists click here’...now what...’upload
your data or search the available data’. And then when they do want to
upload they/we again need it simple. Is all the meta data necessary,
could that not be optional following a simple ‘addition of a dataset’?
I suppose it depends what the overall goal is. If you want linked data
in a standardised form, then I understand. But personally I feel
acquisition of data in any form should be the priority. The power of
the platform lies in the API, and the subsequent analysis of the data.

When it comes to data uploading, the average researcher would not know
which license to put their data under and they do not have the time to
research which one it should be. Especially not all 75 that you have
an option of on the CKAN data upload form. Due to the legal
connotations of a ‘license’ I think this question might be enough to
put a lot of my colleagues off and this could be the point that their
data upload ends. I don’t even know what the ‘extras’ are for and I
like to think that I’m quite well versed in this kind of thing.

My second point is incentives. I know about the benefits of open
access etc but a lot of people do not, again it needs to be spoon fed,
“WHY should I do this” – Maybe even have a tab saying this. All we get
is “CKAN makes it easy to find, share and reuse open content and data,
especially in ways that are machine automatable.” – Is this a good
thing? How can I personally benefit from spending 5 minutes of my
precious time doing this?

I should note that Rufus has since made a few changes which may have
corrected some of the things I’m referring to. I should also note that
I am aware that this is just the viewpoint from one angle and that the
site is for many types of data, each of which will have different
specs. These are not complaints, I love what OKFN and CKAN are doing,
I’m just trying to help.

Thanks for your time,

Mark Hahnel.