W3C home > Mailing lists > Public > public-swisig@w3.org > December 2012

Libraries, linked open data and the SWISIG

From: Svensson, Lars <L.Svensson@dnb.de>
Date: Sat, 22 Dec 2012 08:15:49 +0000
To: "public-swisig@w3.org" <public-swisig@w3.org>
Message-ID: <24637769D123E644A105A0AF0E1F92EFA3FAD6CF@dnbf-ex1.AD.DDB.DE>
All,

I forward an off-list discussion I had with Paola, Roberto and Heiko. Starting point was my announcement of the workshop "User Interaction based on Library Linked Data" (UILLD 2013) [1]. Paola saw the overlap between our workshop and the special issue of the SWJ [2] and invited me to have papers from UILLD published in the special issue which I unfortunately had to decline since we plan to publish our proceeding through IFLA.

Short about me: I work in the IT department of the German National Library with a focus on semantic web, linked data and identifiers, investigating how libraries can benefit from semantic technologies and how others can benefit from our data. For those interested in more details, a good starting point is our paper from the last World Library and Information Conference on the German national bibliography as linked data [3,4].

On to the conversation (statements with an odd number of arrows are from Paola, the others from me)

>>> I am active because I am fedup with government corruption, 
>>> ineptitude and bad politics and I am waiting that the web is going 
>>> to change all that, seems that eGov is the right place to try to 
>>> make a difference and expose hypocrisy. Library for me fits with 
>>> eGov where eGovernance processes rely on well indexed, uptodate and 
>>> easy to find information to support decision making.
>>>
>>      What I can offer is expertise on how linked library data works 
>> (it's a bit special...) and we have staff at our place who are busy 
>> building linking platforms. Can you provide some use cases showing 
>> how library data could help?
>
> will think- at the moment all I can think is that
> a) library data is generally highly organized and designed to be 
> findable by librarians/users (although not without defect)

True (both statements, particularly about defects). Transforming the data for true semantic processing of course helps to reveal all defects there are, since e. g. reasoners are not very forgiving about improper/incomplete semantics. Together with tree colleagues I shall teach a course about semantic technologies in libraries in the upcoming term and we hope the students will give us additional ideas what can be done.

> b) linked library data, I am sure, adopts the best available practices 
> both from linked data and library data domains

My view always was (and still is) that linked data turns data silos into what librarians call authority files (reusable information about people, places and other interesting things) so from that point of view there is a large overlap.

> c) one of the challenges of governance, and eGov in general, is to 
> make well informed, possibly data driven decisions, yet the body of 
> knowledge that needs to be leveraged in this domain is fast and 
> interdisciplinary

Which essentially means that what libraries can offer is tools to organize the data and (possibly) provide access to it. The large caveat is that libraries collect _published_ resources (and that we are notoriously bad when it comes to journal articles) and that e. g. my library does _not_ collect research data but only the publications resulting and building on that data.

> d) for accessing and making governance and eGovernance knowledge 
> accessible during decision making, evaluating approaches from your 
> domain seems the right thing to do

I'd say that our approach currently is to get the data out there so that people can do something with it. So far I have not seen much re-use, but I hope that will change... If our data can help people make better decisions, control their government and prevent corruption, the ROI will be HUGE.

What the linked library data community desperately needs to go forward is practical examples of what people can use the open/linked library data for. There is a group working on how to represent library data using schema.org <http://schema.org>  and one large discussion point still is: what do we want/expect the search engines to do with the data (Help people find books on specific topics? Point them to a library in the neighborhood where they can borrow that book? ...).

[1] http://uilld2013.linkeddata.es/
[2] http://www.semantic-web-journal.net/blog/special-issue-call-semantic-web-interfaces
[3] http://nbn-resolving.de/urn:nbn:de:101-2012052306 
[4] http://de.slideshare.net/larsgsvensson/the-deutsche-nationalbibliografie-as-linked-open-data

All the best and season's greetings,

Lars
Received on Saturday, 22 December 2012 08:16:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 22 December 2012 08:16:20 GMT