- From: Karen Coyle <kcoyle@kcoyle.net>
- Date: Wed, 27 Apr 2011 09:56:25 -0700
- To: public-lld@w3.org
Owen, it sounds to me like you have outlined some of the issues that need to go into a cost/benefit analysis! I'll see if I can get this into the wording on the recommendations page without it getting to be too specific or too long. kc Quoting Owen Stephens <owen@ostephens.com>: > Thinking about this... > > I haven't had time to look at the latest version of the 'benefits' but > presumably this is complimentary to the 'costs'? > I think that the benefits of expressing 'traditional' library data as LD is > more than just discovery (and indeed, many 'discovery' benefits might be > realised without doing full LD representations) - so aspects of data re-use, > machine processable data etc. is part of this. > > It might be worth exploring this question of native vs converted LD further. > I'm thinking there are probably different sets of costs. Doing basic LD > representations from MARC should be straightforward once a few have > trail-blazed and shared conversion routines etc.? However, cost of matching > entities etc is high. > > Some of these costs go away when you create library data natively as LD - > but the associated costs of doing this feel v high (redeveloping systems to > cope etc.!) > > Owen > > On Wed, Apr 27, 2011 at 3:50 PM, Karen Coyle <kcoyle@kcoyle.net> wrote: > >> Owen, good point. So perhaps there needs to be something about discovery >> services ... and moving library data initially onto the SemWeb for >> discovery? >> >> The trick that I see to the idea that LD will be built on top of current >> systems (which I agree is likely to be the first step) is the question of >> how truly useful LD will be created if we continue to create MARC records in >> our systems. MARC can't handle any of the WEMI relationships ("is >> translation of") and so these would have to be coded outside of the library >> cataloging systems. This implies that we will actually be re-creating our >> data in an LD space, something that many will see as duplicate work. >> >> kc >> >> >> Quoting Owen Stephens <owen@ostephens.com>: >> >> I don't have an issue with what is proposed here, but it feels like it >>> addresses the most extreme vision of library adopting linked data >>> practices >>> - that is baking LD into library data at a low level in terms of metadata >>> practices. >>> >>> However, most of the activity I'm aware of with LLD is taking existing >>> library data and expressing it as LD. I feel that there should be an >>> analysis of the cost/benefit of this approach as well - winning the >>> argument >>> 'start creating library data as linked data' feels too hard to win >>> outright >>> and it may be a matter of justifying some incremental steps first? >>> >>> Owen >>> >>> On Tue, Apr 26, 2011 at 6:32 PM, Karen Coyle <kcoyle@kcoyle.net> wrote: >>> >>> The recommendations are broken into three major sections, one of which >>>> speaks to management actions that are needed. Of these, the first >>>> addresses >>>> costs and return on investment. The group felt that without some kind of >>>> cost benefit, it would be hard to make a good case for LLD. At the same >>>> time, we don't really have a good grasp of what our current practices >>>> cost >>>> us. Thuus: >>>> >>>> *Identify costs of current practices, and costs and ROI to moving of LLD* >>>> >>>> There must be some measurement of the relative costs of current library >>>> data practices and the potential of Linked Data to aid in making >>>> decisions >>>> about the future of library data. There are various areas of library >>>> metadata practices that could be studied, either separately or together. >>>> Among these are: >>>> >>>> * The relative costs of the Record v. statement approach: for editing >>>> by >>>> humans, as full record replacement in systems, and the capabilities for >>>> sharing >>>> * The use of text versus identifiers approach has costs: actual records >>>> must change when displays change (Cookery to Cooking); international >>>> cooperation requires extensive language mapping processes; some needed >>>> data >>>> elements must be extracted from textual field using algorithms, which >>>> also >>>> hinders sharing; and some library data formats require catalogers to >>>> duplicate information in the record, providing both textual fields and >>>> coded >>>> data for same information. >>>> >>>> --- >>>> >>>> Again, comments, suggestions, alternate ideas are all welcome. >>>> >>>> -- >>>> Karen Coyle >>>> kcoyle@kcoyle.net http://kcoyle.net >>>> ph: 1-510-540-7596 >>>> m: 1-510-435-8234 >>>> skype: kcoylenet >>>> >>>> >>>> >>>> >>> >>> -- >>> Owen Stephens >>> Owen Stephens Consulting >>> Web: http://www.ostephens.com >>> Email: owen@ostephens.com >>> >>> >> >> >> -- >> Karen Coyle >> kcoyle@kcoyle.net http://kcoyle.net >> ph: 1-510-540-7596 >> m: 1-510-435-8234 >> skype: kcoylenet >> >> > > > -- > Owen Stephens > Owen Stephens Consulting > Web: http://www.ostephens.com > Email: owen@ostephens.com > -- Karen Coyle kcoyle@kcoyle.net http://kcoyle.net ph: 1-510-540-7596 m: 1-510-435-8234 skype: kcoylenet
Received on Wednesday, 27 April 2011 16:56:55 UTC