Friday, March 04, 2011

Open Data I

The idea of open data has gone from an extremist rallying cry to a mainstream movement. In the next few posts I'll highlight just an iceberg tip's worth here, but expect to see more about this every day that passes.

The UK's educational research arm, JISC (something like NSF but more for education rather than pure science) and the research libraries' organization, RLUK, undertook a study about the advantages and possibilities afforded by opening data from libraries, archives and museums. They have produced the Open Bibliographic Data Guide, which investigates the business case for providing bibliographic data that can be re-used.

This is a practical, not a utopian, vision of open data.
"In earlier times, observers may have considered the ‘open data movement’ as the preserve of a certain type of fanaticism also associated with Open Source Software (OSS) and Open Content, emotionally and ideologically linked to the spirit of 1969.

However, OSS and Open Content have now morphed in to propositions with clear business cases of interest to corporations, institutions and governments. National strategies and Chief Information Officers espouse Open Source Software for financial and business benefit, whilst academic leaders are supporting Open Access Journals and Open Educational Resources (OER)."(link)
The report gives 17 different use cases -- situations in which an institution might want to provide its data with some degree of openness.

1 – Publish data for unspecified use
2 – Publish open linked data for unspecified use
3 – Supply data for Physical Union
4 – Allow Physical Union Catalogue to publish data
5 – Expose data for federation into Virtual Union Catalogue
6 – Publish grey literature data
7 – Contribute data to Google Scholar
8 – Publish activity data
9 – Supply holdings data for Collection
10 – Expose holdings / availability data for Closest Copy location
11 – Share data for Collaborative Cataloguing
12 – Supply data for Crowd Sourced Cataloguing
13 – Supply data to be enhanced for own
14 – Publish data for LIS research
15 – Allow personal use of data for Reference Management
16 – Publish data for lightweight application development
17 – Allow commercial use of data in mobile application

For each of the cases the report discusses pros and cons for the institution, its users, and the world, as well as the business case for making ones' data open. They acknowledge the complexity of our current environment of bibliographic data ownership:
"Our problems with bibliographic metadata are quite specific:
  • Non-profit and commercial players have built businesses around datasets of MARC records, indexing / TOC services and journal Knowledge Bases – but what is original about those accumulations?
  • Bibliographic records in the circulation amongst libraries are of uncertain and complex provenance, with the exceptions of those explicitly tagged by a ‘vendor’ or exclusive to a special collection" (link)
JISC doesn't stop at this report but is sponsoring projects and ongoing activities in this area. Already the British Library has produced its British National Bibliography data openly for reuse. You can keep up with these activities through the newsletter (subscribe here) whose logo reads: One to many; Many to one: Towards a virtuous flow of library, archival and museum data.

No comments: