From Library Linked Data
Potential XG topics
Please add your name to indicate interest in a topic
AI: the following two points are more Semantic Web representation-oriented. I feel they could be mentioned even if linked data publication was not involved (but of course they are still crucial!). Note that I won't defend much keeping the two main points separate: they could well be merged. I prefer to consider this after redundancy between the more detailed points is removed, and for this I prefer to wait for you ;-).
1. legacy/traditional vocabularies and linked data [antoine,kcoyle]
- Inventory of vocabularies published. Emerging consensus about which to use? [jyoung4, gneher, wwaites2, rsinger, AMicsik, dlukose, emmanuelle]
- Status of library-related vocabularies in development, e.g., RDA, FRAD and FRBR [jyoung4, wwaites2, GordonD, rsinger, dlukose, emmanuelle]
- Translation of data in MARC format to linked data [wwaites2, GordonD, rsinger, dlukose, antoine, kcoyle, StuWeibel, emmanuelle]
2. knowledge representation issues / Describing library and museum authorities and KOS resources as Linked Data [jyoung4, GordonD, dlukose, antoine, marcia]
- Using SKOS to describe knowledge organization systems (KOS) that have been used by library, archive, and museum communities (LCSH, RAMEAU, AAT, RGN, TGM I and II, DDC, etc.) [marcia, bvatant, ndelaforge, keckert, GordonD, AMicsik, dlukose]
- Persons, families, corporate bodies, works [ahaffner, ndelaforge, emmanuelle]
- Check if SKOS extensions are needed for describing particular types of KOS (term list, name authority file (not limited to agents and works), digital gazetteer, list of subject headings, taxonomy, thesauri, classification, etc.) and provide SKOSified KOS examples. [marcia, keckert]
- Dilemma between skos:concept and foaf:person (rda:person, frbr:person, frad:person) for person authorities [jyoung4, ahaffner, bvatant, ndelaforge, keckert, wwaites2, rsinger, marma, emmanuelle]
- FRBR group 1 entities [jyoung4, ahaffner, Jschneid4, wwaites2, kcoyle], frbr entities in general [StuWeibel] AI: this may be moved to the point above on legacy/traditional approaches. The point above could be then reformulated to emphasize that SKOS is not the only proposal for authority/KOS data.
- Applying FRBR, FRAD, RDA [jyoung4, ahaffner, Jschneid4, wwaites2, emmanuelle] AI: this may be moved to the point above on legacy/traditional approaches
- Check if and how to use SKOS to describe integrated KOS or mapping results. [marcia, keckert]
- Confirm the use of SKOS for multilingual KOS vocabularies. [marcia, bvatant, keckert]
- Conversion issues, e.g., URIs, content negotiation, RDF compatibility [jyoung4, ndelaforge, rsinger, dlukose, emmanuelle] AI: some of it may be positioned in identifiers (URIs) and linked data publication (content negotiation)
- Discuss how to reflect changes in a concept scheme in linked data set(s). What are new challenges and approaches? [marcia, bvatant, keckert]
- How to deal with obsolete concepts? [bvatant, ndelaforge, keckert, wwaites2] AI: relates to "versioning, updates", but could be orthogonal
- Types of library data other than bibliographic and authority, e.g., circulation, statistics (COUNTER, SUSHI), full-text sources (e.g. DOAJ, Project Gutenberg, etc) [Jschneid4, keckert]
- "Application profiles"
- Specifying/documenting the content of metadata for use by other information producers and by information consumers [Tom, ahaffner, Jschneid4, keckert, kcoyle] AI: I'm really not sure whether this fits here...
AI: I feel that the following three points are really specific to how libraries should behave in a linked data environment, that is, one where the data is being really 'networked'.
3. What obstacles do libraries, museums, archives, publishers, currently face in publishing and managing Linked Data? How can the obstacles be removed? What role for libraries? [jyoung4, ndelaforge, digikim, GordonD, rsinger, dlukose, StuWeibel, emmanuelle]
- Available pools of linked data (inventory): quality, granularity, link integrity, vocabulary usage [gneher, ahaffner, rsinger, antoine, kcoyle]
- linked data management, hosting, and preservation [digikim, GordonD, kcoyle, emmanuelle]
- vocabulary-specific aspects of management, hosting, and presentation [digikim, GordonD,kcoyle]
- Versioning, updates [bvatant, rsinger, antoine, emmanuelle]
- Issues of Web architecture, e.g., persistent URI design best practices, HTTP [jyoung4, ndelaforge, keckert, digikim]
- Dissemination mechanisms: RDF schemas, RDFa, bulk download, feeds, SPARQL... [jyoung4, keckert, digikim, dlukose, antoine]
- Licenses, IP, DRM, other availability/rights/access restriction info [antoine, kcoyle, emmanuelle]
4. Use of Identifiers [GordonD, dlukose, StuWeibel, emmanuelle]
- HTTP URIs, DOIs, handles, ARKs, shorteners, hash, slash, 303 redirects, PURLs [jyoung4, ndelaforge, keckert, digikim, StuWeibel, emmanuelle]
- Reuse or urlification of traditional identifiers [jyoung4, StuWeibel]
- Identifiers for properties and classes, concepts, and "real world things" [bvatant, keckert, StuWeibel]
- Namespace policies [keckert, StuWeibel]
- Identifiers for libraries? ISIL, possibly urn:isil: namespace [StuWeibel]
5. Linking across datasets [digikim, dlukose, antoine]
- How much linking is there? What links to what? [Jschneid4, GordonD]
- Alignment (cross-linking) of vocabularies [GordonD, AMicsik, dlukose, antoine]
- Interoperability, "mapping frameworks", etc
- Alignment of element sets [ahaffner] AI: I'm not sure I get that one…
- Alignment of authorities [jyoung4]
- Alignment of real-world-resource identifiers [bvatant]
AI: the following may go to legacy vocabularies (which would then become 'legacy vocs and approaches), but I feel it could be part of the linked data-specific issues too (and maybe that would be better).
6. The Linked Data paradigm and the Metadata Record paradigm [GordonD, AMicsik, dlukose, emmanuelle]
- Models for packaging Linked Data in records, e.g., Named Graphs [jyoung4, keckert, StuWeibel, emmanuelle]
- Provenance of Linked Data [keckert, wwaites2, rsinger, emmanuelle]
AI: I don't know where that one would go. Maybe to the linked data cluster above...
7. Implementation issues
- REST patterns for Linked Data [jyoung4, ndelaforge, keckert, digikim, AMicsik]
- Caching strategies [ndelaforge, keckert, digikim]
AI: the 2 points on 'strategic guidance' below may belong to the documentation item below, but I feel it would be interesting to have feedback on thes for our use case, so they could give a category for the template. This is different from other community-building issues, below, I feel.
8. 'strategic guidance'
- Workflows or roadmaps for different kinds Linked Data projects [keckert, emmanuelle]
- Examples of business models of managing linked library resources (metadata, vocabulary, and KOS resources) [digikim]
AI: the next issues are really interesting for the group, but I think we have to keep them apart, not to interfere with the ones that are really interesting for the use case and case studies.
9. Community-building, education and outreach
- Need for training and documentation (a Linked Data primer for libraries ?) [gneher, Jschneid4, keckert, digikim, antoine, emmanuelle]
- Liaison with standardisation bodies and initiatives (ISO and national bodies, IFLA, International Council on Archives, CIDOC...) [GordonD, emmanuelle]
- Outreach to other communities (archives, museums, publishers, the Web) [Jschneid4, GordonD, antoine]
10. Next steps
- Future working groups? In W3C, in library world, or elsewhere? [digikim, StuWeibel, emmanuelle]
- Set up other groups to define best practices - what is needed to make that doable for other groups with more time/broader scope?
- Common patterns in Linked Data, with examples, and with best practices for "Linked Data friendly" output from traditional library data - to provide guidance and save time - maybe several best practices when there are several good ways to solve a problem. Antoine: isn't related to "education" above?
- Publishing citations, e.g. best practice for citations in RDF http://www.semanticoverflow.com/questions/963/best-vocabulary-for-converting-bibtex-to-rdf/971
- Best practice for the description of personal names and name variants Antoine: see "Persons, families, corporate bodies, works" above
- Characteristics of a common model that embraces the legacy of MARC data and the abstractions of FRBR, and makes evident how such a model could be grown coherently to include the varieties of other metadata relevant to the community?
Who is who?
- antoine = Antoine Isaac, VUA & Europeana, Netherlands
- bvatant = Bernard Vatant, Mondeca
- digikim = Kim Viljanen, Aalto University, Finland
- dlukose = Dickson Lukose, MIMOS BHD, Malaysia
- emmanuelle = Emmanuelle Bermes, BnF
- JSchneid4 = Jodi Schneider, DERI Galway
- jyoung4 = Jeff Young, OCLC Research
- kcoyle = Karen Coyle
- marcia = Marcia Zeng, Kent State U. US
- rsinger = Ross Singer, Talis
- ndelaforge = Nicolas Delaforge, INRIA Sophia-Antipolis France