Talk:Cluster Digital Objects

From Library Linked Data
Revision as of 15:14, 15 February 2011 by Jschneid4 (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Old todo's

  • WHO? Topics/agenda for meeting
  • ALL Carefully read the use cases in our cluster -- all
  • Split this up? All?/ Examine the unclaimed use cases to see if they belong with us
  • Split this up? How? /Look at all use cases to see if they're relevant to digital objects, in case we want to share them
  • WHO Clarify with William about the Scholarly Debate UC--is there more data than the record examples such as

http://bibliographica.org/graph?uri=http://bibliographica.org/aggregate/301b111e-0dc0-5e34-a5e6-06c461d51789/37561/work http://bibliographica.org/graph?uri=http://bibliographica.org/aggregate/301b111e-0dc0-5e34-a5e6-06c461d51789/32635/work <http://bibliographica.org/graph?uri=http://bibliographica.org/aggregate/301b111e-0dc0-5e34-a5e6-06c461d51789/37561/work>and the searches that show Works and Expressions. -- who?

Groups or Collections of Objects?

Note by Mark: are we going to call them "groups of objects" or "collections of objects"? I'd say the former, as collection for me refers to metadata record collections. You see that also in the text now when referring to the NLL UC as an "existing collection"

I take collection to mean something different -- the collection of the objects themselves. But we can change this if you feel strongly. Jodi Schneider 16:12, 13 January 2011 (UTC)
I think either works. I'd use collections myself, but, like Jodi, don't mind turning it all into groups. Mark, if you want to, go ahead and change that. Asaf Bartov 19:17, 15 January 2011 (UTC)

Cluster Admin (moved from page)

Next meeting

  • Skype Thursday 2010-12-2 - CANCELLED
  • Skype 20:45 UTC Wednesday 2010-11-03
  • Skype 14:00 UTC Friday 2010-11-26

Agenda

  • ...

Useful Links

UseCases

Email with Mark's initial mapping of UCs - goals

Pending tasks

  1. check with the UC shepherds if their UC is complete and correct.
  2. develop the cluster page

Completed tasks

Overall goals

  • Homogenize the texts -- use the same wording for the same goals
    • add/rename standard goals
    • Note--is there some standard? good example?
  • Make sure that there are links to the schemas ("vocabularies" in this group's parlance)
    • Note--let's talk about whether schemas and vocabs are the same here
  • Make sure that there are links to (example) records.

Minutes 2010-11-03

  • What are we trying to do?
    • Find overlap in use cases--since we can't put all 40+ into an end document. Do this starting with goals.
    • see mailing list post
    • May help to get actual schemas and records to see how they are being used
  • to-do:
    • find shepherds for all our use cases (Mark)
    • add inline references to dimensions/topics in UCs
  • Discussion of Goals
    • URIS: provide URIs for entities in the UC [INSERT ENTITY NAMES]
    • RELATIONS: represent relationships between the entities that already existed in the data [INSERT RELATION TYPES]
    • RELATIONS-NEW: add relationships between the entities that did not exist yet in the original data. CONSIDERING NAME CHANGE, E.g. to CREATE-RELATIONS or ASSOCIATION
    • PUBLISH: publish the data as {RDF, RDFa, HTML, ...}
    • SEARCH/BROWSE: provide search and browse facilities over the RDF data
    • REUSE-SCHEMAS: reuse of existing schemas {FOAF, ORE, ...} in publishing the data as RDF
    • REUSE-VOCABS: reuse existing vocabularies already published as LOD {LCSH, GTAA, ...} in publishing the data as RDF, i.e. replacing internal identifiers with URIs from these vocabularies.
    • (LINK-VOCABS: link the published data to other LOD vocabulaires, i.e. mappings.)
    • MAPPING: create equivalence relationships (owl:sameAs, skos:closeMatch) between value vocabularies or data items.
    • API: provide an API to access / change the data.
    • DISCOVER/SUGGEST: use existing links to discover/suggest other items.

Minutes 2010-11-26

  • Discussion of existing work by other clusters
  • ACTION: Mark to solicit more info from Emma about how to improve goals; waiting for her suggestion of goals based on Emma/Karen group
  • Add a qualifiers field - RELATE (existing, aggregations) (Jodi to replace)
  • Discussion of how to distinguish LCSH, AAT,... from FRAD, FOAF, ...
    • LCSH, AAT,... - list: thesauri, glossary, ...; domain vocabularies
    • FRAD, FOAF, ... - conceptual models, RDF vocabularies, value vocabularies, element vocabularies; modelling vocabularies, ontology
    • "people will abbreviate"
    • may need to further distinguish conceptual models (FRAD, FRBR as ontologies/modelling vocabularies vs. as conceptual models)
  • Discussion of concrete use cases - see sample from Emma, Mark to continue work on the example he started
  • Discussion about Emma's suggestion to create abstract scenario(s) for each cluster
  • Asaf was at Interedition, talked to people about digital editions from manuscripts TEI. Heavily marked up digital editions of manuscripts, which represent even deletions, crossed-out, margins, and the different scholars' views. -- is this in scope?
  • Mark suggests as related Knowledge Engineering for Historians on the Example of the Catalogus Professorum Lipsiensis, which was ISWC best in-use paper
  • Next meeting Thursday 4:15 UTC 2010-12-02, Mark to write agenda

ACTIONs:

  • Mark: contact emma about goals, reply to Antoine about UC P20, write up smthg about terminology to list, draft a scenario text, send ISWC report to Jodi and Asaf, make agenda for next week,
  • Jodi: rewrite goals page wrt qualifiers
  • Asaf: draft Interedition UC
  • All: add goals in UCs

Minutes 2010-12-02

  • ...

Scenario

We need to come up with one or more scenarios like those of the Image Annotation XG that cover our use cases. Attempt by Mark:

Suppose a group of people collects several digital objects that somehow belong together. Digital objects can be anything, ranging from PDF files of scientific papers to JPEG images and MPEG4 movie files. There are different examples of cases where people want to group digital objects. One is a group of researchers that write papers which have images, datasets and project partners associated with them. Another is newspaper articles that are grouped together in a dossier. Users of the system would like to represent this grouping of objects in digital form. Some of the objects will not be within their own control, but published on the Web by other parties. In some cases there are more detailed relationships between particular objects, e.g. that a book discusses a similar topic as another book, or was written by a particular author who is described by an authority file available elsewhere on the Web. Besides representing all this information, the users want to be able to conveniently search and browse this data, and be supported in discovering new relationships between objects. All the information should be published on the Web in a way that allows others to browse and process the objects and groups.

Notes by Mark:

  • it's really hard to write the text in a way that it covers all our UCs (i.e. agnostic on type of resource being collected)
  • why again are we using the term "Digital Objects"? Are the resources this cluster handles any different from the resources in other clusters? Shouldn't we call them "resources" or does "Digital objects" have a well-understood meaning in the library world. But even then, this cluster is not about resources, it's about aggregation of resources. ANSWER: "Digital 0bjects" has well-understood meaning in Library world.