LOD

Linked Open Data -- for local (i.e., not national) digital collections

Background


Cleveland Memory Project and other memory projects (UA, AM) need to be merged. They are thinking about going through a commercial system, but that will create another silo. So, Linked Open Data (LOD) would be a better idea.

Goals:

Provide a road map for the local (i.e., not national) digital collections
Note: There are small collections here and there, they are in silos, and they themselves have no capability to leverage to a higher level, e.g. linking/ together in the LOD cloud.
Identify the best available open source technologies, like using Drupal, to do the human-user services interface.
Research on user experience for such services (including contents and interface)
Exam what outside data in the LOD cloud can be used by these digital collections, to enrich the user experience when people look for something through here. (currently only images and descriptions are informational).

Become a good complete road map for digital collections for cities, township, etc. in the U.S.

Topics


  • 1, Turning the current data into linked data, -- how to do it, and what gaps need to be filled.


An example

one example we will be using is Cleveland Union Terminal”, which looks like:
http://images.ulib.csuohio.edu/cdm4/item_viewer.php?CISOROOT=/print&CISOPTR=286&CISOBOX=1&REC=1



  1. A) Sorting out the types of ‘things’ that are involved in the digital collections. This will be in very high level, e.g., buildings, events, and... Their images captured.
  2. -- Note, currently the records are ‘aggregates’ which mixed place, the publication and publisher, and the digital images together. We will need to draw some data models to see their relationships in order to avoid duplication and inconsistency.
  3. B) Mapping the metadata elements to the available vocabulary (e.g., Dublin Core, FOAF) in the LOD.
  4. --Note, once the ‘things’ are put into the data model, we can use different vocabularies to represent place, architecture, agents (person, corporate body, family), etc.
  5. C) As we discussed before, there will be also some ‘things’ that will need to be put together in an ‘authority’ data format.
  6. -- Note, e.g., the “Cleveland Union Terminal”, designed by “Graham, Anderson, Probst, and White”, built by “the Van Sweringen brothers”, facing the “Public Square”, and became “Tower City Center”). Those proper names, if not in the LCSH or TGN, have to be established and given a unique URI (cool URI maybe useful.




  • 2, Mapping to outside authority data in the cloud (e.g., through subject headings, thesaurus terms, dbpedia entries, etc.)
  • 3, Linking to appropriate data in the cloud, especially museum, historical societies, media, government data.
  • 4. Using the best available open source technologies, like Drupal, to design meaningful human-user interface.
  • 5. Recommendations for human-use interface based on user experience.
  • 6. Exam what outside data in the LOD cloud can be used by these digital collections, to enrich the user experience when people look for something through here. (currently only images and descriptions are informational).
  • 7. The business model or workflow, reusing existing data as much as possible.
  • 8. Sustaining and promoting all products to educational and society use.