Planning | Description | Requirements | UI Mockup |  Triples Examples | Queries Examples


Example story: As a faculty member or librarian, I want to create a virtual collection or exhibit containing information resources from multiple collections across multiple universities either by direct selection or by a set of resource characteristics, so that I can share a focused collection with a <class, set of researchers, set of students in a disciplinary area>.

This use case is focused on individual creation of a set of resources – a virtual collection or exhibit – along with descriptive information for the collection and possible arrangement and annotation of resources in the collection. The base use case is that any virtual collection will be publicly accessible and thus authentication is required only for user creating the collection (and not for users of the collection). The inter-institutional aspect of this use case is not supported by current systems. It is expected that the collection URI will support both human and LD views.


Out of scope: Discovery of virtual collections. Annotations of a single item (Rob: should that be a separate use case? Simeon: would need to understand discovery.)     

Potential Demonstrations

A. Faculty member constructs a reading list to share with a set of students: Faculty member at one institution browses local discovery system, finds item of interest for a class and creates a new virtual collection with this one item (where is link? when do they log in?). As part of creation process they give the collection a name, add a description and perhaps other collection metadata. They then go back to the catalog, enter search terms resulting in a short result set, all of which they wish to include, they "select all" and add these to the collection. When their collection is complete they share the virtual collection URI with the set of students in their class. They do this by cutting and pasting the collection link into an email to the students. When the students follow this link they have access to the collection and access shows summary information and links to each item.

Status 2015-02-22

Completed:  A user can create a new collection.  A user can navigate the Cornell catalog to an individual resource and add it to the catalog.   The URI for the virtual collection can be emailed to others allowing other to access the collection.

Incomplete: A user can not add a short result set.

B. Extend A to allow selection of materials from other institutions. Should users be able to see what remote collections an item is in? And thus able to use the intelligence of multiple universities to guide selection?

Status 2015-02-22

Completed:  A user can navigate catalogs at Stanford, OCLC, VIVO Cornell, and other institutions that serve rdf and copy a URI.  One at a time, an URI can be submitted to the virtual collection system and it will be added to a virtual collection.  Remote items are identified with the source providing the metadata.  Limitation: The system will not succeed interpreting rdf from all sites.  It is limited to certain ontologies: BIBO/VIVO, Schema, Bibframe.

Incomplete: Items in the catalog do not identify collections to which they belong.  Cross-site selection of multiple items from a single search implementation is dependent on Use Case 6.1

C. Extend A/B to allow ordering of items in the virtual collection and comments about each one

Status 2015-02-22

Completed:  A user can set a comments and a set of tags on items in a virtual collection.

Incomplete: Ordering of items is not available yet.  Comments and tags are not set on virtual collections.

D.  Extend A/B/C to allow authentication for non-public lists, either private or group level. (e.g., lists in progress to publish, personal scratch space not ready to publish, my research team is doing cutting edge research and don’t want to have a public list)

Status 2015-02-22

Completed:  N/A

Incomplete: Authentication and access controls are not implemented; therefore, all controlled vocabularies are public.

Data Sources

  • Catalog records
  • Records for any other resources that are to be available for inclusion (demonstration of the use case does no depend on more records being available but would be more powerful with a variety of media (e.g. books, journal articles, images, video, etc.)
  • If extended to non-public access to the need appropriate information to control access. This might be subject area or class membership information if sharing with sets of students in area or class. Such information might not be needed if controlled channels are used to share the collection (e.g., Blackboard)
  • Person information related to authentication

Ontology Requirements

  • Model for a virtual collection resource with associated metadata including title, description, authorship, etc. (and, in local system at least, owner(s) to control edit privileges)
  • Ability to add items to virtual collection
  • Ability to add comments/annotations for items specific to the context of the virtual collection
  • Ability to specify order of items in virtual collection

Engineering Work

  • Authentication as user in class allowed to create lists (at least librarian, faculty member)
  • Model of user classes/groups/individuals and authentication of these in access system to see a particular virtual collection
  • Method for selecting resources not only in local discovery system but also other university discovery system (or are resources from other university displayed locally for selection?)
  • Authentication as user allowed to view virtual collection if extended to support non-public uses

Who will do what?

  • Cornell needs this to replace the CuLLR infrastructure supporting “virtual libraries”
    • Lynette/Simeon
  • Harvard has a non-linked data Library Cloud infrastructure in prep that will have a UI for building virtual collections
    • could have the developer (Michael Vandermillen) consult on UI issues but not write code. (LibaryCloud probably using a modified version of Omeka, so probably not very useful.)
    • LibraryCloud is going down the maker path (?)
  • Stanford -- about 6 documented use cases this would address
    • need to understand/refine distinction between 1.1 and 1.2
    • who? look at in aggregate first
    • Stanford thinking of extending Blacklight to do this; like Spotlight tool for creating exhibits
  • No labels