- 1 Use Cases to test proposed extensions against.
- 1.1 Scenarios for end-users
- 1.2 Scenarios for search engines
- 1.3 Use case: Publication event in library applications
- 1.4 Use case: Describe a FRBR Item (e.g. a single identifiable book)
- 1.5 Use case: Describe a FRBR Manifestation
- 1.6 Use case: Validate Schema Extensions Against the Developing BSR
- 1.7 Use case: Describe library holding/availability information
- 1.8 Use case: Electronic materials "appropriate copy" information
- 1.9 Use case: journal articles and other periodical publications
- 1.10 Use case: Content/carrier categories and RDA/ONIX Framework
- 1.11 Use case: RDA relationship designators unconstrained by FRBR entities
- 1.12 Use case: Describe a digital Item (i.e. a copy available online)
- 1.13 Use case: a slightly longer-range focus via Fuller, Foster & Gould
- 1.14 Use Cases for Publishers
Use Cases to test proposed extensions against.
Publish short summaries on this page - link to more detailed descriptions if available
Scenarios for end-users
Support real-time buying decisions
- Library software (e.g. Ex Libris Alma) would use a vendor site to support real time decision to buy a book (i.e. from Amazon) and have delivered to student as opposed to routing the user request to ILL. If multiple vendors support schema.org than this would be easier to support more vendors, comparison and buy based on Library rules (price, time to delivery, etc.).
What is needed for the scenario to work
Smart phone directs to appropriate copy
- Some smart phone travel app want to direct its customers to history books (even better eBooks and eArticles) available to him in his library that describe his travel location (e.g. some place in China). Libraries supporting schema.org would enable the smart phone app to use library resources and display rich information regarding the possibilities. I doubt that they would want to use MARC J. This scenario requires discussion of appropriate copy, what entails rich information for such a solution, etc.
What is needed for the scenario to work
Embedding citation metadata in HTML (COInS++)
Target service: a web resource that surfaces bibliographic information such as library catalog, an abstracting & indexing database, an institutional repository, a digital exhibit, a review site...
Citation managers such as Zotero, EndNote, RefWorks, and the like generally work by:
- relying on targeted services to support direct export of citations in a standard format (in some cases offering a choice of standard formats via the lightly-used unAPI lookup mechanism);
- converting COInS metadata (http://ocoins.info/) into a citation, when available; or
- implementing the equivalent of screen-scraping to generate citations from the generated HTML
A sufficient definition of citation metadata in schema.org would make it possible to simplify the work of the citation managers, for those target services that adopt schema.org.
What is needed for the scenario to work
(Potentially): work backwards from what citation managers want to identify the metadata elements that we need to surface in schema.org microdata
Scenarios for search engines
Please refactor these into one of the above categories
Use case: Publication event in library applications
Submitted by: Gordon Dunsire
Several RDF schema for library/bibliographic metadata include provision for publication statements/events, including ISBD (http://iflastandards.info/ns/isbd/elements/P1162), RDA (http://rdvocab.info/Elements/publicationStatementManifestation), and FRBRoo (http://iflastandards.info/ns/fr/frbr/frbroo/F30 - under construction; see http://www.cidoc-crm.org/docs/frbr_oo/frbr_docs/FRBRoo_V1.0.2.pdf). The British Library has also include similar provision (http://www.bl.uk/schemas/bibliographic/blterms#PublicationEvent - see http://www.bl.uk/schemas/bibliographic/blterms)
There is some synergy with the PublicationEvent Class proposals from the TVRadio Schema proposals: http://www.w3.org/wiki/TVRadioSchema
In all cases the element (property or class) has components; the simplest common pattern includes date of publication (equivalent to start and end date of the event, typically a year date for book-type resources), place of publication, and name of publisher. PublicationEvent classes are used in place of blank nodes where a resource has more than one publication event, and as statements aggregating the values of components.
Use case: Describe a FRBR Item (e.g. a single identifiable book)
Submitted by: Jeff Young
Suggest this mapping between (FRBR) Item and GoodRelations Individual:
<http://purl.org/library/Item> a owl:Class; rdfs:subClassOf <http://purl.org/goodrelations/v1#Individual>, <http://schema.org/IndividualProduct> .
Use case: Describe a FRBR Manifestation
Submitted by: Jeff Young
Suggest this mapping between (FRBR) Item and <http://purl.org/goodrelations/v1#ProductOrServiceModel>:
<http://purl.org/library/Manifestation> a owl:Class; rdfs:subClassOf <http://purl.org/goodrelations/v1#SomeItems>, <http://schema.org/SomeProducts> , <http://purl.org/goodrelations/v1#ProductOrServiceModel> .
Use case: Validate Schema Extensions Against the Developing BSR
Submitted by: Philip Schreur
I'd like to start out by saying that I am in total agreement with Antoine and others that any expansion to Schema must be kept simple. That being said, The Program for Cooperative Cataloging (http://www.loc.gov/aba/pcc/), is completing work on specifications for a Bibliographic Standard Record that can be used with any format. Although this is intended as a guide for the creation of a catalog record, it should be possible to extract the high level elements in the specification and compare them to Schema to expose any major area that might be missing in the description of standard library materials. The Policy Committee of the PCC meets in early November and I hope to get their help in this review.
Use case: Describe library holding/availability information
Submitted by: Richard Wallis
It should be possible to describe the availability of a CreativeWork for loan from an organisation - either from the organisation publishing the information, or a third party agent referencing that availability. Additional information including quantity available, loan periods, costs etc. should be possible.
[kcoyle] see my example based on Google Rich Snippets: http://kcoyle.net/libRichSnip.png Some screen shots of actual library holdings displays are at: http://kcoyle.net/holdings.html
[RJW] this example introduces a potential need for being able to describe type of availability (on shelf) and an actionable URI (place hold, etc.)
[kcoyle] schema.org doesn't seem to use, much less require, standard lists of values. I imagine that the library community will want to use such lists, but that this will be outside of the schema.org property definition. Within our community we can create controlled lists; I suspect that other entities using schema.org will do the same.
[pkiraly] Possible starting points would be: DAIA - Document Availability Information API (http://www.gbv.de/wikis/cls/DAIA_-_Document_Availability_Information_API), NISO Circulation Interchange Protocol (http://www.ncip.info/).
[kcoyle] These are good starting points. I wonder also if there isn't some way that OpenURL can be employed here? We may have to design for a future, unknown source of holdings data. Most local library systems could not sustain direct querying, and WorldCat is large but far from complete (although perhaps a good start for experimentation). Note that http://schema.org/Offer has "ItemAvailability" and there is also http://schema.org/Organization has "Location" -- so we might be almost there if not fully there with these elements. We would need to define our own "availability" values.
[Jyoung4] Here's one possible way to express a "holding" using http://www.schema.org/Offer: e.g. Ohio State University "holds" Zen mind, beginner's mind:
@prefix schema: <http://schema.org/> . _:A0 a schema:Offer ; schema:itemOffered <http://www.worldcat.org/oclc/136259> ; schema:seller <http://viaf.org/viaf/169435843> ; schema:url <http://library.ohio-state.edu/search~S7?/o136259/o136259/1%2C8%2C8%2CE/frameset&FF=o136259> ; schema:businessFunction <http://purl.org/goodrelations/v1#LeaseOut> ; schema:price "Library loan" ; ] .
Use case: Electronic materials "appropriate copy" information
Quoting from Dan Scott email:
"...similar to how scholar.google.com recognizes user IP address ranges and serves up links that lead directly to electronic resources licensed by the corresponding institution (assuming that you've gone through the process of exposing your electronic holdings in Scholar format and told Scholar to go ahead and read them), I think the schema.org principle search engines (hereafter referred to as "SoPSE") would benefit from rewarding users who search via their search interface with links to instances of the requested resources at institutions associated with the current user.
Geo-IP could point directly at public libraries where the resource can be accessed rather than relying on the OCLC middleman, and the user's ability to access restricted resources (e.g. borrowing by GLAM institution members only) could be via some combination of IP range or explicit affiliations in the user's SoPSE account, so that inaccessible resources wouldn't clutter search results. The explicit benefit to the SoPSE is that explicit affliliations strengthen signals that the SoPSE can then use to serve up relevant ads to that user, and of course the SoPSE can offer users the option of immediately purchasing a personal digital license for the resource via the SoPSE's Music / Books / other licensing & vending services.
A possible side benefit that PSEs would gain from having physical & electronic resources surfaced via a schema.org extension is that they might be able to evolve from the Google Scholar "export and refresh a list of resources in a specific format" model towards a more standard "crawl based on published sitemaps & microdata" model."
Use case: journal articles and other periodical publications
Mark up bibliographic citations of journal articles such that they can be identified and possibly located ("appropriate copy" or other functionality). One model to follow could be to mimic the data in the OpenURL, which is already proven in this area, and which would interact with the OpenURL resolver databases that exist.
Phil Schreur reports that HighWire Press is interested in the following fields (which appear to also satisfy the OpenURL case):
article Abstract - The abstract or summary of the article
volume - the print volume that this article was published in
issue - The print issue that this article was published in
pages - The printed page range for this article
locator - A description (often numeric) that locates this article within its publication
doi - The digital object identifier (DOI) used to uniquely identify this object
pmid - The PubMed Identifier (pmid) assigned to this article's PubMed record
publication - The full name of the journal or publication this article was published in
publicationAbbreviation - The PubMed or other scholarly abbreviation of the journal this article was published in
issn - The International Standard Serial Number used to identify the print edition
eissn - The International Standard Serial Number used to identify the electronic edition
owner - The copyright holder for the article
rights - The license status of the article
cites - A reference that this article cites
A similar case and work is taking place at http://www.w3.org/wiki/WebSchemas/PeriodicalsComics.
aTitle = article title jTitle = journal title bTitle = book title v = volume n = number date = date (year or more detailed) pp = pages (page range) lang = language pubPattern = publication pattern ed = editor issn = ISSN doi = DOI
AU | aTitle | jTitle | v | n | date | pp AU | date | | aTitle | jTitle | v | n | pp aTitle | AU | lang | [jTitle or bTitle] | date | pp aTitle | AU | [jTitle or bTitle] | ed | date | pp AU | pub | lang | jTitle | v | n | date | pp
jTitle | issn | pubPattern | publisher jTitle | issn | v | n | pp jTitle | issn | v | n | date | pp
Journal issue with contents
jTitle | issn | v | n | date aTitle | AU | pp | DOI | date aTitle | AU | pp | DOI | date etc.
Use case: Content/carrier categories and RDA/ONIX Framework
Submitted by: Gordon Dunsire
The RDA/ONIX Framework for Resource Categorization provides a basis for creating content and carrier categories suitable for the library and publishing environments.
It is used for the RDA Carrier Type, Content Type, and Media Type vocabularies. A Mapping of ISBD area 0 vocabularies to RDA/ONIX Framework is available as part of the activities proposed in Mapping ISBD Area 0 vocabularies to RDA carrier and content vocabularies .... Further development and use of the Framework will be discussed under item 12 on the agenda for November 2012. This includes representing the element set and value vocabularies in RDF. The Vocabulary Mapping Framework is an extension of the RDA/ONIX Framework.
Value vocabularies for media categories used or proposed by communities outside of RDA and ISBD tend to mix content and carrier concepts, resulting in limited support for interoperability and resource discovery based on content or carrier characteristics.
Use case: RDA relationship designators unconstrained by FRBR entities
RDA Toolkit appendix J lists designators for relationships between resources. These are represented in RDF as properties in the element set RDA Relationships for Works, Expressions, Manifestations, Items.
The Vocabulary Mapping Framework covers relationships between agents and resources, including inter-resource relationships, from multiple schema including RDA, but is not synchronized with amendments made in the past three years.
The RDA element set includes a parallel set of properties based on the relationship designators but without dependency on the FRBR Work, Expression, Manifestation, or Item classes; for example, the Derivative relationship between two Work-type or Expression-type resources. The element set is not stable, and remains in "New-Proposed" status. The JSC/RDA will discuss further development of the relationship designators under item 12 on the agenda for November 2012, and in particular the proposals in RDF representation of RDA relationship designators.
The RDA relationship properties are a rich extension of the high-level relationships given in FRBR; the unconstrained, FRBR-free versions should offer simple representations of FRBR-ish relationships between creative works in a way that the wider web will understand.
Use case: Describe a digital Item (i.e. a copy available online)
Submitted by: Romain Wenz
For instance, digital items associated with a book, digital copies of a book, and so on. It would be good to link other items (FRBR manifestation, for instance) with these elements, so as to identify if the document is directly available.
Use case: a slightly longer-range focus via Fuller, Foster & Gould
Submitted by: Jerry Persons
Schema.org taken as a whole is one recent example of an emerging web-wide trend toward increasingly well-structured representations of information—all a part of the so-call migration from a web of pages or documents toward a web of data. Extending the bibliographic segments of this particular schema (filling in the gaps as Phil Schreur suggested) is an important part of improving the navigability of information resources spread across references to all manner of cultural heritage resources found throughout the web.
The use cases outlined here assume that a richly populated web of data already exists -- one that supports facile navigation and discovery of resources connected by links that label the interwoven relationships among people, places, dates, organizations, events, things, ….
So, yes, there are many use cases that involve the consumption of improved schema.org markup to the considerable and measurable benefit of those who search today’s web.
It’s also worth reminding ourselves that adjusting the lens to a slightly longer focus gives some inkling of the potential longer-term benefits to be had from this present-day effort to extend schema.org such that library (and dare one suggest all GLAM) resources become integral components of the increasingly well-structured raw materials being consumed by tomorrow’s navigation and discovery environments.
Use Cases for Publishers
Submitted by: Laura Dawson
Publishers have developed highly structured data for digital communication among trading partners - and with ONIX, have managed to create an XML schema to deliver information that's ultimately meant for the consumer. Which is to say, publishers have grown accustomed to creating structured data.
But they are not accustomed to the search box as the primary retailer. Increasingly, this is becoming the case. How to structure data on the web to maximize discoverability of books (and hopefully lead to purchase)?
I have a couple of ideas about this. Obviously metadata on websites has to be optimized for search, and using common vocabularies has proven quite effective. I don't think publishers will push back a great deal here, so long as we can agree to look to ONIX codelists as a source of standardized tags for this specific use case - making books discoverable on the web for the purposes of sale.
But furthermore, it will become critical to structure the book itself. An epub file is not much more than a very long HTML file with some additional information zipped up with it. Search engines that index book content are ingesting the HTML file. If that content is marked up semantically, there's a greater chance that book content itself will be returned in a search result, just as ordinary web pages are.
That is, of course, down the road a piece. It's worth keeping in mind, though, that this is a potential direction in which we may be headed.