- From: Richard Wallis <richard.wallis@dataliberate.com>
- Date: Wed, 1 Jun 2016 16:25:01 +0100
- To: Wes Turner <wes.turner@gmail.com>
- Cc: "phil.barker@hw.ac.uk" <phil.barker@hw.ac.uk>, "public-schema-course-extend@w3.org" <public-schema-course-extend@w3.org>
- Message-ID: <CAD47Kz5Tcg2CA67RNxxnfqPK43Tj=_i1E188-yroPK9M49XtrQ@mail.gmail.com>
As I think is clear from this and similar threads in other areas, the chances of consensus across overlapping and partially overlapping groups, areas of expertise and interest resulting in a workable *One Enumeration Set to Rule Them All* is probably unachievable. Such a set becoming part of Schema.org or an extension vocabulary being even less likely. Ideally URIs/Values from authoritative industry or academic groups should play a significant part here. Some already exist (Library of Congress Subject Headings being a classic example). As Wes points out, the technical barriers are not too large in publishing such a set of enumeration values - an authoritative group of people committing to create and maintain them at a persistent location. Alternatively identifying a good existing authoritative source and helping them publish their values as a persistent URIs. On the Markup side of things, as Phil pointed out, there is a proposal < https://github.com/schemaorg/schemaorg/issues/894> focused on marking up enumerations on external [to Schema.org] sites. ~Richard Richard Wallis Founder, Data Liberate http://dataliberate.com Linkedin: http://www.linkedin.com/in/richardwallis Twitter: @rjw On 1 June 2016 at 15:49, Wes Turner <wes.turner@gmail.com> wrote: > > > On Wednesday, June 1, 2016, Phil Barker <phil.barker@hw.ac.uk> wrote: > >> Hello Wes, everyone. >> Yes, enumerations are good when you can reach consensus on what they >> should be. The problem is agreeing on them across the scope of all people >> who might use schema.org. External enumerations [1] would help by >> reducing the scope of the required agreement. In general, I think the >> enumerations can be added later. I think we can propose a property and >> then, when we see how it is used, add in enumerations as required. >> >> Taking the properties that you suggest would benefit from an enumeration >> in turn: >> >> AccessibilityPhysicalFeature >> - I think this is beyond our competence. Physical accessibility for >> events / locations needs to be considered in a much wider context than >> Courses and by people who understand the issues involved. I would rather >> leave it to a different working group. (This incidentally is what LRMI did >> for accessibility of creative works, and it worked out quite well.) >> > > So it was in scope for LRMI. > It's a property (accessibilityPhysicalFeature) and like 10+ Enumerations > (braille, largePrint, wheelchairAccessible,). > > >> >> >> CourseMode >> - Agreed. The option set for the CEDS Course Section Instructional >> Delivery Mode https://ceds.ed.gov/element/001161 seems like a good >> starting point. I know Jim Goodell is interested in whatever it takes to >> make the CEDS vocabularies more usable in the wider RDF world. >> >> > > CEDS: CourseSectionInstructionalDeliveryMode > > schema.org: courseDeliveryMode ? > > [external] enumerations (name, URL) would hopefully reduce the number of > misspelled schema.org/Text properties (so that faceting by this attribute > is reasonably feasible without extensive de-duplication) > > we could include Text in the range for courseDeliveryMode but expect URLs > of Enumeration Things with names (because those are translatable) > > > > In general, I think the enumerations can be added later. I think we can > propose a property and then, when we see how it is used, add in > enumerations as required. > > -1 on "everyone will realize they need to go back and re-work to upgrade > from Text to URL later" > > * "IDK, the person who did it last semester must've done it wrong and it's > too late to change now" > > Here's how those go (folksonomy -> structured vocabulary): > > > - "Interactive Audio/Video" > - "Audio/Video" > - "AV" > - "Audio-Video" > - "Audio" > - "Video" > - https://ceds.ed.gov/element/001161#AudioVideo > > > With just Text, there's not enough information. > > * RDF requires URLs (IRIs) as subjects (not Text) > * <#AudioVideo> a Enumeration > * <#AudioVideo> schema:name "Audio/Video"@en > * <#AudioVideo> schema:name "ljud video"@sv > * <#AudioVideo> schema:description "Course is taught via remote > interactive receiver or via streaming media technologies." # CEDS > > * When you want to GROUP BY courseDeliveryMode, if they're not all > consistent URLs, it's very noisy/lossy/insufficient (and then faceted > interfaces don't work without deduplication) > > > > >> >> >> CourseDigitalFormat >> - I am not convinced we even need this as a property. Can we start a >> separate thread to discuss it? >> > > CourseDigitalFormat is necessary in order to identify courses which offer > digital formats (such as complex XML schema). > > There's no need for a separate thread; or a separate mailing list; or a > separate issue ticket. > > - [ ] Add the property (courseDigitalFormat) > - [ ] Add the External Enumeration class (CourseDigitalFormat) # see > above/below > - [ ] Create a wiki page to host said Enumerations > - [ ] REQ: @vendors: pick a camelCase URL for a SMWiki page > > > > >> >> >> W3C support for Semantic MediaWiki >> - What level of support? The community group wikis are Semantic MediaWiki >> (at least according the badge on the bottom right of the >> schema-course-extend and web-schemas wikipages). Do you mean support in how >> to use it? I tried using SMW for a different task last summer and found it >> a difficult combination of RDF thinking and some of the more esoteric >> features of MediaWiki. >> > > We should now be able to define Enumerations in a [SMWiki] table of > structured data > (such that, when we parse an [Enumerations] wiki page for {RDFa, > Microdata, JSON-LD, Turtle} with e.g. OSDS > https://github.com/openlink/structured-data-sniffer , > we can generate what we need (e.g. labels, description) in our > applications from RDF without copy-and-pasting); IDK who would be helpful > with that. > > [*] Inline links are much more accessible than (not linked) footnote > citations. > > >> Phil >> >> 1. https://github.com/schemaorg/schemaorg/issues/894 >> >> On 21/05/16 16:22, Wes Turner wrote: >> > In reading the examples that Phil added here: >> <https://github.com/westurner/schemaorg/pull/3#issuecomment-220782903> >> https://github.com/westurner/schemaorg/pull/3#issuecomment-220782903 >> >> I wonder whether, for faceted interface support, we should have >> courseMode Enumerations? >> >> We could define said Enumerations just like these: >> https://www.w3.org/wiki/WebSchemas/Accessibility >> >> >> AccessibilityPhysicalFeature (for Place, Event, CreativeWork, ) >> https://www.w3.org/wiki/WebSchemas/Accessibility PhysicalFeature >> - name >> - description >> - url >> - image / logo / photo (Creative Commons {SVGs,} would be great here) >> - https://github.com/schemaorg/schemaorg/pull/972#issuecomment-173698449 >> >> >> CourseMode >> https://www.w3.org/wiki/WebSchemas/CourseMode >> - name >> - description >> - url >> >> >> CourseDigitalFormat >> https://www.w3.org/wiki/WebSchemas/CourseDigitalFormat >> - name >> - version >> - description >> - url >> >> >> Suggestions: >> - These Enumeration URLs should/could be copy-paste-able >> - Semantic MediaWiki supports RDFa >> - IDK how feasible it would be to get W3C support for Semantic >> MediaWiki (or really who to contact for such a request) >> >> >> >> -- >> -- >> Phil Barker @philbarker >> LRMI, Cetis, ICBL http://people.pjjk.net/phil >> Heriot-Watt University >> >> Ubuntu: http://xkcd.com/456/ >> not so much an operating system as a learning opportunity. >> >>
Received on Wednesday, 1 June 2016 15:25:30 UTC