Category or Sub-category of learningResourceType?

[John Graves]

I am looking at the specification:

http://www.lrmi.net/the-specification

specifically, learningResourceType

where there is an example, “presentation”

SlideSpeech is developing several new types of presentation, building up from the base type (a set of PowerPoint slides), by adding first one computer-generated voice over, and then two (or more) computer-generated voice overs in different languages (press up arrow to toggle English-German).

What if LRMI includes a taxonomy using MIME-type like category/sub-category designations?

These indicate the file format of the presentation

presentation/ppt

presentation/keynote

presentation/odp

presentation/tex

while these might be used to designate these new types, by language

presentation/slidespeech-en

presentation/slidespeech-de 

presentation/slidespeech-multiple

[Greg Grossmeier]

<quote name="John Graves" date="2012-06-15" time="09:41:56 +1200">

> What if LRMI includes a taxonomy using
> MIME-type

> <http://en.wikipedia.org/wiki/Internet_media_type#List_of_common_media_types>
> like category/sub-category designations?

Schema.org, and LRMI by extension, mostly side-step the vocabulary issue
and prefer to let the communities of use organically come to a
consensus. You actually see this happen quite quickly for things like
places and people (where the vocabulary tends to be: use Wikipedia
URLs).

LRMI is a special actor in that it is both a community and an extension
to Schema.org. Thus, we hope to see what vocabularies are most useful
for the community, through both feedback on this list and via just
watching what happens in the wild, and making recommendations where
appropriate.

In this case, however, I don't think LRMI is the right place to
enumerate a category/sub-category of presentation.

If others on the list disagree, please feel free to chime in with
support and a recommendation on how/what it would look like and who
would maintain it.

That is an important aspect to note: LRMI (the individuals and
organizations behind it) are not going to maintain a growing vocabulary
(or really any vocabulary). Instead, we would prefer to point to an
already agreed upon and well maintained one somewhere else.

Hope that helps clarify the options here,

Greg

--
Greg Grossmeier
Education Technology & Policy Coordinator
twitter: @g_gerg / identi.ca: @greg / skype: greg.grossmeier

[Monty Swiryn]

If you've seen my previous posts, you may not be surprised that I support the notion of Category and Subcategory properties.  In the K-12 sector, there are many publishers that use this schema to give users a method for browsing and drilling-down to the products they need.  In many cases, the entire site is structured this way.  Their customers are used to looking for products in this manner, so if their search behavior resulted in high Google rankings, it would be very pleasing to the publishers, to say the least.

If you'd like examples of sites that use this paradigm, let me know and I'll be happy to come up with a list.

thx

Monty

[Greg McFall]

I totally agree that it is important to "let the communities of use organically come to a consensus" about the terms in a vocabulary.  However, I would argue that using a data type of "Text" is somewhat limiting.  I would rather see the data type modeled as a Thing.  This way, it can be treated as an object in its own right.  In practice, this means that one would end up using URI values as identifiers for learningResourceType objects.  Yes, it is possible to stuff a URI into a Text value, but the semantics are broken when you do that.

This same comment applies to many other properties in LRMI.  In general, LRMI frequently uses Text as the data type for a value that is better modeled as a resource in its own right.