Comment on RDF 1.1: use case for HTML5 data type

Dear RDF working group,

This is a comment on RDF 1.1. I want to make you aware of the work
undertaken in the MultilingualWeb-LT (MLW-L)T working group
http://www.w3.org/International/multilingualweb/lt/
We are defining metadata to enhance language technology applications and
localization services on the Web. An example application is machine
translation. Here recently a "translate" attribute has been added to HTML5,
see
http://rishida.net/blog/?p=831
for further information and for uptake by current online machine
translation services.

The MLW-LT working group is developing other kind of metadata also to be
added to HTML5. Details of the mechanism are under discussion, but we are
likely to continue an attribute approach like with "translate", using a
specific prefix for the time being.

The goal of the MLW-LT group is to bring this metadata to other types of
content as well.
RDF is an important type of content for us. We would like to avoid
developing content type specific metadata syntaxes as less as possible.
This is the motivation for my mail.

I have heard that there are discussions about adding an HTML5 datatype to
RDF 1.1. Such a data type would help us tremendously, since we could just
re-use our attribute approach: the attributes would be part of HTML5
syntax, no matter whether HTML5 pieces show up as HTML5 itself or whether
they are part of e.g. a label in RDF. For the latter, RDF (1.1.) of course
would need to allow for the HTML5 datatype.

I hope that this use case is of interest for you and helps to motivate the
introduction of an HTML5 datatype in RDF 1.1. Looking very much forward for
your feedback.

Best regards,

Felix Sasaki (co-chair and staff contact of the MLW-LT Working Group)


-- 
Felix Sasaki
DFKI / W3C Fellow