[OGDev] Ratings and Ratings on the OpenGuides (via Revyu and SPARQL?)

[Tom Heath]

Hey Chris,

Glad you like it. OpenID support seems to be a popular request ;) I'll
look into it and let you know how I get on. The kind of scenarios
you're talking about are exactly the kind of thing I'm interested in
supporting.

In the meantime, basic smushable descriptions of Revyu users are
available from addresses of the format
<http://revyu.com/people/username/about/rdf>, e.g.
<http://revyu.com/people/tom/about/rdf>

Speak soon,

Tom.

On 11/12/06, Chris Prather <chris at prather.org> wrote:
> Tom Heath wrote:
> > Hi all,
> >
> > Apologies it took me so long to pick up this thread. Slightly later
> > than planned, here goes:
> >
> > Background: there seems to be some enthusiasm for supporting
> > reviews/ratings on the OpenGuides, both of articles, and the things
> > the articles are about (ie. "Rate this article about the Red Lion Pub"
> > and "Rate the Red Lion Pub").
> >
> > Issues: is it worth developing Yet Another Reviewing and Rating
> > Infrastructure, specifically for the OpenGuides?
> >
> > Potential Solution: I have developed Revyu.com <http://revyu.com/>, a
> > site that acts as a generic platform for creating reviews and ratings
> > of things (anything at all), on the basis that people can easily get
> > reviews and ratings back in RDF/XML (via simple GET requests or SPARQL
> > [1] queries) for reuse in their own applications.
> >
> > In the context of the OpenGuides, it would be pretty trivial to
> > provide "Review/Rate this article about the Red Lion" and "Review/Rate
> > the Red Lion" links on OpenGuides pages, that pass the user to
> > Revyu.com to provide the review, then sends them back to the OpenGuide
> > page afterwards, very much like the way "add this to del.icio.us"
> > links work on sites like The Register. SPARQL queries from the
> > OpenGuides sites could then retrieve all reviews related to a
> > particular page, from Revyu.com, and present the results as desired.
> >
> > This approach has some nice advantages: it doesn't require changes or
> > additions to underlying database structures, creation of new reviewing
> > forms or scripts to do RDF output of reviews (this is all provided by
> > Revyu), and may help the OpenGuides reach new audiences as content
> > would show up on other sites. It would also serve as a nice
> > demonstration of the power of doing things Semantic Web-style, and how
> > amenable Open Guides are to this. Disadvantages are few IMHO, mainly
> > that it would require SPARQL support to be added to the OpenGuides at
> > some level. On the upside, I think there are decent Perl libraries for
> > SPARQL, and it is a fun thing to learn. As an interim solution we
> > could always use iframes or some other method to show content from
> > Revyu on the OpenGuides.
> >
> > So, what do people think? Is this something people would be interested
> > in doing? Please fire back any questions, or check out the site
> > <http://revyu.com/>, or both. Anyone going to Oxford tonight for the
> > Hackfest will probably be lucky enough to get a demo ;)
> >
> > Looking forward to hearing thoughts, comments, etc.
> >
> > Cheers,
> >
> > Tom.
> >
> > [1] This is a nice intro to SPARQL:
> > http://www-128.ibm.com/developerworks/xml/library/j-sparql/
> >
> >
> Ooh, I really like the idea of Revyu!
>
> One suggestion: OpenID support. I'm about to try to attack it again with
> OpenGuides and a couple other projects. It would be really nifty to be
> able to use a distributed "Identity" across Revyu and OpenGuides ... it
> would make meta queries like "show me all guide pages Edited or Reviewed
> by chris at prather.org" possible via smushing and SPARQL. In a longer
> context it would make "show me all pages in the London Guide that people
> I trust like" which would smush your personal Foaf for foaf:Trust, the
> Guide and Revyu ... in a nicely semantic way.
>
> -Chris
>
> --
> OpenGuides-Dev mailing list - OpenGuides-Dev at lists.openguides.org
> http://lists.openguides.org/cgi-bin/mailman/listinfo/openguides-dev
>