Re: Can we lower the LD entry cost please (part 1)?

Allow me to clarify, regarding Knoodl as a LD publishing tool.  It is today
mostly a collaborative ontology/knowledgebase builder.  As far as features
for linked data publishing, I will try to outline some of the most important
ones and our expectations for delivering them:
1) RDF/OWL import/export - available now

2) RDF/OWL editing - available now

3) SPARQL endpoint - we plan to deliver this functionality shortly.  The
initial delivery will most likely be without the benfits of
our permissions model, merely a sort of radio button-type functionality
allowing communities to either expose their datasets as public SPARQL
endpoints or not.  Eventually, we will incorporate all of our permissions
into the SPARQL endpoint functionality but, obviously, this is a much more
difficult task.
(note: we do support SPARQL queries within Knoodl now)

4) Ontology mapping - this we have separated into two buckets - (a)
JOIN-type functionality that would allow users to execute complex queries
against multiple data sources and (b) simple creation of assertions that
link ontologies together (i.e. owl:equivalentClass, etc.).
(a) is being developed now and will be delivered shortly.
(b) is available now, as we support all owl constructs, including
owl:imports, though we are considering developing a UI specific to this task
(any feedback as to the appropriate features of such a UI would be
appreciated).

5) URI search - available now

These are the features that I see as being the most helpful to publishing an
RDF dataset.  If I have missed any that anyone thinks are necessary or would
be very helpful, please don't hesitate to let me know.

Thanks,

Mike Lang
Revelytix, Inc. <http://www.revelytix.com>

phone: 410-584-0009 (office)
          443-928-3782 (cell)
skype: michael.allen.lang.jr
aim: MikeJrRevelytix


On Mon, Feb 9, 2009 at 11:47 AM, Hugh Glaser <hg@ecs.soton.ac.uk> wrote:

> It's a fair cop, guv. You got me banged to rights.
> [Translation: I plead guilty as charged.]
> I was being too provocative.
>
> And thank you for the Knoodl plug - I had seen it more as a collaborative
> ontology builder than an LD publishing activity.
> As feedback, I can say that it still looks rather like that - apparently
> the first thing I have to do is create a community, which is not necessarily
> my first thought when I want to bang up some RDF into the LD world.
> But I will revise my view.
> And all strength to you on it.
>
> Hugh
>
> On 09/02/2009 16:16, "Michael Lang(Jr.)" <michaelallenlang@gmail.com>
> wrote:
>
> Hugh,
>
> I do not understand the notion of "permitting a site to be a member of the
> Linked Data cloud".  As I understand it, being a member of the Linked Data
> cloud simply involves linking your RDF data to a dataset that is part of the
> cloud.  How do you intend to add any governance to this process?  If I am
> confused about this, please correct me.
>
> However, I do not agree that there should be any governance to this
> process.  Rather, I believe that self-governance is the only possible form
> of governance that could work in this system at the scale of the Web.
>  Publishers of datasets should make up their own minds about what they will
> supply with their datasets.  In time, proper best practices will emerge and
> proper incentives to follow those best practices will emerge and all
> publishers will have to publish "good" datasets to be successful.
>
> The problem here, which I think several people on this thread have noted
> separately , is pretty simple: there are still too few datasets and too few
> apps (both for publishers and consumers).  Too few datasets means not enough
> competition which means that publishers can get away with sloppy practices.
>  Too few apps for consumers means not enough benefit to publishers which
> means they have little incentive to follow good practices when publishing.
>  Too few apps for publishers means its too hard to publish data which means
> publishers are sometimes forced to use sloppy practices.
>
>
> (product plug)
> On the apps for publishers front, part of what we are trying to accomplish
> with Knoodl.com <http://www.knoodl.com/ui/home.html>  is to deliver a tool
> that provides all the necessary functionality to publish a dataset from
> scratch, including enabling mapping to other ontologies/datasets, turning it
> into a SPARQL endpoint, having search for vocabulary terms, allowing for
> rich documentation of terms, etc.  We are still working on many of these
> features but they are all on our roadmap.  Of course, we are always glad to
> have any feedback regarding the current state of the product or the future
> direction.
>
> Mike Lang
> Revelytix, Inc. <http://www.revelytix.com>
>
> phone: 410-584-0009 (office)
>           443-928-3782 (cell)
> skype: michael.allen.lang.jr
> aim: MikeJrRevelytix
>