- From: adasal <adam.saltiel@gmail.com>
- Date: Mon, 5 Apr 2010 17:45:06 +0100
- To: Danny Ayers <danny.ayers@gmail.com>, semantic-web@w3.org
- Cc: "ping.fm" <drktuj@ping.fm>, adam.saltiel.epaddress@blogger.com
- Message-ID: <g2me8aa138c1004050945n87e20ac9n5a300abb71608f3e@mail.gmail.com>
The subject is suitably emotive. Maybe it also resonates with a doubt about the balance between research and realisation? I think it will all come together over a short space of time. Apropos internal applications semantic technologies are a superset of recommender systems? I think so. But what are the business drivers to recommenders, as against a verifiable and unequivocal set of results from the deployed algorithms. How can we know? What criteria veracity in the business context? Just a thought. Best, Adam On 5 April 2010 13:51, Danny Ayers <danny.ayers@gmail.com> wrote: > So basically no-one in the Semantic Web community has a clue about > mailing lists. That is quite cool in itself. > > On 5 April 2010 14:45, Danny Ayers <danny.ayers@gmail.com> wrote: > > Lots of different ideas. Still the one thread. About time someone > > change the subject. > > > > But if it gets up to a hundred, I have something to tell my neighbours. > > > > On 5 April 2010 14:37, Hugh Glaser <hg@ecs.soton.ac.uk> wrote: > >> Certainly :-) > >> And you make a good point: this could have been applied to any emerging > >> technology; for example, replace "Web" with "Database" circa 1965. > >> > >> I think we can and should look around for new and exciting applications; > >> however, Linked Data and Semantic Web Technologies offers solutions to > >> existing problems of data fusion in the corporate world. > >> > >> The most interesting thing I have seen in this thread was a little > tagged-on > >> paragraph from Afraz: > >> ------ Forwarded Message > >>> From: Afraz Jaffri <afrazjaffri@hotmail.com> > >>> Date: Wed, 31 Mar 2010 10:26:36 +0000 > >>> To: <semantic-web@w3.org> > >>> Cc: <danbri@danbri.org>, <afraz.jaffri@capgemini.co.uk> > >>> Subject: RE: call to arms > >> <snip> > >>> FWIW we have 2 large projects that use linked data 'under the hood' > that the > >>> customer doesn't even realise is there. We have a few more in the > pipeline, it > >>> really is just a matter of time. > >> This is not an isolated case. > >> > >> And once a company is using Linked Data internally, it is only a matter > of > >> time before the appropriate subset is made open. > >> > >> For example this is sort of how qdos came about out of garlik > >> ( http://qdos.com/user/Danny-Ayers/26c95f0c5aeac91c7f54b1914db44f2d ) > >> > >> Best > >> Hugh > >> > >> On 01/04/2010 20:16, "Danny Ayers" <danny.ayers@gmail.com> wrote: > >> > >>> Hugh, may I tweak your post a little: > >>> > >>> The use case: > >>> My organisation's support staff have decided that the Web > >>> is the way ahead. > >>> They want to replace their document storage with a web site in a > seamless > >>> fashion. > >>> They ask me for advice, and I point them ... where? > >>> > >>> Where is the site that tells them how to re-target their data layer > >>> from .doc to HTML? > >>> > >>> Where are the specifications of Web performance, so they can do the > >>> procurement process (hardware and software)? > >>> Where are the utilities that help get documents in and out? > >>> Where are the simple descriptions of ... > >>> > >>> Of course some of these things may exist but I don't know about them - > if so > >>> apologies. > >>> But remember - the people concerned are just interested in using good > >>> technology - they really don't want to know about the Web. > >> > >> > > > > > > > > -- > > http://danny.ayers.name > > > > > > -- > http://danny.ayers.name > >
Received on Monday, 5 April 2010 16:45:40 UTC