Re: Use of Content Management Systems, ranking and usage in Semantic Web

Paul.
What does this mean?
Disclaimer: while I've never actually run Fedora, I *do* work for them.
You are talking about Fedora 9 (0r 10) which is the OS for one of my PCs?
And you work for Red Hat?
I hadn't thought semantics came into the OS so much as into the efforts
being made by the KDE community. This is a brave attempt in early stages,
but with some results. You have to use KDE to see this.
Is there something else Fedora are involved in, I would like to know?

With regard the main issue it is very complex.
I notice how often when these things are discussed the semantic web, as some
public space like the existing web, is usually referred to.
I think this is right in a way as in discussing the public space it is also
being defended.
But it is also far too vast for me to think about.
What I do find I can think about are the odd things I come across, for
instance an International Law firm that is using semantic technologies to
create a far more useful and usable document management system, managing
case law and showing how it evolves in context.

There are principles that are emerging.

   1. An optimum atomic content entity must be made available.
   2. Create views on that entity.
   3. Views are governed by semantic principles.

This is just my way of putting it. The discussion of personas and schemas
belongs at 3.

I work on one of the largest content sites in the world.
I would very much like to introduce a semantic enablement of this content,
in both how it is created and managed and in how it is served.
But as you may imagine there are processes already in place and an
application already built that have their own momentum.
An evaluation of benefit has to be made in this context.

Let's take the internal environment.
Content is divided into several themes. For each theme there is a theme
manager.
I believe it follows that most of the content theme managers are concerned
with is that that belongs to their theme.
Before we go rushing that there is something simple here, by concerned with
I may mean a spreadsheet with items of work on it, a meeting that discusses
the nature of a piece of content or any one of a number of other things that
I am not myself fully aware of.

Certainly one of the questions that would be part of such an evaluation
would be whether a logical framework as would be found in schemas
representing the organisation of this work would aid in organising the
complex and varied tasks with which people have to engage so as that they
may more surely follow agreed items.

Let us say that this is all good, that a schema is devised, that the schema
has persona - in this example the individual themes, and that we have
managed to sit this somehow on or along side the existing system.

And let us say this didn't prove too expensive to implement or disruptive to
the business.

The next, obvious, issue is if or how this can be used in the application
that serves the site content?

Again there are three obvious points where this can be done.


   1. By re-engineering part of the existing application.
      - Too many unknowns.
   2. By re-engineering or replacing part or all of the existing CMS.
      - As above, too many unknows.
   3. By creating a semantic layer.
      - To me this would seem the only alternative.

But what exactly would the benefits be?

At the moment the application is Java based and the site is being PHP
enabled.
I understand that this must be in order to take advantage of PHP templating
and widgets for Web 2.0 enablement.

Again I have reduced this to three areas -

   1. A site that I can personalise for myself.
   2. A site that personalises content for me.
   3. A site that offers widgets that do either or both of the above.


There are also the pricipals by which personalisation is made in 1. or 2. In
otherwords, what exactly do we mean by personalise?

Surely personalise in the context of the site I am describing is
semantically relevant content in the context of this thread?
It is a function of data retrieval.
My argument would be that proper logicallly arranged semantically tagged
content out of all the possible alternatives would be the most useful
addition here.

I wounder what your thoughts are?

Adam




2008/11/15 Paul Gearon <gearon@ieee.org>

>
> 2008/11/14 Sw-MetaPortal-ProjectParadigm <metadataportals@yahoo.com>:
> > Dear Paola,
> >
> > Getting back to the root question of CMSs, I am having a hard time
> figuring
> > out why so little semantic web functionality is built into the current
> most
> > popular CMSs being used.
> >
> > The functionality of CMSs resembles to a large extent what intended users
> > (web designers and users) expect of them in order to build sites, and
> > semantic web issues apparently are not yet high on their agenda.
> >
> > As a mathematician and global sustainable development activist I had a
> hard
> > time when I tried to come up with a practical approach based on existing
> > infrastructure and ICT technology to see how the UN ICT Task Force tried
> to
> > envision global empowerment of stakeholders in sustainable development
> > through the use of ICT and the internet in particular.
> >
> > I came up with an obvious shortlist, the internet, open source software
> and
> > toolkits, and open (internet) access digital repositories of knowledge
> and
> > information.
>
> I notice that the CMS systems you provide as examples (Drupal, Joomla,
> etc) are all CMSs with semantic web extensions applied. While not as
> popular, systems like Fedora are certainly gaining in popularly and
> are built on Semantic Web structures from the outset. (I don't know of
> others, but I'm sure they're out there). I hope you are including
> these in your assessments.
>
> While these systems may not meet the levels of Semantic Web
> integration that you appear to be looking for, Paola has a very good
> point in that it is much more important to meet people's needs than it
> is to push Semantic Web concepts as far as possible. The Semantic Web
> standards enable features to be easily implemented in software. I
> don't believe they should be driving the features for the pure sake of
> being more in tune with the "Semantic Web".
>
> You might also question why the popular CMSs aren't built on semantic
> web technologies, and why you have to look at less well known projects
> to find this. I would say that it is due to ignorance about what the
> semantic web has to offer. The Semantic Web may have been around when
> Joomla was started, but did the developers of Joomla know how to use
> the semantic web? Did the users know what kind of functionality would
> be readily available if their CMS was built on top of semantic web
> technologies?
>
> Obviously these benefits are starting to be recognized now, with
> semantic web modules becoming available for the well known systems,
> and the lesser known systems like Fedora becoming more popular.
>
> Regards,
> Paul Gearon
>
> Disclaimer: while I've never actually run Fedora, I *do* work for them.
>
>