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

Dear Paola,

As a mathematician I am relatively new to the field of Semantic Web technology, but having a deeply ingrained interest in anything related to "reality modeling" (e.g. quantum reality) and knowledge modeling I am convinced that a "one size fits all" approach will not work.

If we just stick to the general fields of science and technology we will encounter dozens of clusters of interrelated disciplines with sometimes common but more often slightly different fields of formal and informal concepts, definitions, theories, terminology and bodies of work derived from these.

For our global project we have been forced to survey various fields and ascertain the state of the art in Semantic Web technology development and implementation, and most importantly the standards developed for the respective fields.

What we have seen so far is that the health sciences field in particular yields some of the most promising approaches in getting to what is a unifying common denominator set of tools to start out with.

In particular the approach used in the OpenEHR (www.openehr.org), based on archetypes and templates is the closest thing we have seen to a knowledge modeling paradigm in which a restricted language domain is used.

For those familiar with natural language processing and formal logic systems for linguistic research, the archetype/template approach is a novel approach, and actually more related to the fields of cognitive science and psychology.

To get back to the issue at hand, as I have pointed out earlier in personal emails to various subscribers to W3 lists, templates are a promising field of research, in that the archetype and templates represent excellent atomic or primal concepts to start out with in knowledge and reality modeling, which are at the heart of semantic web technology.

The OpenEHR initiative realized that archetypes and templates are the way to go, and their work deserves more attention in the sense that a generalization of archetypes and templates needs to be investigated.

If we just look at two required ingredients for the success of semantic web technologies, browser and search engine capabilities for handling semantic content, we will see that archetypes and templates are actually a very good hands-on approach to dealing with information extraction and formatting.

Too much effort in the field of the semantic web goes in trying to create a single uniform approach with monolithic technologies that ignore the reality that humanity is divided by cultural and historical heritage, language and other societal factors, which become most acute in the bodies of work laid down in written language.

Archetypes and templates are a clever way of getting around a lot of the problems and yes you may not be far off at all that "reasoning parsers" for browsers and search engines may just be a way to get around of the problems.

As web users we "personalize" our web pages in domains like MySpace, Facebook, Yahoo, Hotmail, on professional web sites etc.

Why not personalize our browsers and the search engines we use by having them use "reasoning parsers" to extract the information we need?

Now here comes the mind blowing part!!! We can set up web pages by tagging them in some way to be processed for information extraction and semantic content generation in such a way that digital repositories on the web will actually identify these sites as part of a "vernacular" form belonging to a particular category using ontologies out of a particular "language group" (the Yahoo Category versus the Google Brute Force approach to indexing information).

Thus we acknowledge the fact that we live in a Babylonian language confusion dominated world, yet are still able to make sense of it all, in the way intended by the creators of the Semantic Web.

Templates and archetypes are fresh ideas and using reasoning parsers that can pick up which vernacular from which language domain group is being used may prove very useful!

In such a context the question if RDF can be useful in outputting triples that make sense can be rephrased to how we output triples that make sense in predefined vernaculars or language domains (language here not being the linguistic, but the field of scientific discipline /technology type).

We submitted an idea Project 10 to the 100 at Google in which we outlined using "tag-templates for semantic content detection and information extraction purposes" and it is funny to see that Yahoo and Microsoft are jointly working on "category based Semantic Web Search".

They must be on to something!

Milton Ponson
GSM: +297 747 8280
Rainbow Warriors Core Foundation
PO Box 1154, Oranjestad
Aruba, Dutch Caribbean
www.rainbowwarriors.net
Project Paradigm: A structured approach to bringing the tools for sustainable development to all stakeholders worldwide
www.projectparadigm.info
NGO-Opensource: Creating ICT tools for NGOs worldwide for Project Paradigm
www.ngo-opensource.org
MetaPortal: providing online access to web sites and repositories of data and information for sustainable development
www.metaportal.info
SemanticWebSoftware, part of NGO-Opensource to enable SW technologies in the Metaportal project
www.semanticwebsoftware.org


--- On Thu, 11/13/08, paola.dimaio@gmail.com <paola.dimaio@gmail.com> wrote:
From: paola.dimaio@gmail.com <paola.dimaio@gmail.com>
Subject: Re: Use of Content Management Systems, ranking and usage in Semantic Web
To: metadataportals@yahoo.com
Cc: semantic-web@w3.org
Date: Thursday, November 13, 2008, 12:54 PM

Milton

it's an important question, but the answer is non linear

I am convinced that when CMS (and blogs) will output their stuff as
RDF, then the semantic web will be a done thing.

but, assuming such a functionality exists,  we (me anyway) still have
not quite understood  how to model RDF into meaningful content
structures

 shall we use metadata structures to infer RDF triples? or do we use
reasoning parsers  (thats a new technology I just made up) to digest
content and output triples that make sense?

>
> In order to transition from plain vanilla web sites with run-of-the-mill
> content management systems (CMSs) to semantic web enabled sites with
> corresponding CMSs, a lot of functionality has to be added.

yes, but....
in a content management context we need to define better what we mean
by semantic. In CMS  logic (my take) semantic = relational

if so, then all CMS have semantic capabilities, in principles, because
they all use relational dbases, therefore the relationships between
objects just need to be defined, and  the functionality that needs to
be developed is just some kind of user friendly interface to ' let a
relate to b' and output content to XYZ schema

I was just looking for Dan Brick's profile on Facebook, and thinking,
how on earth is Facebook doing the friends networks without using
FOAF? its loose relational technology a bit augmented perhaps to
capture expanded value sets, I think

if however, by semantic you mean CMS that can produce content
published as 'rdf'  format,  then to derive RDF schemas from HTML
content  looks relatively trivial, is just a schema mapping exercise

The question for a cms user would be to devise logically and
meaningfully consistent RDF schemas - that's where I got stuck last
time I looked at this issue, a couple of years ago or so.


> We are trying to find out if any reports exist on the ranking and usage in
> Semantic Web context of CMSs and also if comparison tables exist, e.g.
> listing SW standards, metadata sets and other items compared for the CMSs
> and also, who are the users of such CMSs?

so many questions in one sentence...
AFAIK , all CMS can be set up to output content using ANY schema, as
long as the schema is defined, therefore, in principle, all CMS s have
semantic capabilities, provided they have been configured to do so

When I requested an RDF feature for Drupal I was thinking of a module
that could translate content into RDF, not sure how Drupal is setting
up for RDF,   havent looked into it yet. Other nice functinalities
could be 'visualise' and manipulate queries
>
> Many non-profits are migrating to web sites with CMSs, of which Plone,
> Joomla and Drupal seem to be quite popular.
>


> But we feel that the popular ranking and usage of available CMSs does not
> necessarily reflect the ranking as per SW functionality.
> Another issue of particular importance for users of such sites is browser
> add-ons and plugins for semantic web technologies.

once the functionality is defined, its relatively easy to implement
it, what we are having difficulties in understanding is what exactly
you would consider a requirement in this arena


> Are there any reports,list, directories and comparison charts for
currently
> available browsers, listing available plugins and add-ons?


Please let us know what you find out,

best
PDM
>