Re: ‘New ontology pages’ as Semantic Web foundation

Dear Alex,

you wrote
"this is a reminder that an original idea of Semantic Web is based on 
three foundations, namely, XML, RDF and ontology pages."

but this is by far not complete, much more is possible. The Semantic Web 
could also integrate VRDs (Vectorial Resource Descriptors) which can be 
perceived as members of Conceptual Spaces (see e.g.
http://luisa.open.ac.uk/publications/Dietze-BlendingReal&Virtual.pdf
or simply as describer of quantifiable resources (in the broadest sense, 
also sources for feature extraction) in arbitrary resolution.
http://www.orthuber.com/wpa.htm contain most information, with some old 
conventions and partially old nomenclature

Every VRD contains a VSI which is a HTTP URI which identifies the Vector 
Space to which the VRD belongs, and which points to a VSD (a Vector 
Space Descriptor) which contains all necessary information about the 
vector space, also about the distance function for similarity 
comparison. This "Similarity" is well defined and does *not* need 
explicit creation of links and Vector Spaces can be used to expand the 
Web of Linked Data, both concepts could enrich each other.

All the best

Wolfgang Orthuber



Alex Abramovich schrieb:
>
> Hi all,
>
>    This is a reminder that an original idea of Semantic Web is based 
> on three foundations, namely, XML, RDF and ontology pages. Web 
> resources’ content must be duplicated in the machine-readable form 
> (ontology page); RDF will link a total Web content into one semantic 
> network (Semantic Web); intelligent agents, defined on this semantic 
> network, will serve the Web visitors.
>
>    As it seems to me, ‘Berners-Lee at al’ expected that Web resource’s 
> owners will write ontology pages themselves. Unfortunately, their 
> expectation failed.
>
> Why?
>
>    On the one hand, Web resources’ owners don’t ready to pay more for 
> their Web resources maintain, and any Web design’s complication 
> conflict with real tendency of the simplification (and even 
> automation) of Web resources’ construction.
>
>    On the other hand, ‘Berners-Lee at al’ didn’t provide both any 
> unified scenario for the ontology pages creation and connective 
> semantic mechanism.  
>
>    The ontology pages idea was rebranded, as rightly observed Dan 
> Brickley, into other variations on the theme, related to Linked Data 
> and such instruments as RDFa, GRDDL etc., which aim to integrate RDF 
> more closely into user-facing Web content.
>
> As a result Semantic Web engineers must link now more than 13 billion 
> RDF triples and unknown quantity of independent ontologies.
>
>    I suggest returning to the original ontology pages’ idea based on 
> the new knowledge representation language Need Language (NL). 
> Herewith, I assume that Web publisher is extremely interested in 
> success of the publication, but he is against additional and 
> unmotivated expenses on the maintenance of Web resource, and also he 
> does not wish to penetrate into additional technical details.
>
>    You know that any Web publisher has in mind a satisfaction of a 
> certain need of Web visitors. The main problem is to detect what need 
> exactly may be satisfied by the given Web resource.
>
>    NL based engine will provide a query-answering session with both 
> any Web publisher and any Web visitor using their professional or/and 
> everyday slang.  
>
> As a result Web resource’s content and Web visitor’s specification are 
> represented the same semantic marked syntax, and NL based engine will 
> get an opportunity to find for the visitor an appropriate Web 
> resource. If except a need description Web publisher provides a way of 
> this need satisfaction, we will get an opportunity to meet Web 
> visitor’s need directly or to compose a new way of the given need 
> satisfactions using available need-resources.
>
>    In other words, I mean that Web publishers will rewrite (in the 
> scope of the query-answering session) their published information in 
> the new specific form (or input a description of certain need’s 
> satisfaction) that *includes all necessary constructive elements 
> including documents and audio/video data* in the corresponding places 
> of the new presentation of their Web resource.
>
>    As a result Web visitor will be relieved of necessity to look 
> through Web content in search of relevant information. He will deal 
> mainly with Web of needs. System engine will interview Web visitor and 
> find or generate the actual way of the given need satisfaction. 
> Herewith, system engine will demonstrate to the customer *only those* 
> documents and audio/video resources, *which* *are related to the found 
> way of the given need satisfaction.*
>
>    Optional, Internet provider supplies Web of needs in form of 
> configurations ordered by the customer that will allow to the 
> governments to regulate the information flow.
>
> Neither Web publisher nor Web visitor will be obliged to know 
> something else except the particularities of their needs. They will be 
> interacting with Web of needs using their professional slang.
>
>    With the purpose of realization of mentioned above possibilities in 
> Need Language I have formalized the representation of domain knowledge 
> and defined both commonsense knowledge and commonsense reasoning.
>
> I need your opinion, your advice, your help and your cooperation.
>
> All the best,
>
> A.Abramovich   
>
>

Received on Saturday, 6 February 2010 17:41:35 UTC