Re: [Semantic_Web] Re: web to semantic web : an automated approach

Yeah, you are right about that but you have to put the current technology
into perspective here, everything is a nothing but a set of rules.

Ontologies are restrictive no doubt about that but how do you plan to
structure your rdf around, you can call an ontology a set of constraints and
classes which help the machine, understand the data in a better manner,
somewhat and they are required for a reasoner.

Ontologies merely a set of classes/concepts within a certain domain and
relationships between, sort of a tool  for reasoners to understand and
reason better.

thats what any NLP/AI tool does they have set of rules and through which
deduce/ reason, they will be restrictive as well.

So more the set of rules or more granular the ontology the less restrictive
it is and harder it is to maintain the quality.

One thing, I guess we need to understand here, all the semantics we try to
introduce, they will still be bound by certains rules/ constraints, So will
be restrictive, and also classified in some manner or another by any machine

And on top of that we have a separate problem of  "Objective information"
and "Subjective" information and whatever is in between


Thats what I mean by understanding of Scope here, may be you decide not to
use ontologies or you do, the fact that remains that a decision needs to be
made on what level of granularity you would want to go, because all of this
data on the web needs to fit into a set of rules which will always be finite
or untill we are able to come up with some kind of learning technology which
creates / understands rules by itself, like our brain does in real life.

Its possible that I might not talking gibrish here :).. i am just thinking
aloud,

Cheers

-- 
Rishav Rastogi

Cell : +91 9008103913
Yahoo : rishav_work
Gtalk : rishav.rastogi
Skype : rishavrastogi


On Thu, Oct 23, 2008 at 9:32 PM, रविंदर ठाकुर (ravinder thakur) <
ravinderthakur@gmail.com> wrote:

> actually i am not a big fan of ontologies. infact i think trying to fit
> data into some ontology might be wrong. the problems are that ontologies are
> overly restirctive if they are finite and well defined and if they are
> ambiguious then they are not ontologies.
>
> eg. take the general example of one's parents. its very normal to have
> ontology with parents defined as one father and one mother but what happens
> to the case of surrogate mothers, or doner fathers/mothers etc. and what
> happens to the adopted kids. how can be their parent relationships be
> represented completely with any well defined ontology ? on the other hand
> only using RDF without any creating any relation to some associated ontology
> might be a complete system.
>
>
> if ontology is finite and well defined then it won't be complete and if its
> not well defined and not finite, well then i would say its not an ontology.
>
>
> ravinder thakur
>
>
> --~--~---------~--~----~------------~-------~--~----~
> You received this message because you are subscribed to the Google Groups
> "Semantic Web" group.
> To post to this group, send email to semantic_web@googlegroups.com
> To unsubscribe from this group, send email to
> semantic_web+unsubscribe@googlegroups.com<semantic_web%2Bunsubscribe@googlegroups.com>
> For more options, visit this group at
> http://groups.google.com/group/semantic_web?hl=en
> -~----------~----~----~----~------~----~------~--~---
>
>