I have often said that the currently available tools for ontology
development and use are still rather primitive. I don't believe that
the field of ontology can make much further progress without going
significantly beyond them. (01)
Following is a slightly edited version of my response to some questions
that were sent to a different forum. I apologize for the duplication
for anyone who also subscribes to the iaoa-member list. (02)
In any case, I would ask them or anyone else to suggest ideas about
what kinds of tools and methodologies could satisfy the needs of the
person who was asking the questions -- *and* the needs of many others
with similar or different requirements. (03)
For the record, following is an article I wrote about what can be
done to develop better tools: (04)
http://www.jfsowa.com/pubs/futures.pdf
Future directions for semantic systems (05)
John (06)
-------- Original Message --------
Subject: Re: [iaoa-member] Need expert opinions on ontologies for
human-human interaction
Date: Fri, 26 Apr 2013 17:43:29 -0400
From: John F Sowa
To: iaoa-member@xxxxxxxxxxxxxxxx (07)
On 4/26/2013 2:30 AM, Syed Irfan Nabi wrote:
> I am looking for answers to the following 4 questions. (08)
Those are very good questions, but the answers I'll give are not
the suggested options in your list. (09)
> 1) Is there an IS (or applied ontology) perspective of ontology where
> ontology is primarily for human-human interaction (i.e. mutual
> understanding among humans) and not for computer-computer/human-computer
> interaction (i.e. specification for annotating data for interoperability
> and information extraction)? (010)
I would say that *every* ontology is primarily about and for human-
to-human communication about issues that are important for humans. (011)
All applications of computer hardware and software are designed to
support some human needs. Whenever we use ontologies in a computer
system, the ultimate goal is to make the computer systems more flexible
in adapting to human requirements and easier to use by and for humans. (012)
> According to Hepp (2007) “it is highly arguable whether formal logic
> is the only or even the most appropriate modality for specifying the
> semantics of a conceptual element in an ontology”. (013)
Formal logics are nothing more nor less than stylized versions of what
we can and do say in ordinary language. The primary reason why we use
formal logics to communicate with our computers is that they aren't
smart enough to understand the languages we speak every day. The
formal logics are *simplified* versions of language that are easier
to process by computer. (014)
An article that analyzes that point and develops its implications: (015)
http://www.jfsowa.com/pubs/rolelog.pdf
The role of logic and ontology in language and reasoning (016)
> 2) Hepp (2007) has mentioned a distinction between CS and IS researchers
> in their understanding of what an ontology is. According to him “a collection
> of named conceptual entities with a natural language definition – this is,
> a controlled vocabulary – would cont as an ontology.” How important and
> useful is this distinctions? (017)
I agree. But if you want the computer to process that ontology in
support of your computerized databases, knowledge bases, and application
programs, then you need to map it to the kinds of languages that the
computer is able to process. That means you need to simplify it to
the extent that it can be processed by a computer -- a formal logic. (018)
> 3) According to Guarino (2012) “An ontology is first of all for understanding
> each other … among people first of all” and that “ontology is invariant,
> necessary information”. Also, the purpose of an ontology is to provide
> an accurate picture of the language, entities and their relationships in
> a particular domain (Corcho, Fernández-López, and Gómez-Pérez 2003). (019)
I agree with everything except the phrase “ontology is invariant,
necessary information”. I'd have to see what Nicola meant in context. (020)
> Thus, how important is it to find the relevant concepts/entities and their
> relationships than their representation using a particular tool? (021)
It should be possible to use an ontology in an open-ended number of
different ways with many different kinds of tools. But I would also say
that the currently available tools for ontologies are still primitive
compared to what we need to make ontology-based systems more useful. (022)
> 4) I am looking into the possibility of utilizing Grounded Theory inspired
> technique for ontology development.
>
> Such that the resultant ontology:
>
> 1. Is meaningful to humans.
> 2. Is useful for Human-human interaction.
> 3. Correctly interprets published research.
> 4. Relates ontology to corpus (closely follows the text). (023)
Since the phrase "Grounded Theory" is capitalized, I assume that you're
referring to some specific system. What is it? (024)
> In this case, if the ontology is represented using simple graphical
> structure using boxes and arrows, would it suffice? (025)
A graph can be just as formal and just as precise as any linear
notation for logic. Or it could be even more ambiguous than any
natural language. I don't know what you mean. (026)
When you say "simple", I suspect that you mean a notation with
very few operators and very few syntactic options. That means
that it can't say very much without becoming highly ambiguous. (027)
Following is a talk I presented in 2006 about concept maps,
which are very simple to learn, but highly ambiguous: (028)
http://www.jfsowa.com/talks/cmapping.pdf (029)
With sufficient conventions, it is possible to design concept maps
that can be interpreted correctly by a computer system. But what
happens is that you end up with a two-dimensional formal logic. (030)
But many people use those notations in a way that is so ambiguous
that nobody except the person who wrote them can explain what they mean.
It would be easier to design software that can interpret unrestricted
English than to design software that can determine what some people
mean by some of these graphic notations. (031)
John Sowa (032)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (033)
|