I think that once one starts to look at what terms would need to be in
the foundation ontology for any industry vertical effort, there are sets
of terms that suggest themselves as being needed, and many of these may
themselves be governed by some suitable authority already. (01)
As an example, in creating an ontology for the financial services
industry (focusing on securities terms and definitions) I needed to have
an upper ontology for the various more basic meanings that are
specialised for existing standard securities terms. These included terms
for time (for schedules), Legal, Geography (for countries, which among
other things are issuers of certain sort of security) and so on. One
area of particular interest was basic accounting terms, and these are
handled by the XBRL standard. XBRL is interesting because it has what
they call a taxonomy, but this is really a kind of ontology (or several
jurisdiction-specific ontologies) implemented in XML using Xpath links.
This means that for us to build an ontology that captures the business
meanings behind the existing financial industry data and message
standards, we needed to draw on terms in the XBRL existing standard. (02)
In order to capture business meanings for securities I have put together
a framework of basic common terms covering financial, legal, time,
geography, mathematics, information, business/commerce and a few others.
I would expect in due course to find (and align with) terms managed by
standards bodies for each of these, in the same way that most ontologies
would expect to make use of FOAF and Dublin Core, but many of those are
either not ontologies yet, or are written in some non standard ontology
format as XBRL currently is. (03)
So I would not expect to find many unadopted sets of content for the
material that other industry-specific ontologies need to draw on. For
instance much of what is meaningful to a business is rooted in legal
meanings (I think of legal as a sort of sensory input to the business),
so there is a need for basic legal constructs like contract, law,
jurisdiction, right etc., all of which must presumably be owned and
defined by some legal community or standards body if we can find them. (04)
Where the work is I suggest, is in finding these communitis of practice
and standards bodies and assisting them where needed to provide the
business view that XML and UML and the like have not delivered for them,
and in so doing, make their basic terms available for more specialized
industries to derive meanings from. (05)
I think it would be a mistake to think of all industries as being at the
same "level" in any taxonomic hierarchy of the things they deal with,
and of there being some level above this where there is nothing. That
may be how the world is organised but it is not how the facts that
different domains deal with are related. So for example a financial
security is a kind of contract, which is a basic legal construct, but
law is an industry as well, with people who must already have
definitions for those terms. Only the level directly below "Thing" is
not looked after by any existing community or standard, and that means
maybe half a dozen terms. (06)
In this way, to follow your example, the widget community would know
that there is no sense in simply declaring Widget to be a direct
sub-class of Thing, but would look to the general manufacture
standardisation effort, find their "Man Made Object" and use that as a
parent class in their ontology (or maybe find there is a more
specialised "Factory Made Object"). This would enable them to generate
widget messages as before, but also relate these to their ontology where
the meaning of widget is formally established to as to avoid collisions
with the new Truncated Grumlet message that someone in their community
wants to add to their standard. That being the business case for having
an ontology in the first place. (07)
Mike (08)
Ian Bailey wrote:
> Hi Pat, Matthew,
>
> I'm not sure completeness and/or scope is necessarily a major problem. Some
> degree of semantic correlation can be achieved by mapping to higher levels
> in the specialisation hierarchy - e.g. your ontology might not have "widget"
> in it, but it does have "man made object", so I can simply add widget to the
> ontology as a subtype of it, and grow the scope of the ontology as I go on.
> We followed pretty much this process with IDEAS, using DoDAF, MODAF and
> DNDAF as the sources which fed into IDEAS.
>
> The problem we encountered is one of usage - i.e. how the users and vendors
> interpreted the standards (in this case DoDAF, MODAF and DNDAF). In SC4
> alone, there are a lot of models which are similar, but often used
> differently. First of all, each application protocol has used the SC4
> integrated resource models differently (occasionally, there is commonality
> between APs that had to interoperate). Secondly, different implementations
> of the same AP often interpret it differently...then the users often add
> their own interpretations and workarounds.
>
> My point is that you can't do a proper semantic analysis by looking at the
> data models alone - you need to get your hands dirty and look at the legacy
> data. You can carry out academic exercises mapping models and doing gap
> analyses, but these never work when it comes to the real world. And this is
> just for SC4, where the models are very well defined (if a little byzantine
> in places). I made a living for a number of years hoovering up and mapping
> legacy data - I quickly learned to ignore the data models and look at what
> the users had actually put in the tables. When I first read Chris
> Partridge's book, it was his forensic approach that really appealed to me as
> a practitioner. It took me a bit longer to believe in all that ontology
> nonsense mind you ;)
>
> Another point, and perhaps this is specific to SC4, is that engineering data
> seems to lend itself very well to extensional ontologies - there are lots of
> physical things, and types of physical things to deal with, and precision in
> identification is crucial. SC4 already has an extensional ontology in
> ISO15926. I don't do much work outside of the engineering domain so I'd be
> interested to know if anyone has similar feelings about what types of
> ontology suit other domains the best - e.g. finance, HR, etc.
>
> Cheers
> --
> Ian
>
> -----Original Message-----
> From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx
> [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Matthew West
> Sent: 10 February 2009 08:56
> To: '[ontolog-forum] '
> Subject: Re: [ontolog-forum] standard ontology
>
> Dear PatC,
>
>
>> There is one problem I have with this effort at using ontologies to
>> help
>> define or organize standards; there is as yet no standard for an
>> ontology
>> that has a enough content to actually represent all the things that the
>> non-ontology standards talk about. Apart from the droll situation of a
>> community without a standard trying to formalize the knowledge of
>> communities with a standard, there is the practical question of whether
>> we
>> intend to recommend one foundation ontology as the basis for the
>> formalization, or take a hands-off position and let a thousand
>> incompatible
>> flowers bloom? I don't recall whether there is a consensus on this
>> point.
>>
>
> [MW] The way I would see it, it is about who is responsible for what, so
> rather than having a thousand ontologies for units of measure, we get BIPM,
> or the relevant ISO committee they develop their standards through to
> develop it, and the rest of us just use it.
>
> Regards
>
> Matthew West
> Information Junction
> Tel: +44 560 302 3685
> Mobile: +44 750 3385279
> matthew.west@xxxxxxxxxxxxxxxxxxxxxxxxx
> http://www.matthew-west.org.uk/
>
> This email originates from Information Junction Ltd. Registered in England
> and Wales No. 6632177.
> Registered office: 2 Brookside, Meadow Way, Letchworth Garden City,
> Hertfordshire, SG6 3JE.
>
>
>
>
>
> _________________________________________________________________
> 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
> To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>
>
>
> _________________________________________________________________
> 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
> To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>
>
>
> (09)
--
Mike Bennett
Director
Hypercube Ltd.
89 Worship Street
London EC2A 2BF
Tel: +44 (0) 20 7917 9522
Mob: +44 (0) 7721 420 730
www.hypercube.co.uk
Registered in England and Wales No. 2461068 (010)
_________________________________________________________________
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
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (011)
|