ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] How long to useful?

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: William Frank <williamf.frank@xxxxxxxxx>
Date: Wed, 30 May 2012 22:05:49 -0400
Message-id: <CALuUwtD3UJ-5Zbck5yMvgm-F9zkvrbEQBORZc6WfDZBKoGuKiQ@xxxxxxxxxxxxxx>
David,

In a drier tone, It sounds as if the scenario you are painting involves the following features. 

1. The information technology we are discussing is concerned with creating some kind of application. (this one is perhaps the most doubtful as a reasonable assumption, except at most big institutions, this is indeed how they mostly do think of information technology)

2. This is happening in some kind of big company bureaucracy, which as always is held captive by its IT division, (so resume-ready technology is the main goal, and the 'solutions' are technology-bound like never before) where replacing, upgrading, reengineering an application typically takes more than  3 years. (I have known it take more than a decade, under these conditions, sometimes with no end it sight, so this one is hard to dispute).

3. No sane manager in this environment is going to care about anything he does that takes more than 3 years, because he or she needs to score in a year or two.


The question you are asking then seems to be, what would the role for ontology be in this environment?

Most of us in this forum would agree that the pairing of these three features is evidence of the massive ignorance and incompetence of the institutions where they occur.  But occur they do. So, it is not directly relevant to your question that I, and many others would be happy to explain why each of these three features is highly undesirable, completely unnecessary, with well-known and well-practiced ways to successfully avoid them.   What I wanted to do first was challenge these three features as if they were "premises".   But  you did not state them as such, only as describing a scenario. 

Putting aside challenging the features, then, how could ontology be helpful in THIS scenario.  

Well, first, it seems that it would follow that no sane manager in this situation would care about any application product finishing, but would make sure to move on, before the three years were up, so that NOBODY will care whether ANY application ever finishes.  This seems to be more and more the case, and, I believe, is more the cause of the time applications take than any purported complexity of the technology.   Nobody is left to care, indeed, more and more profit by non-completion.

Putting that aside, if somebody DID care about anything getting done at these places, they would need to come to understand that

1. in any project taking a year or more, a single domain ontology, which can be produced in about a quarter, if done by experts, and then maintained, and used to underly all the different kinds of technology work, will produce the result MUCH faster than the flailing and thrashing that will otherwise occur.

2. even more importantly, as more and more of business practice is automated, without a (lower, domain) ontology, it is impossible to describe in externally understandable, translateable,  terms what business is **actually being done.**   Making it  also easier to HIDE what business is being done, to those outside the inner linguistic circle.  This is a plague to corporate governance, to regulatory compliance, and from where I stand, is obviously one of the main reasons why it was so easy for the mortgage debacle, as well as the other disasters in financial services, in pharmaceutical quality control, etc. to occur.   

In short, to paraphrase a friend, the less a company knows about "information technology", the more they emphasize the "technology", and the less they emphasize the "information".  They spend more to get less and less more and more slowly, until the business has fully disappeared into the technology, which obsoletes faster and faster,  and disappeared behind a semantic wall where the users of it sit, who can then do as they please.   I ask, how is this sustainable, and how will it end? 






On Wed, May 30, 2012 at 2:37 PM, David Eddy <deddy@xxxxxxxxxxxxx> wrote:
John -
7. Time, how long will it take, do the parties have that much time?

Way back in the early '80s when were we coming to the end of green fields systems (e.g. pretty much all the big, important, necessary functions had been automated at least once), it was observed that it took a minimum of 4 years to do a serious system... & most organizations simply did not have the attention span... sponsors change, markets change, technology fashion shifts, etc.

Thirty years later, things are far, far more complex—mainframe, midrange, client/server, web, Mobile coming, etc.  These additional layers of complexity—each with their own twists on language, jargon, slang, organization, taxonomy, etc.—do not make for moving quickly.

Plus, management's attention span is well under 3 years.  If I'm on the upward bound management track & something isn't going to punch my ticket in less than 3 years, it's not going to happen.

Just how does ontology fit into that equation?

___________________
David Eddy



_________________________________________________________________
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
 




--
William Frank

413/376-8167



_________________________________________________________________
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    (01)

<Prev in Thread] Current Thread [Next in Thread>