440 Davis Court #1602
San Francisco, CA 94111-2496
415 781 5700

June 6, 2000

04 00067 61 00060602

Mr. Jack Park
Street address
Palo Alto, CA Zip

Subject:   Ontologos and Ontology Need Further Review

Dear Jack,

Responding to your letter today explaining that implementation information is available on the Ontologos web site, and providing another web resource to support application of Ontologos, further review shows that more information is needed to accomplish the results you anticipate from Ontologos.

I have taken additional time, at your urging, to investigate ontology because "organize" is an important part of augmenting intelligence, which is a goal of the DKR project. Since people inherently organize information, we have to improve what people are doing so it takes less time and cost, and yields better results. Methods that focus on a narrow range of high priority needs justify high cost, such as national defense. Doug seems to be focusing more broadly on improving use of daily working information, as occurs in programming software initially, and later other work like a knee operation, digging a ditch, creating a DKR, and so on. This makes cost, including ease of use, and value of results, key criteria in assessing ontology as a method or process to augment human intelligence. Assessment requires work product, explanation of process, estimates of time spent and saved on handling daily working information from meetings, calls, correspondence, books, and activity that generates experience which impacts knowledge, performance and earnings.

When time permits, please submit the web address for obtaining implementation information, or provide the information you downloaded that illustrates implementation of Ontologos. This can help the DKR team assess architecture relative to costs and benefits.

Second, the web address you provided for...

Ontology Exchange Languages for Bioinformatics

...provides excellent background showing ontology has a variety of meanings that support DKR requirements to organize information into categories and manage relational issues announced by Eric Armstrong in the CDS requirements, now v0.8, received on June 5, 2000.

The web site, however, does not illustrate, nor explain any of the meanings it sets out for ontology.

Implementation is not illustrated nor explained for Ontologos, nor for other ontology vendors and methods. The web location provided for GeneClinics, and represented as showing the complexity of ontology work, reveals an alphabetical list of diseases. This is useful, but not sufficiently advanced to warrant notice. I sent a letter to GeneClinics requesting assistance locating their work with ontology.

The study prepared by Ontology Exchange Languages for Bioinformatics appears to be entirely theoretical. No work product is displayed to show a methodology for organizing information, no explanation is offered on how long it takes to create an ontology, nor to use it day-to-day, nor is there information on application and time saved relative to present methods.



Rod Welch