Meeting Report

L8 Special Interest Group

Held at the National Institute of Standards and Technology

May 22 and 23, 1997

Attendees:

Brent Archinal, U.S. Naval Observatory

Bruce Bargmeyer, U.S. Environmental Protection Agency

Ronny Elofsson, SC 14 Secretariat

Larry Fitzwater, U.S. Environmental Protection Agency

Dan Gillman, Bureau of the Census

Alan Goldfine, National Institute of Standards and Technology

Henry Heffernan, EDPNS

Michael Huhns, University of South Carolina

William Kenworthey, Jr., Self

Douglas Mann, Battelle

Judith Newton, National Institute of Standards and Technology

Phong Ngo, Science Applications International Corporation

Burton Parker, Paladin Integration Engineering

Andy Schoka, Mitretek Systems

Dan Gillman chaired the meeting. The following reports the issues raised and decisions made:

1. Derived data

Dan Gillman lead a discussion of derived data, giving an example and raising questions about how to handle this type of data in the X3L8 Metamodel for the Management of Sharable Data. Derived data is generated algorithmically (or by some procedure) from other data elements by concatenation, filtering, categorization, mathematical manipulation, etc. E.g., data about individual persons could be stored in "age", "race", "sex", and "geography" data elements. This data might be manipulated to derive counts of individuals by age, race and sex for one or more geographic areas. The result is "derived data". The original data elements and their conceptual domains are well handled by the current metamodel. The derivation procedure might produce changes to the conceptual domains. E.g., for the original data about individuals, the conceptual domain for "sex" might contain the potentially valid values "male" and "female". If summary data is derived from the original data, the conceptual domain (for the summary data) for "sex" might be "male", "female", and "total"; where "total" is a sum of both males and females. Our metamodel needs to accommodate the changes in the conceptual domains (and their related value domains) as well as provide a place to capture the derivation algorithm (or procedure).

Some derived data could also be considered as data elements. For example, where the derivation procedure results in the production of rows and columns of data, the "derived" data could also be characterized as a new set of data elements. If the intended use of the derived data warrants it, each of the new data elements could be documented within the metamodel.

Changes were suggested for the metamodel. The changes will generalize the "data composite" part of the current metamodel. Doug Mann will revise the metamodel issues document to record the issue and the potential resolution.

Doug Mann proposed a revised outline for the sequence of the text in the dpANS X3.285 document. The revised outline was accepted. Doug will continue with preparation of the document for public review.

Doug also proposed to change some names within the metamodel. It was agreed that he should remove the words "data registration" from any names within the metamodel, since the whole model is about "data registration".

The CD ballot for ISO/IEC 11179, Part 1, resulted in approval. Some comments were received on the ballots.

Ronny Elofsson, from the SC 14 secretariat, was able to join the meeting via teleconference. He helped us work out an approach to quickly progress Part 1. Dan agreed to provide to Ronny a new version of Part 1, along with notes describing the proposed resolution of comments. Ronny will then take the appropriate SC14 actions to progress Part 1. Ronny will utilize the new JTC1 procedure. This requires that Part 1 be circulated for a "final CD" ballot and then be circulated for a brief (two month) up or down DIS ballot.

We worked with Dan Gillman to review and resolve the comments.

This document is out for JTC1 CD ballot.

Frank Olken reported on plans for the Workshop on Metadata Registration. The program committee has met to organize the structure and content of the workshop. People should watch the web pages for changes. The URL for the workshop is: http://www.lbl.gov/~olken/EPA/Workshop/

A new registration form is available. It must be filled out by June 8, 1997. This is different than the "intent" form. The new form enables participants to specify the type of room desired, form of payment, etc. See the Workshop web pages.

Burt Parker led a discussion of the status of the Date and Time standards. Burt reported that a ballot for the date standard has been issued for a final vote by X3L8 members. Upon passage, Phong Ngo will forward the date standard for approval and publication. Burt summarized editorial changes made on the time standard and thanked Brent Archinal and Peter Noerdlinger for their technical contributions and careful editing reviews. Several editorials that Burt didn't fully understand were discussed with satisfactory resolutions. Burt will complete the edits, post the new version on the FTP site and notify Phong Ngo of its readiness for ballot.

Andy Schoka brought a copy of JTC1 document JTC1 N 4631 and raised it to our attention. This document requests SC and member body to comment on JTC1 re-engineering proposals. The group developed a response. Bruce Bargmeyer will prepare it and send it out on the SC14-News reflector for response by any who are concerned. The response will then be combined with any comments received from other member bodies of SC14 and will be sent to JTC1. This is already past deadline, so it must be completed quickly.

Tom Kurihara of NIST is proposing a one-day workshop to discuss Terminology aspects of ontology development. EPA will co-sponsor the workshop. He is considering the potential date of August 19, 1997 at NIST. Interested persons will be invited to attend.

9. Joint Meeting with X3T2

We discussed the proposed joint meeting between X3L8 and X3T2. We accepted the place and time proposed by X3T2 for the meeting: August 14 & 15 in Seattle. Doug Mann will contact X3T2 to complete the arrangements. X3T2 will arrange a meeting place, probably at Boeing or IBM. Doug will request that a conference call arrangement be set up so that people can attend from remote sites. Doug is also requested to develop a tentative agenda with T2.

12. Future meeting schedule

SIG meeting, 9AM, June 18-19, 1997 at Battelle. X3L8 Plenary meeting June 20 at Battelle

Workshop on Metadata Registration, July 8-11, 1997 at UC Berkeley, California

Joint X3L8 - X3T2 meeting, August 14 & 15, Seattle. Meeting place to be announced

X3L8 SIG meeting 9 AM, September 25 & 26, 1997 at NIST, Administration Building, Lecture room C

X3L8 SIG and Plenary meetings 9 AM, October 22-24, 1997, Battelle

X3L8 SIG and Plenary meetings, February 18-20, 1988 at ITI

-------------------

X3L8 documents, including this meeting report, are available via Anonymous FTP from speckle.ncsl.nist.gov. Log in as anonymous, use your email address for a password, Change directory to X3L8 (X3L8 in lowercase letters). Read the "readme" file for a listing of filenames and a brief description of each file. Many of the X3L8 and SC14 documents are also listed on the X3L8 Home Page http://www.lbl.gov/~olken/X3L8