NetNews Usenet Archive 1992 #20/NN_1992_20.iso/spool/comp/database/6687

Internet Message Format | 1992-09-11 | 9.9 KB

Path: sparky!uunet!ogicse!pnl-oracle!news
From: [e-mail]
Newsgroups: comp.databases
Subject: workshop - call for participation
Message-ID: <[e-mail]>
Date: 11 Sep 92 20:35:17 GMT
Article-I.D.: oracle.1992Sep11.203517.15892
Sender: [e-mail]
Organization: Pacific Northwest Laboratory
Lines: 217



                    CALL FOR WORKSHOP PARTICIPATION
                       SDM 92, A Planning Workshop

                Scientific Data Management Workshop, 1992
                           The Role of Metadata
             in Managing Large Environmental Sciences Datasets

                   Salt Lake City, UT, November 3-5, 1992


Sponsored by:
      the National Science Foundation (PENDING),
      the U.S. Department of Energy, and
      the U.S.  Environmental Protection Agency.

SDM 92 is the successor  to the NSF Workshop on Scientific
Database Management held March 12-13, 1990, at the
University of Virginia.

The purpose of this workshop is to bring together computer
science researchers and environmental sciences data
management practitioners to consider the role  of metadata in
managing large environmental sciences datasets.  The
objectives include establishing a common definition of
metadata, identifying categories of metadata, defining
problems in managing metadata, and defining problems related
to linking metadata with primary data.

Workshop participants will be selected on the basis of brief
position papers addressing the objectives of the workshop.
Review of position papers will be conducted by the workshop
organizers and volunteers from among the sponsors and invited
key participants.

The workshop agenda is appended to this call and will include
selected presentations from the position papers and invited
participants.  These will cover both the computer science and
environmental sciences data management perspectives.
Invited talks will cover:
    1)     Computer Science View and Summary - An overview of
             the current research into metadata management and
             related topics.
    2)     The Environmental Science Enterprise Summarized - An
             overview of environmental sciences research
             practices, particularly with respect to how
             environmental sciences data is managed.

Selected position papers will be organized into three viewpoints:
    1)     User Scientific Data System Experience Reports -
             Presentations describing specific environmental
             sciences data management systems.  Includes a
             description of data levels, data structures,
             relationships between data and metadata, data
             volumes, and rates.  Issues related to data collection
             and dissemination.
    2)     MetaData Standards & the User View - Presentations
             defining metadata, identifying categories of metadata,
             identifying any existing standards, outlining possible
             new standards.
    3)     MetaData Management - Presentations that describe
             current technology for scientific data management
             enterprises, describe research areas that address
             issues of scientific information management, identify
             scientific metadata usage in current technology,
             discuss scientific data models and contrast with
             business data models, and discuss the question:  is a
             scientific information enterprise a different kind of
             information business with unique problems?

Following each set of position paper presentations, the
participants will break into small working groups to define
pertinent issues and propose courses of action that address
those issues.  Each working group will report its findings to
the full workshop before the workshop proceeds to the next set
of presentations.  The final activity of the workshop will be to
identify research problems related to metadata and planning
for documenting the results of the workshop.  The goal is to
provide guidance to a major sector in the research community
by focusing attention on metadata research issues that will
directly support the real national issues in cleanup and
management of our environment.

The results of the workshop will be documented in a technical
report edited by the organizers and issued by Battelle / Pacific
Northwest Laboratory(*), an article summarizing the results
submitted to an element of the computing literature selected
by the participants (e.g., IEEE Computer or SIGMOD Bulletin),
and through a panel session to be proposed for the SIGMOD
1993 conference.

Metadata, i.e. data about data, is an important element of
scientific data sets.  Traditionally metadata might be thought
of as the notes in the lab notebook that accompany the record
of the primary data, i.e. the data values observed by the
scientist.  The automated collection or generation of large
datasets, i.e. terabytes and up, requires formal means of
capturing and managing metadata.  Without integrated
metadata these large datasets are difficult to use because
scientists cannot establish the context within which the data
were collected, generated or analyzed.  It is the metadata that
establishes that context.

Current activities and practices in the environmental sciences
provide specific examples of the importance of metadata.  In
Atmospheric Science, for example, modern field programs are
typically large, multi-discipline, multi-agency, and multi-
investigator efforts with land, sea, air and space based
instrumentation operating under a wide range of conditions.
Data which documents the platform and instrument conditions,
the operational environment and interfering sources of noise
are critical to primary data analysis.  These data are all
metadata and underscore the importance of developing
standard approaches to recording and handling metadata as a
critical dataset in its own right, and in linking metadata with
the primary data.

There is little discussion in the literature of metadata in the
context of scientific data and its management.  Metadata is
primarily discussed as information about the structure of a
database, rather than as information about the contents of the
database.  There is a need to focus attention in the Computer
Science research community on the problems of dealing with
the latter type of metadata in an integrated manner.  The
expectation is that new approaches to managing large
scientific databases with integrated primary data and
metadata will enable the scientific research activities that
depend on such data.

_TOPICS_    We solicit papers describing original ideas and
new results on theoretical and practical aspects of all forms
of scientific data management.  Major themes to be covered
include, but are not limited to:

      1)  What is (the logical data model for ) metadata?
              =>  Where does metadata come from?
              =>  How is metadata (to be) generated?
              =>  How is metadata (to be) collected?
              =>  What is the taxonomy of metadata?
                     -  associated with the source of primary data
                     -  associated with the context of primary data
                     -  associated with abstractions of primary data
                           (e.g., reduced views, visualization,...)
      2)  Problems in managing metadata.
              =>  How should metadata be stored
                      (with or separate from primary data)?
              =>  How metadata is (will be) used?
      3)  How to maintain the relationship between metadata
              and the primary data?
      4)  What is the value-added of metadata?  Conversely,
              what is the cost of not having metadata?
      5)  Discussions of practical experience in acquiring and
              managing metadata.

_PAPERS_   Authors are invited to submit 3 copies of a short
position paper of at most 2,000 words by mail, fax, or e-mail
before September 25, 1992, to:

   Ronald B. Melton                  David E. Maier
   Pacific Northwest Laboratory      Oregon Graduate Institute of
   PO Box 999, m/s K1-74                 Science and Technology
   Richland, WA  99352  USA          19600 N. W. Von Neumann Drive
                                     Beaverton, OR  97006-1999  USA

   Tel:   [phone]             Tel:   [phone]
   Fax:   [phone]             Fax:   [phone]
   Internet:   [e-mail]     Internet:   [e-mail]
   Omnet:   r.melton

_PROCEEDINGS_   The proceedings of SDM  92 will be published
by Battelle / Pacific Northwest Laboratory(*) and will be
available following the workshop.  Requests for proceedings
may be sent to:

   Mike DeVaney
   Pacific Northwest Laboratory
   PO Box 999,  m/s K1-87
   Richland, WA  99352  USA

   Tel:   [phone]
   Fax   [phone]
   e_mail:  [e-mail]

_ORGANIZING COMMITTEE_

   D.M. DeVaney           R.B. Melton
   D.E. Maier             J.J. Thomas

_SPONSORS_

   National Science Foundation (NSF)
   United States Department of Energy (DOE)
   United States Environmental Protection Agency (EPA)

_IMPORTANT DATES_

   Deadline for submission:                  September 25, 1992
   Notification of acceptance/rejection      October 5, 1992
   Workshop                                  November  3-5, 1992

_STIPENDS_

Academic participants may request stipends for travel and per
diem.

The granting of stipends is dependent on approval of the
supporting grant from NSF.

Academics requesting a travel and per diem stipend must
indicate with their position paper submission if their
participation depends on the stipend.
______________________________________________________________

*  -  Pacific Northwest Laboratory is operated by Battelle
        Memorial Institute for the United States Department of
        Energy under Contract DE-AC06-76RLO 1830.