WPC2 2B Z ?0I Meeting Notes X3L8 12/5-7/94 Bruce BargmeyerBruce Bargmeyer   HP LaserJet IIIHPPCL5MSXx6X@X@ЫXx6X@X@< 9Z .Courier New RegularX2(03|x Meeting Notes X3L8 Special Interest Group Meeting Held at the National Institute of Standards and Technology July 1921, 1995  Attendees: Bruce Bargmeyerhh#EPA Larry Fitzwaterhh#EPA Dan Gillman hh#Bureau of Census Paul Hawes hh#Bellcore William Kenworthey, Jr.hh#Self Judith Newton hh#NIST Phong Ngo` `  hh#SAIC Bert Parker hh#MITRE Jim Pipher hh#DISA David Stowell hh#Stowell and Associates, Inc. The following reports the issues raised and decisions made: 1. ISO/IEC 11179, Part 1, Framework Dan will have a new version of this part next week. The draft will be put up on the FTP server. It will be sent to the WG4 members by Phong Ngo. A cover letter will be developed with Bill Kenworthey, convener of WG4. As input to Dans editing, Bruce Bargmeyer will put the Buzios agreements on Part 2 onto the X3L8 FTP server. 2. ISO/IEC 11179, Part 6, Registration Comments were prepared for the CD ballot on Part 6. The SIG participants recommend that the comments be sent out with approval of the X3L8 committee. The X3L8 members present voted unanimously in favor of the changes. Phong Ngo will write up the comments and distribute them on the X3L8NEWS reflector. 3. Presentation about EPA Data Registry Larry Fitzwater made a brief presentation on EPA plans and directions. The presentation engendered lively debate on several issues. There is a need for the registry to go beyond data elements. X3L8 may want to step up to some of these issues, such as extension of the metadata from data elements to datasets and data systems. 4. X3L8 Home Page Frank Olken submitted the X3L8 Home Page URL to several WWW indexing sites. He also made several changes per comments from X3L8 members. Potential additions were discussed. E.g., FAQs and something about the upcoming metamodel review. As Paul prepares the materials, we will put them up on the net. 5. BSR BSR workshop was held in Geneva, Switzerland last week. A package was sent out in advance. Many messages were exchanged about package materials before the meeting. Paul Hawes reported on what he had learned about the meeting. We are concerned about several aspects of the BSR development. The BSR is going to attempt to put together a metamodel in a very short period. We will attempt to contact David Gradwell, who is working on developing a metamodel for the BSR project. We want him to be aware of the X3L8 metamodel and the fact that it addresses many of the issues raised at the BSR Workshop. Judy Newton will send a message to him and Paul Hawes will followup. 6.ISO/IEC 11179, Part 2 Bill Kenworthey discussed Part 2. It is agreed that we need to more fully develop a property taxonomy. An example would help to show how the property taxonomy works and how it plays in the metamodel. We discussed sorting the representation terms from property terms where the term does not serve as the property. 7. Document review 7.1 ISO/IEC N 776, Computational aids in terminologyData element categories. We discussed this document and find that it could be useful as a possible source of attributes for the metamodel. Where we already have the attributes, Paul will check them off in the document. We can then discuss whether we need any of the others. We do not have comments on the document in terms of its use for terminology. We do not particularly like the definition it contains for data elements. 7.2 ISO/IEC N 779, Terminology Work Principles and Methods. This document has material that appears to be useful in the development of a taxonomy for the informative annex of 11179 Part 2. The editors of Part 2 are encouraged to utilize it as input. We did not have comments to forward to ISO/TC 37/SC1 on the document. 7.3 SC21 N9344 CD 13237 IRDS Content Module to support a Naming and Thesaurus Facility. Judy Newton commented on this document at the request of David Gradwell. It is apparent that the metamodel in this document is working its way into the BSR work. We believe that the X3L8 metamodel should also provide a foundation for development of the BSR schema. 7.4 N 784 ISO/IEC WD 14957, Information technologyRepresentation of data elements values: Notation of the format (character types and lengths) Additional notation for figures The SIG recommends forwarding the following vote and comments. Disapprove Question 1 Consideration of Registration as CD. And Disapprove Question 2, Consideration of CD. Comments: This standard could be particularly useful in establishing a specific content for the values of the ISO/IEC 11179 Part 3 basic attributes Form of Representation, Datatype of Data Element Values, Maximum Size of Data Element Values, Minimum Size of Data Element Values and Layout of Representation. It could also be useful in the metamodel work underway in X3L8. This would assist interchange of metadata between data registries as well as assist use of registry information by intelligent software. However, as written, the proposal appears to put all of the above and more attributes together in a single multifact attribute. It makes it into a data composite which is not good practice for design of data elements or attributes (although it reflects common usage in program languages). The draft is at odds with ISO/IEC IS 11179 Part 3. While we support the intention of this document, this draft needs considerable work before it is registered as a CD. At present it is incomplete in its coverage, inconsistent within itself as well as not conforming to ISO/IEC IS 11179 Part 3. The following are a some of the problems we find in this document. In Section 3, the definition of data element, which is attributed to ISO/IEC 11179 is not the definition found in the IS parts. The definition taken from part 3 is: data element: A unit of data for which the definition, identification, representation and permissible values are specified by means of a set of attributes. In Section 6, decimal computable format specifications uses the decimal point as the sign for decimal whereas section 5.3 uses the comma as the symbol for the decimal point.Also in section 6, the example D12.3" is uninterpretable or wrong as given. The table at the bottom of page 6 only partially addresses the issue of numeric base. Many other base numbers are possible. Linking a letter to a numeric base is confusing particularly when the same letters used in other contexts mean other things. It becomes very hard to interpret. A more general scheme could be proposed which would cover any numeric base. The document often refers to coma where it should use comma (at least in American English, a coma is a state of unconsciousness, not a punctuation mark). Due to time limitations we could not develop an exhaustive set of comments. Therefore, we did not want indicate that fixing the above would change our vote to yes. Vote of X3L8 members 6 for, none against. 7.5 N 785 Summary of letter ballot for CD registration of ISO 6523 Information technology Structure for identification of organizations and organization parts We briefly reviewed this document which shows that the CD registration passed. Phong Ngo will review this document in more detail. If he finds anything requiring out comment, he will circulate it via the reflector and taking into account any comments, will forward our comments. 8. Discussion of Metamodel for Data Representation We resolved several issues and created a few. For the particulars, see the file issuesmm.wp6 (or .wp5 or .txt) on the X3L8FTP server. See end of report for info on X3L8 FTP server. In attempting to quickly resolve the easier issues, we keep running into more fundamental issues. We decided to prioritize the issues. To do this, we first reviewed the purpose and use of the metamodel. The following are statements by members of the SIG as to their interests. A. What is the scope, purpose and use of the metamodel? 1) The scope and purpose is: ` ` ` a) Schema for a data registry (run by registration authority ` ` ` b) System Engineering ` ` ` c) Facilitation of human understanding` ` ` d) Facilitation of intelligent data processing ` ` ` e) Data Standardization, i.e., metadata management ` ` ` f) Data modeling (common understanding of essential DE structures) (data element in context) ` ` g) Schema integration ` ` ` h) Support federated data bases ` ` ` I) Support warehouse ` ` ` j) Data element conceptualization (how you think about data elements in a common way) ` ` k) Facilitation of data interchange ` ` ` l) Metadata analysis (e.g., a place to find things about the design of a survey; the ability to do queries metadata in a repository) ` ` ` m) Provide routine entry points to metadata ` ` ` n) Routine acquisition of supporting metadata ` ` ` o) Scope of metamodel: metadata routinely required to characterize data ` ` ` p) Redundancy equivalency analysis ` ` ` q) Structure for recording what we know about data ` ` ` r) Facilitate exchange of metadata between CASE tools and repositories. ` ` ` s) Maintenance of meaning over time and distance 2) Users of instantiations of the metamodel: ` ` ` a) EDI data element designer ` ` ` B) Application System Developers ` ` ` c) Case tool developers ` ` d) Data registry developers ` ` e) Data element designers ` ` ` f) Software trying to do schema integration ` ` ` g) People trying to find data ` ` ` h) People trying to understand data ` ` ` I) People/systems trying to formulate queries ` ` ` j) People who share data within and among organization ` ` ` k) data Administrators ` ` ` l) Data modelers ` ` ` m) Enterprise data architects ` ` ` n) End users of data ` ` ` o) Subject matter experts (e.g., capturing the knowledge of) ` ` ` p) Data stewards Suggested categories of users: Designing data Finding data Understanding data ... Three broad scopes were defined: ` ` ` Abstract metadata about data elements considered at the schema level. At this level the descriptions can be used in many applications. ` ` ` Application metadata about data elements and the data collected in particular applications. This includes the abstract scope and adds additional metadata covering such things as the purpose for collecting the data, the statistical methodology used, and other information needed to understand and use the data element values found in one or more particular applications. ` ` ` Repository a more complete set of metadata including directory (locational) information, structure information needed for formulating queries, etc. This scope includes metadata (related to data elements) that conforms to broader notions of repository technology. Paul will use this information to prioritize the issues and to write the front matter describing the metamodel. X3L8 will address the issues in the priority order established by Paul. ` ` ` 9. Future work Phong Ngo lead a discussion to elicit from SIG members their interests in future work. Interests: Data dissemination on the Internet (capturing the metadata that needs to be disseminated with the data) System interoperability data sharing EDI developing EDI data elements, transaction sets and representing the telecommunication industry Efficiency in software development through metadata reuse Further work on Object taxonomy Further work on Property taxonomy Further work on qualifiers Further work on connectors (relationships) Multimedia accommodation (representation) Data Element Concept attribution Domain research Upgrading of metamodel G7 data interchange (Bruce and Judy will send Phong some papers to distribute) Domain Ontology concept and creation More fully describe the structure and content of databases (extend beyond data element) Underpinnings for message standardization (semantics and fundamental structure of content) Federated database and Intelligent Integration of Information Data warehouse Toxic site cleanup data interchange Extend into domains, understanding of domains Metadata attributes beyond the data element (datasets, subject areas, sub data element) Basic research in data structure break into discrete areas in which a common set of rules and attributes can be used and propagated (analogs: music notation, blueprints) Developing rules for common social aggregates of data (e.g., health, environment) Interconnecting heterogeneous databases (federation, common schema, I3) Migrating data from legacy to target systems Constructing intelligent command and control systems (C4) Fusing the disparate results of queries to varied databases EDI, BSR, Mapping and geodesic data Environmental Data Registry (Internet access, EDI, Data Administration) Portable system integration at conceptual level reusability between tool sets and methodologies. Standards for conceptual and logical model levels (apply to Finance, Healthcare, Environment...) Warehouse EDI Intelligent Integration of Information/Intelligent Information Services Standardized domains (code sets) Dynamic metamodel Object oriented extensions Permissible values in domains 10. New work items Phong Ngo will develop the two work items discussed at the last SC 14 meeting. He will bring them to the next X3L8 meeting and send them out on the X3L8 reflector. These are to be new and not part of ISO/IEC 11179. Volunteers for data element concept Phong Ngo. For Domains, Judy Newton. 11. Description of EDI organizations In discussing interests, we discussed EDI standards activities. Bill Kenworthey described the following EDI efforts. The structure that Bill laid out on the board is not preserved below: One effort: United Nations, Working Party 4, EDIFACT Working with ISO Headquarters Central Secret to help on EDI > BSR Related effort: ISO/IEC JTC1 SC30 EDI US TAG to SC30 Open EDI Related effort: X12 EDI 12. Future X3L8 SIG Meetings and X3L8 Plenary meetings: Meetings marked NIST are held at the National Institute of Standards and Technology Gaithersburg, Maryland Meetings marked ITI are held at the Information Technology Industry (ITI) Council Second Floor 1250 Eye Street Washington, D.C. Note: The following meeting was listed as being at ITI, but the rooms listed were at NIST. The meeting is to be at NIST. September SIG and X3L8 Plenary meetings at NIST: X3L8 SIG Meetings: Wednesday, September 13 Building 225, Room B157, 9:00 a.m. to 4:30 p.m. No dining room for lunch Thursday, September 14 Building 225, Room B157, 9:00 a.m. to 4:30 p.m. Admin. Bldg., Dining Room A X3L8 Plenary: Friday, September 15 Building 225, Room B157, 9:00 a.m. to 4:30 p.m. Admin. Bldg., Dining Room A X3L8 SIG: October 1011 at ITI SC14 and SC14/WG4 October 2327 at Paris, France X3L8 SIG: November 30 December 1 at NIST X3L8 SIG: February 78, 1995 at NIST X3L8 Plenary: February 9, 1995 at NIST X3L8 documents, including this meeting report, are available via Anonymous FTP from speckle.ncsl.nist.gov. Change directory to x3l8 (X3L8 in lower case letters) and check the "readme" file for details.