Bath Profile Meeting
April 6-7, 2002
OCLC, Dublin OH
Summary Discussion & Record of Decisions
1. ONE-2 and danZIG Profiles
Poul-Henrik Jorgensen gave a presentation on the features and development of these profiles. (See http://www.portia.dk/pubs/ZIG/One2AndDanzigProfilesApril2002.PDF )
2. U.S. National Z39.50 Profile for Library Applications
Bill Moen gave a presentation on the work going on with the U.S. National Profile. Functional Area A will be moving forward for ballot with NISO. There are some minor differences with Bath.
The U.S. committee is proposing three element set names for Functional Area B holdings. They also propose that this functional area be considered as a draft standard for trial use. This would allow for the community to acquire some experience with the holdings specifications before formally proposing them for ballot by NISO.
3. Functional Area B: Holdings for Bath
The discussion on holdings proceeded with the goal of reaching consensus on the specifications for the retrieval of holdings information in the international arena. Carrol had done a brief survey of librarians to determine if detailed or summary holdings were required. While some librarians indicated that summary holdings were sufficient the majority of U.S. and Australian librarians wanted detailed holdings and circulation data.
There was discussion on the specifications for holdings already defined by ONE-2/danZIG and NISO and the differences between them. For Bath the following was agreed:
- An ESN would be defined for title level holdings that would provide information on the location of the title. This is the equivalent to the minimal bibliographic level holdings defined by the U.S. committee. Provision would be made for both a location symbol (for librarians) and an institution name (for the general public). Some elements would become optional that are now considered as mandatory or excluded. These changes should not impact existing implementations and thus were considered acceptable to the ONE2/danZIG representatives. The changes to the Z39.50 Holdings ARS are:
-
Change actualBibItem to optional
Change targetItem ID from excluded to optional
Change servicePolicy from mandatory to optional
Change locationName from excluded to optional
The danZIG specification for summary bibliographic holdings will be used by the Bath Profile for the specification of summary holdings. This level will provide the type of information generally available in a physical union catalogue or union list of serials.
The U.S. definition of summary level holdings at the copy view is the basis for defining an ESN for holdings related to local systems and virtual union catalogues. Elements for CopyLendingInfo and CopyNumberOfPieces will be added.
Bath will define a list of authorities that are to be used for the InstitutionOrSiteID. It was also suggested that the name of this element be changed to make it more understandable for users. The proposal is to use LocationID since LocationName is used.
It was agreed to use the XML schema developed by Poul-Henrik Jorgensen for the holdings abstract record syntax and to develop sub-schemas for each of the proposed ESNs. The sub-schemas will be maintained on the Bath web site.
Requirements for espec-Q and reference to the holdings attribute set will be removed as part of the revision Joe and Mark will be doing. References to levels of conformance in Functional Area B will be removed and replaced with references to the specific ESNs.
Joe Zeeman, with assistance from Mark Needleman, agreed to draft the Bath Profile Functional Area B: Holdings Retrieval before the end of May. The draft will be integrated into a version 2 2nd draft and sent to participants for comment. Comments will be needed before June12th.
4. Character Set Negotiation
At Level 0 it was agreed that clients should recognize Latin-1 as common practice but be prepared to negotiate Unicode. If a character set is not negotiated, a server should assume that the characters in the search term are Latin-1.
At Level 1, both the client and server should be capable of character set negotiation. An example of how to negotiate Latin-1 will be included. Joe Zeeman and Ray Denenberg will work on this and determine what the negotiation record will look like and the OID for Latin-1.
5. Negotiation Model
Over the last couple of years, there has been discussion about the need to negotiate a specific profile in situations where more than one profile may exist. The U.S. profile had suggested the use of Behavior element OIDs for various functional areas.
Ray Denenberg discussed the negotiation model and how it works in Z39.50. The general feeling was that negotiation of behaviour models was not a good idea. As no consensus was reached, this will not be added to the Bath Profile.
6. Functional Area D: Authority Search and Retrieval
Larry Dixson walked the group through this proposed functional area. The group was very positive about this new functional area. It was agreed that it will become Functional Area D in version 2 of Bath.
7. Profile Harmonization
Slavko Manojlovich had prepared a table that compared the Bath Profile searches to those specified in national/regional profiles. There were some questions about the document but it was decided that this type of information was useful. Carrol Lunau was asked to verify the information. After verification, it will be added to the Bath web site as an informational document.
8. Version 2 of Bath and the IRP Process
There was discussion about whether it was necessary to submit version 2 to ISO TC46 for IRP status. It was decided that the designation IRP gave credibility to the profile and was useful for marketing it, especially within Europe.
9. Version 2
The proposed timeline for Version 2 is:
-
Early June comment on revised draft
Late June distribute to Bath list and ZIG for comment
Autumn 2002 submit to ISO TC46
There was discussion of the Bib-2 attribute set and the recommendation from the Boston Spa meeting that the new version of the profile should outline a migration path from Bib-1 to Bib-2. Based on the discussion at the ZIG meeting of some of the Bath searches specified with the Bib-2 attributes, it was decided to remove all references to Bib-2 from version 2 of the Bath Profile. It was felt that the vendor and library community would be better served by focusing on development of Functional Area B for holdings, rather than on the new attribute set. The profile will, however, indicate that version 3 of Z39.50 allows support for multiple attribute sets in a query, and that such support for multiple attribute sets is necessary for national profiles that extend Bath to meet domestic requirements.
The table in 4.3.2 that summarized the search types will be removed from the profile. It was felt that the table was initially useful in clarifying the searches but was not required in version 2.
The U.S. National Profile does not include title, subject or any keyword scans. It was decided to remove these from the Bath Profile (5.A.1SCAN.3, 5.A.1.SCAN.5, and 5.A.1.SCAN.6).
A new search will be added to Bath at level 2. This is a search for the possessing institution of a work and is taken from the U.S. National Profile. The attribute combination for this search is:
| Use (1) |
1044 |
possessing institution |
| Relation (2) |
3 |
equal |
| Position (3) |
3 |
any |
| Structure (4) |
1 |
phrase |
| Truncation (5) |
100 |
do not truncate |
| Completeness (6) |
1 |
incomplete subfield |
As the documents which were formerly Appendix B: Diagnostics: Suggested User Display and Message Logging Using Bib-1 Diagnostic Set and Appendix E: Creating a Search from Scan Results will now be maintained on the web site of Z39.50 Maintenance Agency, they will be removed from version 2 of Bath. Text will be added giving the URLs at the Maintenance Agency.
The question of specifying the definition of full and brief records was discussed. It was decided that, while it was a nice idea, it could not be done; so Bath will remain silent on the issue.
No comments or suggestions for modification have been received for Functional Area C Cross-domain Search and Retrieval. It was decided to not make any changes to this Functional Area other than cleaning up any language ambiguities that may exist. The issue of using the already defined XML DTD or the RDF schema was raised. It was decided to ask the Bath list what the impact would be if we changed to the RDF schema. The ZIG list, Interoperability list, Dublin Core list, DC Libraries Group and CIMI should also be consulted. The question will need to be framed carefully to avoid opening up a broader discussion of Dublin Core.
Dana noted a typo in 5.A.1.4 where it should say "incomplete subfield" not "incomplete field" in the text.
Prior to sending the document to ISO, the Maintenance Agency will contract with a technical writer to clean-up the document. Two possible names were suggested for this activity.
10. Future Work for Bath
There was a discussion about the future activities of the group. The possibility of looking at search and retrieval from A and I databases was raised. The TZIG is currently developing a profile for this.
Carrol mentioned that she would like to see work begin on holdings searching. The feeling of the group was that this should wait until after holdings retrieval is implemented and vendors have had more experience working with the specs.
The question of the relationship between Z39.50, Bath, ILL, and NCIP was raised. It was agreed that it would be useful if somebody could look at these and define the interrelationships between them. At a minimum, it was suggested that an ESN should be defined to retrieve minimal holdings information such as the ItemID for systems that support both Z39.50 and NCIP.
11. Coordination of International Promotion
Carrol has had a suggestion from a group of libraries in the UK that more effort needs to go into coordination of promotion about Bath, especially for library managers and administrators. A significant amount of promotion has been done in Canada, the U.S. and the UK; but we do not know what has been done elsewhere. A LITA session at ALA in Atlanta on Interoperability will feature presentations related to Bath by Carrol Lunau, Bill Moen, Matt Goldner and Larry Dixson.
Janifer mentioned that the profile is viewed as an Anglo-Saxon initiative by many European nations and not relevant to their situation. She mentioned that, in particular, efforts need to be made in relation to Germany and France. Carrol committed to ensuring that version 2 of the profile will be available in French later this year. Janifer agreed to speak about Bath at AFNOR and to talk to colleagues about it in both Germany and France.
Fretwell-Downing offered the use of their publicity firm to help promote the profile. For example, we might want to issue an announcement about having reached consensus on holdings once we have agreement on Functional Area B.
11. Next Meeting
There was discussion about when and where a next meeting should be held. The last two meetings have been held in conjunction with the ZIG. This has provided valuable input from vendors and Z39.50 implementors, but we have not had much participation from librarians at these meetings. The next meeting should be held in conjunction with meetings or a conference that will attract librarians as well as vendors and implementers. It is important that the profiling initiative continue to receive input from the users of the products to ensure that it remains grounded in the day-to-day reality of libraries.
An EFILA meeting might be an appropriate venue; but the next meeting of this group is in October, which is too soon. If a workshop is arranged in Helsinki in the Spring of 2003, that would also be an appropriate venue since it is expected that such a session would attract a number of European librarians. By next Spring, version 2 will be completed and approved; and we may have vendors with some experience using the specifications in Functional Area B. We also may have the results of the TZIG discussions on profiling A and I search and retrieval.
12. Meeting Attendees
|
