The Bath Profile Four Years On: What's Being Done in the UK?

The genesis of the Bath Profile occurred at a meeting in Bath Spa during August 1999. It sought to address a wide range of issues pertaining to the effectiveness of the search and retrieval processes between Z39.50 client and server services. Over the ensuing months, members of the relevant communities created an ISO-recognised profile specifically intended to have international application. In June 2000, Release 1.1 of the Bath Profile gave precise semantic definition to the abstract search types used by Z39.50-enabled services. The initial focus was to define search and retrieval processes appropriate for bibliographic and cross-domain applications. The suite of search types for bibliographic applications was rationalised and extended in the recently issued Release 2.0. In addition, the new release now provides a detailed specification for the retrieval of holdings information, and defines a further suite of search types for authority records.

Four years after the creation of the Bath Profile, the M25 Systems Team sought to review its implementation within the UK. The resultant conference was funded by the Joint Information Systems Committee (JISC) and organised in association with UKOLN. It offered a series of integrated presentations on the envisaged potential and current actual deployment of the Profile. To summarise briefly , the presentations on 7 July 2003 comprised:

• Carrol Lunau, the principal editor of the Profile, gave the keynote paper on its historical development and current status
• John Gilby and Fraser Nicolaides, of the M25 Systems Team, described the possible problems that the absence of semantic interoperability could cause the implementers and end-users of virtual union catalogues
• Andy Powell of UKOLN presented the contextual role of Z39.50 and the Bath Profile within the JISC Information Environment
• Peter Stubley, Assistant Director of Academic Services at the University of Sheffield, reprised the conclusion of the UK National Union Catalogue Feasibility Study on the inappropriateness of Z39.50 as a technical solution for this type of service
• Rob Bull, Managing Director of Crossnet Systems, detailed the significant interrelated technical and business-strategy difficulties confronting Z39.50-server vendors who potentially might implement the Profile
• Slavko Manojlovich, of the Memorial University of Newfoundland and here representing Sirsi, concluded with an overview of that company’s marked success in effecting implementation of the Profile

The event was chaired by Peter Burnhill of EDINA.

It was a specific intention of the organisers that the presentations stimulate discussion amongst the audience. This was emphasised by an extended panel-moderated session at the close of the event. Many of the fifty-six delegates were invited representatives of leading vendor companies and service organisations operating in the UK. The former comprised: Aurora, Crossnet/DS, Dynix, Endeavor, Ex Libris, Fretwell-Downing, Sirsi and Talis. Amongst the services providers and developers were: EDINA, Institute for Learning and Research Technology, IRIS, M25 Consortium of Academic Libraries, MIMAS; and UKOLN Interoperability Focus.

The purpose of this article is not to restate the content of the papers and the discussion that they engendered, but rather to document certain key statements and proposals that could have a significant bearing on the current implementation and future development of the Bath Profile. These are detailed below. (It is worth noting, however, that the presentations [1] are available in HTML format or MS PowerPoint).

Stability of the Bath Profile

In a particularly cogent presentation by Rob Bull (who had been invited by the M25 Systems Team to emphasise the actual and perceived technical difficulties associated with implementation), it was conjectured that limited implementation was a consequence of the perceived instability of the Profile. The inclusion of proposed developments in Release 1.1, and anticipated extensions of the current version may have generated this unwillingness to commit possibly considerable technical and, by implication, financial resources to the creation of fully compliant Z39.50 server modules. Unfortunately, this perspective was neither refuted nor confirmed by any of the vendor delegates. (Indeed, as one delegate noted, the vendors were peculiarly reticent on most issues throughout the day.)

The legitimacy of such a position was comprehensively refuted by Carrol Lunau. Two particular points were to be noted, namely:

• that developments are responsive to the requirements of the application user communities, and thus constitute necessary revisions of a functionally useful profile
• that any further developments or extensions will not confute or otherwise negate the existing specifications.

In summary, the Profile should be regarded as both progressively developmental and immediately implementable.

Internationalisation

Although devised and presented as an internationally applicable profile, the Bath Profile has received only limited endorsement by countries outside North America and the United Kingdom. This regrettable situation has perhaps been exacerbated by the fact that until recently the Profile was available only in the English language (though I would regard this as being as much a symptom as a cause) [2].

It is arguable that this perforce threatens the status and identity of the Bath Profile as an international standard. Indeed, this situation may be independently compounded by the appearance of Z39.89-200X, ANSI/NISO’s U.S. National Z39.50 Profile for Library Applications [3]. (This forthcoming national profile uses many but not all of the bibliographic search and retrieval elements defined by Release 2.0 of Bath.) Given NISO’s organisational resources and inherent authority, it is conceivable that in the long term Z39.89 could, as Carrol Lunau stated, “ultimately make Bath redundant”.

I would argue that the definition of international interoperability should remain technically and organisationally independent of any national initiative. This does not negate any national or sub-national initiatives. The Z Texas Profile [4] demonstrates how Bath can, as originally intended, comprise the core specification around which are defined additional requirements.

It should be reiterated that this is a highly speculative assessment. Although beyond the formal scope of the conference, it was nevertheless evident that the identity of the Bath Profile required further consideration.

Independence of the Bath Profile from Z39.50

The Bath Profile is subtitled ‘An international Z39.50 specification for library applications and resource discovery’, which functionally associates it exclusively with the Z39.50 information-retrieval protocol. As Z39.50 is no longer the only protocol that requires search types to be semantically defined, it was cogently suggested by several delegates and at least one speaker that the Bath Profile be disaggregated from Z39.50. This would not negate the current content of the Profile, but would require it to assume a more abstract format.

Informed customer requirements specifications

A key inhibitor to adoption of the Bath Profile was identified as the inadequate specification of requirements by customer organisations. Request For Proposal documents may mandate compliance with the Bath Profile, but neither specify to what level nor require confirmatory proof. It was recommended therefore that requesting institutions be far more precise in the specification of their requirements. Within the context of particular consortial and other operational circumstances, they may, for example, require compliance to specific levels of each Functional Area of Bath, augmented by a range of additional search types. To confirm that the specification may be supported, potential vendors should be required to submit their Z39.50 server configuration to the Z-interop Testbed for independent verification.5.

A directory of Z39.50 server databases

One especially notable issue arising from the event was the identified need for a publicly accessible directory of Z39.50 server database implementations. This would document not simply the required connection information (IP address, authentication string, database name(s), etc.), but also would detail the Bib-1 attribute arrays that were supported. As a feature of this latter function, each directory entry would declare which Bath Profile-defined search types were supported. Such a service should greatly assist institutions and consortia in the configuration of their Z39.50 clients.

It is arguable that, at least in the UK, the directory should be associated with or form a component of the JISC Information Environment Service Registry (IESR) 6. The M25 Systems Team intend to discuss this proposition further with JISC and the IESR project team.

Cross-domain searching

Carrol Lunau reported that Functional Area C of the Bath Profile, ‘Cross-Domain Search and Retrieval’, had elicited no comment or criticism from the envisaged user community. Andy Powell and others, however, confirmed the considerable potential usefulness of this Functional Area as many portal and related ‘metasearch’ tools were being developed that enabled the simultaneous searching of multi-domain resources.

Conclusion

This event successfully highlighted the challenges facing the designers, commercial implementers and institutional end-users of the Bath Profile. In doing so, there may be a tendency to feel daunted and overwhelmed. Such a negative response, however, should be regarded as fundamentally inappropriate. Alternatively, in confirming the continued need to engage with the general issue of semantic interoperability, the identified challenges demonstrate the essential vibrancy of the responsively developmental Bath Profile.

The currency of the Profile should also be recognised. It is not a blueprint for future application, but needs to be used now by customer and vendor organisations alike. In this sense, the reticence of most of the vendor delegates to make any discernible contribution to the discussions could be regarded as somewhat alarming. As consortial interoperability comes increasingly to the fore, it may be necessary for the customer organisations to clarify their business requirements.

Finally, I should like to reiterate that the opinions and interpretations expressed here are entirely those of the author.

References

1. Bath Profile Meeting: Bath Profile Four Years On:What’s being done in the UK?,
   Monday 7 July 2003, Bath, UK. Presentations are available in HTML format or MS PowerPoint at:
   http://www.ukoln.ac.uk/events/bath-profile/programme.html
2. The French language version of Le Profil Bath is at: http://www.nlc-bnc.ca/bath/tp-bath2-f.htm
3. U.S. National Z39.50 Profile for library Applications by the National Information Standards Organization is at: http://www.niso.org/standards/standard_detail.cfm?std_id=734
4. The Z Texas Profile: A Z39.50 Specification for Library Systems Applications in Texas (Release 2.0) by The Texas Z39.50 Implementors Group and William E. Moen is at: http://www.tsl.state.tx.us/ld/projects/z3950/tzigprofilerelease20.html
5. The Z-interop Home page is at: http://www.unt.edu/zinterop/index.htm
6. The JISC Information Environment Service Registry Home page is at: http://www.mimas.ac.uk/iesr/

Author Details

Article Title: “The Bath Profile Four Years On: What’s being done in the UK?”
Author: Fraser Nicolaides
Publication Date: 30-July-2003
Publication: Ariadne Issue 36
Originating URL: http://www.ariadne.ac.uk/issue36/bath-profile-rpt/