Bibliographic Enrichment Advisory Team

Annual Report, 1995


Introduction and Overview

BEAT produced some very tangible results in 1995 despite the impact of an interruption in access to its funding for a protracted period during the first half of the year. Though initial BEAT 1995 planning was completed in the March - April time frame, the Team was unable to consider actual funding for projects until late July.

At the request of the Librarian, the Business Research Project (from which BEAT in turn receives its funding) was audited, and though the audit gave the project "flying colors," a consequence of difficulty in accessing funding was that BEAT was forced to delay implementation -- indeed approval -- for projects it wished to undertake until much later in the year than was originally anticipated. The interruption meant that almost all BEAT projects reappraised in July and that still required funding could be undertaken only in the last quarter of the year. In addition, then pending discussions with a private sector firm for the acquisition of tables of contents data precluded any final commitment of funds to be reallocated from that project to any other pending conclusion of those negotiations.

The BEAT activities for this year were varied, particularly since several projects the team had in mind that would have dovetailed with those of earlier years could not be undertaken. Nonetheless, a theme that runs through all the BEAT initiatives is the implicit idea that BEAT will continue to look for and to develop tools to automate many of the processes in projects the team does undertake, and that BEAT aims to make data conversion and bibliographic enrichment and access reliable and economical. Where possible the team wants to put more data into records or at least make the enriched data available to users, and the idea that the searcher will be better able to evaluate the data presented in their search results. The projects described below aimed to provide enriched data, improved access to data, or more information about the data presented.

Library of Congress Classification

Classification -- what readers most often see as call numbers on their books -- is a very useful way to organize information by category. Taken in combination with terminology found in particular areas of knowledge, classification can provide a sophisticated way to isolate material of interest. By continuing to support revision of the structure and terminology found in various classes and subclasses as they relate to BEAT's primary areas of focus, and then to sponsor the conversion of the revised schedules to machine readable form, BEAT continues to improve the accuracy and currency of data that staff use to build records and that then, in turn, provide data to users who need it in increasingly meaningful form.

Class HJ

A BEAT project to complete the revision of class HJ (Public Finance) was undertaken under the direction of Gabriel Horchler, Team Leader of the Business and Economics Cataloging Team. This revision, and its subsequent conversion to electronic form, together with that already completed for Classes H - HG, brings the entire business /economics half of the LC classification scheme into the on-line classification arena. In addition, the substance of the HJ revision, as well as the procedures followed will, after approval of the Library's Cataloging Policy Office (CPSO), serve as a model for additional work in other schedules to be undertaken in 1996. The completed product is more compact (it is 30 percent smaller than its predecessor), is easier to use, and includes the latest terminology.

The work was also important since it allowed the project to be undertaken and completed just as the entire H schedule itself was to be issued in a new edition. Thus, it allowed CPSO staff who did the conversion to on-line form to understand the types of challenges presented by revision of an entire schedule and the incorporation of numerous changes in the HJ sub-class into the existing machine readable database.

In addition, there was some minor final work done on a 1994 project regarding the linking of H-HJ classes to terms in the Library of Congress Subject Headings (LCSH) in the first part of 1995.

Class JZ

Before the hiatus in activity described in the introduction, the BEAT Team approved a project and budget for the conversion of Class JZ (among other items JZ includes those sections of the former JX schedule dealing with international organizations and international aspects of business) into electronic form for inclusion in the on-line classification database. The actual redesign of JX and the creation of JZ was done by Dr. Jolande Goldberg, Senior Law Classification specialist in CPSO, and BEAT Team member.

The conversion project was deferred, but will be included in 1996 activities. The conversion of JZ will also mark the completion of the J schedule, whose internal revision (Class J- JV) and subsequent conversion to electronic form BEAT supported previously.

Serial Record improvement

According to business reference staff one of the major problems in providing reference service for business periodicals is whether the periodical is held by the Library of Congress or not. BEAT approved work that aimed to ameliorate this situation.

Location of titles

The project undertaken in this area aimed to eliminate or better clarify status of material indicated by the legend appearing in serial records retrieved in the LC on-line systems. Under the direction of Regina Reynolds, and undertaken by both technicians and catalogers in the Serial Record Division, the original goal was to identify and process some 2500 business serials, some 2885 records were eventually updated to accurately indicate whether the titles were in the Library. These records were all updated to indicate whether LC holds them or not, add a "current issues only" legend where appropriate and add a data element to indicate the location of those issues. The team actually identified over 3200 records, and the last group of these was awaiting additional funding for processing as this report was written. Additional classification data was also provided for certain records determined to be in the Library's PREMARC collection.

The project had at least three benefits. First, it brings to the attention of LC recommending officers titles which, though displayed by the system, are not actually in the Library's own collections. By so doing, selection officers can move to acquire this material. Second, by reviewing the cataloging, materials which merit full cataloging and analysis can be identified and processed, thus providing more useful information to those searching the material in the Library's collections. Third, in undertaking this work catalogers are now considering a strategy to eliminate the appearance of the ambiguous "May or may not be in LC" message, thus helping to reduce the incidence of this type of data in the future.

Thus, the project eliminates ambiguity encountered at LC by users who find a record for a serial publication and produced a search result that makes a definitive statement about the availability and location of these materials as well as leads to a potential solution for preventing this ambiguity in the future.

Text Capture and Electronic Conversion (TCEC) Initiatives

TCEC activities grew out of the first Table of Contents initiatives undertaken by BEAT in previous years as well as from a convergence with parts of the Electronic CIP experiment. The three principal TCEC developers all sit on the BEAT Team, and 1995 saw the continued and ongoing development of TCEC programs.

The objective of these activities is to develop programs that can improve cataloging operations and which maximize the use of the Library's Bibliographic Workstation (BWS) a high-end technical services workstation that uses the OS/2 operating system, TCP/IP, and customized terminal emulation software among other programs to provide the Library's catalogers with a hardware and platform that their work requires.

ClipSearch

Among the chief accomplishments of the group in 1995 was the development of the ClipSearch program. ClipSearch is an OS/2 program written in Visual Rexx (VX-REXX) that helps staff search authority and bibliographic files, automates the copying of retrieved data into cataloging records-in-progress, creates name authority records and generates cross-references automatically among other things. The program contains on-line MARC codes for languages, countries, and geographic areas. It permits staff to capture, create, and store virtually unlimited personalized lists of frequently used headings, phrases, class numbers and other information for quick reference and recall permits immediate incorporation of this data into cataloging records.

As the year closed, user testing of the ClipSearch program was nearly complete. At that time it had 60 official testers and numerous other LC users who loaded the program on their BWS without waiting for the final release version. The official release of version 1.0 of the program in fact took place on January 25, 1996.

TCEC SERLOC

Also of note was the TCEC SERLOC computer program, which enables the Serial Record Division to create Library of Congress SERLOC (Serial Location) records from the bibliographic records the division creates on OCLC as part of the CONSER program. The TCEC initiative allows the Library to create needed SERLOC records from the exported MARC data files. This is a tremendous time-saver in the division and is more efficient than the method in place previously.

NLM CIP

A noteworthy productivity gain in the processing of CIPS from the National Library of Medicine was also achieved through the TCEC NLM CIP program which eliminated the necessity of re-keying NLM CIP data into the Library of Congress MUMS database. The program allows the NLM data to be reformatted electronically and then merged with the data contributed by the Library of Congress. An initial comparison of the method LC applied formerly and the new NLM CIP method showed a four-fold production improvement as well as improved turn-around time. In addition, since there is no re-keying involved, the record is quite accurate.

The TCEC group also received generous additional financial support from the Cataloging Directorate at the end of the Library's fiscal year. These funds will be used for TCEC initiatives and development in 1996.

Outreach

As TCEC programs were implemented more extensively, the group, particularly David Williamson, also demonstrated these productivity programs both inside and outside of LC. Notable among these demonstrations outside of the Library were multiple presentations at ALA, such as that for the cataloging research interest group at ALA, a presentation to Harvard and MIT cataloging staffs, and to the University Press of both institutions. Representatives or visitors from the Bibliotheque Nationale and from Sweden were shown the Electronic CIP program. Demonstrations within LC included programs for the cataloging and CPSO staff of the Library, NACO Series Training participants, and the LC department Directors, among others.

Table of Contents Initiatives

BEAT has long wished to enrich the content of bibliographic records with table of contents data, but pursuit of this agenda continued to frustrate the Team in 1995. Despite vigorous efforts to reach agreements first, with a private sector vendor, and later with an academic-based consortium, the Library was unable in the first case, and it appears unlikely in the second case -- to reach an acceptable agreement for the acquisition and redistribution of table of contents data from these sources. BEAT had wished to include TOC data with the catalog records distributed by the Library.

BEAT (with discussions with officials of the Library's Business Research Project) set aside funds sufficient to pursue this initiative assuming a successful conclusion to negotiations with these parties in turn, and a consequence to the activities of the Team generally was to make these amounts unavailable for other activities while negotiations continue. Discussions with the consortium continue, and a deadline of June 1996 has been set for either undertaking or abandoning this venture. BEAT will of course remain open to pursuit of the idea, but at the present time it appears that only LC-produced Tables of Contents data are likely to appear in records distributed by the Library anytime in the near future.

To that end, catalogers working with CIP data submitted by publishers participating in the Electronic CIP experiment continue to select Table of Contents data for inclusion in MARC records, though the actual number of such records is still relatively small. Staff follow BEAT-devised as well as other LC policy guidelines for incorporating this TOC data in MARC records.

In the area of R&D related to the TOC idea BEAT scanned and then converted to text some samples of tables of contents for business books. The text was then marked up in HTML (Hypertext Markup Language) and placed on a World Wide Web server at LC. A link to this data was placed in the 856 field of the corresponding MARC record with the result that a search on the test Z39.50 server at LC will reveal the existence of the linked data when a user encounters the bibliographic record, and the related table of contents can be displayed merely by invoking a keystroke or mouse click on the hypertext link. In addition, those users searching only the LC catalog can see the 856 field link which gives the full internet address of the related data which they could subsequently use to access the data on the World Wide Web.


Go to the Library of Congress Home Page

[LCLibrary of Congress

Comments: [email protected] (02/03/96)