31
DESIGNING AN ONLINE PUBLIC ACCESS CATALOGUE Nathalie Nadia Mitev, Gillian M Venner and Stephen Walker

Library and Information Research Report 39

|Bjy
The British Library

WQ

British Library Cataloguing in Publication Data

Mitev, Nathalie Nadia Designing an online public access catalogue: Okapi, a catalogue on a local area network. (Library and information research reports, ISSN 0263-1709:39) 1. Catalogs, On-line I. Title II. Venner, Gillian M. III. Walker, Stephen IV. Series 025.313 Z699 ISBN 0-7123-3058-5

v

DESIGNING AN ONLINE PUBLIC ACCESS CATALOGUE: Okapi, a catalogue on a local area network
NATHALIE NADIA MITEV, GILLIAN M VENNER STEPHEN WALKER

Library and Information Research Report 39

Abstract
Designing an online public access catalogue [OPAC]: Okapi, a catalogue on a local area network [LAN] is the final report of a two-year research project ''Microprocessor networking in libraries" which was funded by the British Library and the Department of Trade and Industry, and based at the Polytechnic of Central London. The aim was to produce an OPAC on a LAN, that would be readily usable without training or experience, without sacrificing effectiveness or being tedious for experienced users. The result was a functioning prototype OPAC called Okapi, which has a number of distinctive features: use is eased by coloured keys and a lack of jargon; the system uses search decision trees to select a suitable action at each stage of a search, and it performs automatic Boolean and hyperBoolean functions where appropriate. The OPAC was installed and evaluated in one of the Polytechnic site libraries. There is a full description of Okapi under the headings: source file: indexing; search functions; user interaction and evaluation. There are also general discussions of these topics, and chapters on OPACs in general, LANs, and recommendations for future research. Library and Information Research Reports are published by the British Library and distributed by the British Library Publications Sales Unit, Boston Spa, Wetherby, West Yorkshire LS23 7BQ. © The British Library Board 1985

ISBN 0 7123 3058 5 I S S N 0263 1709 T h e opinions expressed in this report are those of the authors and not necessarily those of the British Library. Typeset in Great Britain by SB Datagraphics Ltd, Colchester and printed by William Clowes (Beccles) Ltd, Beccles.

V

Preface
This report is primarily about the design and use of online public access catalogues (OPACs). It describes the application of a study of existing online catalogues, and of research in information retrieval and other fields, to the design, construction and testing of a prototype experimental OP AC (called Okapi). The prototype catalogue was designed to operate within a local area network (LAN), although there is little in this report, or in Okapi, which is specific to the LAN environment. The intended readership includes library management and technical services personnel, students of library and information science, and those concerned with interactive computer systems, particularly with systems intended for untrained and non-dedicated users. Most of the report assumes little prior knowledge of library automation or of computing in general, but there are a few sections which require more specialised knowledge. We wish to thank all those who have helped us during the course of the project. In particular we wish to thank Neil McLean and Mel Collier (Head and Deputy Head of Library Services, Polytechnic of Central London) for their constant support, and Paul Baxter and Maureen Grieves of the British Library Research and Development Department, who have given so much time to the project. Other Polytechnic staff without whom the work would scarcely have been possible are Winifred Abbott (Technical Services Librarian) and the cataloguing staff — particularly Penny Pope and Helen Chapman — for endless help on cataloguing matters and the use of MARC; Maura Coghlan and her staff in the Riding House Street Library for providing facilities for us to install and test the catalogue with real users; and Dave Roberts (Head of Computer Services) for access to computing facilities. Our colleagues in the Library Technology Centre, Pat Manson and Mary Rowbottom, gave advice and encouragement on many occasions.

VI

Preface

Outside the Polytechnic, we are particularly grateful to the following people: Philip Bryant and Alan Seal of the Centre for Catalogue Research for many stimulating discussions on the use of catalogues in general and on OPACs in particular; Linda Reynolds for help with screen layouts and with the design of evaluation procedures; Charles Hildreth of OCLC for encouragement and for having written the best book on OPACs [1]; the staff of Nestar Systems (suppliers of the hardware) — particularly Railton Frith (Technical Support Manager); Alan Hopkinson for the benefit of his encyclopaedic knowledge of cataloguing and of bibliographic exchange formats; and Margaret Slater of Aslib for her valuable advice. Finally, there were the several hundred people — Polytechnic students and staff, professionals in various fields, conference delegates and passers by, friends and relations of the project team — who acted as guinea pigs by being persuaded to sit down in front of computer terminals, were told "Just use it!", and were observed and questioned. These individually unacknowledged collaborators made countless constructive suggestions, many of which have been incorporated in the prototype OP AC described in this report. May 1985 NNM GMV SW

Reference 1 Hildreth C R. Online public access catalogs: the user interface. OCLC, 1982.

Vll

Contents
List of figures List of tables 1 1.1 1.2 1.3 1.4 2 2.1 2.2 2.3 2.4 Background, history and aims The project proposal Staffing History of the project Aims of the project Online public access catalogues Sources of information on OPACs 2.1.1 Existing OPACs 2.1.2 Written sources OPACs as IR systems 2.2.1 Coverage and source 2.2.2 Users and usability From conventional catalogues to OPACs Types of OP AC currently in use 2.4.1 First generation OPACs 2.4.2 Second generation OPACs 2.4.3 The third generation 2.4.4 The National Library of Medicine ( N L M ) CITE OP AC Types of catalogue search 2.5.1 Specific item searching 2.5.2 Subject searching User interaction with OPACs 2.6.1 User attitudes and behaviour 2.6.2 OP AC input from users 2.6.3 OPAC output OPAC design 2.7.1 OPAC components 2.7.2 The design of Okapi xii xiii 1 1 2 3 6 7 7 7 8 9 9 10 10 11 12 13 14 15 16 17 17 18 18 19 20 20 20 21

2.5 2.6

2.7

Vlll

Contents 25 26 28 28 28 30 30 30 32 34 34 35 36 39 39 41 42 42 43 43 44 45 46 47 48 48 49 49 50 52 52 53 53 54 54 56 56 57

3 3.1 3.2

Local area networks Types of LAN LAN or multi-user minicomputer? 3.2.1 Relative cost 3.2.2 Performance considerations: LAN advantages 3.2.3 Performance considerations: shared processor advantages 3.3 The project LANs 3.3.1 General description and cost 3.3.2 Performance 3.3.3 Reliability 3.3.4 Some technical information on the PLAN 4000 LAN 3.4 Data security 3.5 The design of applications software for LANs

Creation of Okapi source file Machine-readable catalogue and choice of project test site Selection of fields from the MARC record Discussion of MARC fields omitted and other problems 4.3.1 Statements of responsibility 4.3.2 MARC fields omitted 4.3.3 Records with analytical entries 4.3.4 Other problems of creating a MARC subset 4.4 The Okapi record 4.5 Creation of Okapi file from MARC file 4.5.1 Creation of Okapi record from MARC record 4.5.2 Transfer of data to PLAN network 4.5.3 Final Okapi file creation on PLAN 4.6 File size, mean record length and other statistics 4.7 Order of records in the file and on screen 4.8 File updating and maintenance 5 5.1 5.2 Access to catalogue files: indexing Introduction and guide to the chapter OPAC index contents 5.2.1 Access points — the sources of index keys 5.2.2 Data types 5.2.3 Okapi index contents Filing order and the display of indexes 5.3.1 Browsing and index order 5.3.2 The readability of index displays

4 4.1 4.2 4.3

5.3

Contents 5.3.3 Character coding and filing order in indexes 5.3.4 Filing order and display of the Okapi index The construction of indexes for bibliographic records 5.4.1 Personal names 5.4.2 Corporate and conference names
5.4.3 Titles

IX

5.4

58 60 61 62 64
65

5.5 5.6

5.7

5.4.4 Subject headings Stop lists and authority files 5.5.1 Stop lists 5.5.2 Authority files Index storage and access 5.6.1 Methods of storage 5.6.2 Postings lists 5.6.3 Searching an index 5.6.4 The representation of data in indexes 5.6.5 One index or several indexes? The Okapi index: storage, access and construction 5.7.1 Summary 5.7.2 Choice of index structure for Okapi 5.7.3 The secondary index 5.7.4 The primary index 5.7.5 Postings and beasts 5.7.6 Construction of the Okapi index Search functions and search trees String comparison Possible results of an index search Conditional searching Booleans Hyper-Booleans Computational aspects String searching Search trees Functions not implemented and reasons why User interaction design Introduction Users' reactions to previous online catalogues Previous OP AC designs and their user interfaces 7.3.1 IR aspects

68 69 69 72 72 72 73 74 75 75 76 76 77 78 79 81 81 87 87 88 89 89 90 93 94 94 96 98 98 100 102 104

6 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 7 7.1 7.2 7.3

X

Contents 108 112 112 115 117 121 121 124 126 143 143 143 144 145 146 147 149 150 150 154 154 155 155 156 156 159 159 160 161 162 162 165 167 168 170 174 176

7.4

7.3.2 Dialogue aspects General approach to Okapi user interaction 7.4.1 How many interaction levels? 7.4.2 How to offer search types 7.4.3 Automatic features: behind the screen... 7.5 Structure and description of Okapi dialogue 7.5.1 User input * 7.5.2 Action by Okapi 7.5.3 Okapi displays 8 8.1 8.2 8.3 Okapi in use Introduction 8.1.1 Brief literature review 8.1.2 What can be evaluated and why? Methodology 8.2.1 Automatic logging 8.2.2 Observation and structured interviews Structured interview results 8.3.1 Users' background and experience 8.3.2 Users' opinions of Okapi 8.3.3 Questions about their search Analysis of user sessions 8.4.1 Session length, types of searches, search states 8.4.2 Search failures 8.4.3 Use of keys, spelling and input mistakes Logging results

8.4

8.5 9 9.1 9.2 9.3 9.4

Conclusions and recommendations Summary of results of the project OPACs — advantages and disadvantages LANs and library automation Some conclusions and recommendations for research and development 9.4.1 On the enhancement of subject searching in OPACs 9.4.2 On data input by the user: spelling and keyboarding 9.4.3 On OP AC output 9.4.4 On user interaction in general 9.4.5 On bibliographic files for OPACs 9.4.6 On the evaluation of OPACs and the study of their use 9.5 Concluding remarks

Contents Appendices 1 2 3 4 5 6 MARC field selection criteria Subnelds used from selected MARC fields Okapi record format Special characters used in the Okapi record General comments received about Okapi Results of 70 interviews 1 Students' background and experience 2 Questions about Okapi 3 Questions about their search 4 Some cross-tabulation Analysis of 96 user sessions Okapi logging results Annotated and edited extract from the log of an Okapi station

XI

7 8 9

180 182 186 188 190 199 199 201 202 203 204 208 209 211 221 232 251

References Glossary Index Other reports

Xll

List of figures
3.1 4.1 6.1 LAN diagram Complete MARC to Okapi example Title only search tree 33 51 95 106 109 111 131 131 132 132 133 133 134 134 135 135 136 136 137 137 138 138 139 139 140 140 141 141 142 142

7.1 Subject search in LCS (Ohio State University) 7.2 Main menu in Bibliotech (Advanced Management System) 7.3 User asks for help in Datalib 7.4 Okapi introductory screen 7.5 Okapi first menu 7.6 Okapi first menu with help 7.7 Okapi keyboard help 7.8 Okapi overall view of specific item input 7.9 Okapi specific book help (1) 7.10 Okapi specific book help (2) 7.11 Okapi title input 7.12 Okapi author input 7.13 Okapi subject search input 7.14 Okapi subject search help (1) 7.15 Okapi subject search help (2) 7.16 Okapi search: author only (match) 7.17 Okapi search: author only (no match) 7.18 Okapi search: author only (AND) 7.19 Okapi search: title only (match) 7.20 Okapi search: title only (partial match) 7.21 Okapi search: title and author 7.22 Okapi search: subject (AND) 7.23 Okapi search: subject (hyper-OR) 7.24 Okapi author index display 7.25 Okapi title index display 7.26 Okapi brief record display 7.27 Okapi full record display

xiii

List of tables
5.1 5.2 5.3 5.4 7.1 7.2 7.3 The most frequent terms in a file of 30,081 MARC records Analysis of index terms by type of key Index size and term occurrence statistics The most frequent initial tetragrams Summary of the first two generations of OPACs Two levels of interaction Automatic displays 83 84 85 86 103 113 120