IRE Index index Butterworth & Company Karen Sparck Jones All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including photocopying and recording, without the written permission of the copyright holder, application for which should be addressed to the Publishers. Such written permission must also be obtained before any part of this publication is stored in a retrieval system of any nature. lion (Ed. ~iicaI ision and `ort `be' [BM irnal Index Aberystwyth off-shelf test, 137, 142, 149, 150, `ices 151 977) Aboutness, concept of, 47, 54, 299 [OCRerr]rk AGRIS, 125 Alphabetical subject headings, 136, 147, 221, edn, 259, 264 Analysis of results, 11, 18 uter failure, 16 in Medlars, 20 in Cranfield 2 test, 20 semantic, 318 In: stages, 18 syntactic, 318, 320 30), Anomalies, 42 `rnal ANOVA procedures, 96, 97 Archetypal retrieval test, 11 and analysis, 18 components of, 12 costs and time, 15, 19 coverage and currency, 15 documents and requests, 13 experimental design, 13 explanatory variables, 15 failure analysis, 16 performance limits, 16 performance measurement, 14 recall and precision, 19 relevance assessments, 17 sampling, 13 system, 12 Artificial intelligence work, 172 Assumptions, 52 ASTIA uniterm test, 261 Automatic document classification, 323 Automatic extraction indexing, 21 Automatic indexing, 218, 219, 317, 321 non-statistical, 242 not using relevance information, 239 tests, 226 using relevance information, 241 Automatic keywords, 221 Automatic term classification, 323 Automatic systems, in laboratory testing, 156- 178 advantages, 165 benefits of, 159 data collection, 163 fallibility of apparatus, 175 model building, 170 objectives, 166 programs, 163 program correctness, 173 realism, 167 speed of, 159 technical faults, 160 test collections, 160 Automation, 2 Averaging, 86 BACKFILES, 187 Bayes' theorem, 35 Behavioural evidence of relevance, 196 Behavioural intervention, 122 Benefits of systems, 10 measures of, 107 Bias, 26, 246 in Cranfield I, 269 BOOKS test, 247 Boolean logic, 237, 286, 289, 303, 309, 312 in INSPEC, 290, 295, 311 British Library Research and Development Department, 285 Browsing facilities, 149, 169, 171, 191 Bureau of Ships study, 220, 222, 272 Case-Western Reserve University test, 139, 146, 219, 220, 272 Cataloguing policy, 125 Categories, 83 Changes in system, costs of, 125 343