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. 344 Index Characteristics of systems, measurements of, 144 Chemical abstracts, 144, 244 Clustering bases, 227 Communication, between indexers and searchers, 123 failures in laboratory testing, 160 Computers, 1, 218, 324 evaluation of processing, 116 interaction witb user, 165, 171 output, 83 searcb costs, 307 statistical packages, 99 Computer searching, extraneous sources of variation, 75 Thomas, 21 Concepts, 44 --57 confounded, 47, 55 definitions of, 45 difficulties, 44 ineffable, 44 significance of, 53 interactions, 48 text-related, 46, 51, 54 user-related, 55 Conceptual simplication. 160 (`ontrol of tests, 214, 225 Coordinate indexing, 259 Costs, 37, 69, 106, 192, 228, 248, 259 in evaluation, 110, 120 INSPEC experiment, 306 justification of, 107 ofcbanges, 125 of experiments, 202 reduction of, 77 sensitivity of, 192 simulation, 192 (ost- benefit, 70, 107 (`ost-effectiveness, 15, INSPEC, 307,311 ([OCRerr]()st performance - benefits analysis, 107 [OCRerr]overage, 15,114 as extension of recall, 114 Cranfield projects, 49 55, 202. 219, 220.222. 256-283,321 co-ordination level function. 189 importance o[OCRerr] 256 index languages, 220. 224, 264 precision -rod rec[OCRerr]1, 263 [OCRerr]p[OCRerr]iti[OCRerr]it\ in [OCRerr]7X tr[OCRerr]t [OCRerr]oll[OCRerr]tion[OCRerr] 161 ( r,inti[OCRerr]ld 12 ic Cranfield WRU tests ([OCRerr]ranfic1d I 68 136.152, 215, 225,257 272 ,idmini[OCRerr]tr iii\r decisions 261 l[OCRerr]i r[OCRerr] in [OCRerr]69 ritir i[OCRerr]rn[OCRerr] 0 [OCRerr]67 [OCRerr]ic[OCRerr]ign ot, 25[OCRerr]. 268, 271) c[OCRerr][OCRerr]cnti.iI t[OCRerr]atarc[OCRerr].25[OCRerr] t,ii1arc[OCRerr], 2(14 Liain,in 1)ci-t[OCRerr])r[OCRerr][OCRerr]ri)cc 0(1. 151 k\ 1[OCRerr][OCRerr][OCRerr]thc[OCRerr]c- l-r[OCRerr][OCRerr]m 143 19,70,107,110,114 Cranfield I (cont.) importance of, 271 indexer/searcher memory, 269 indexing statistics, 261 indexing systems, 259 indexing time, 260, 268 index language, 138, 143, 144, 260, 269 influence of, 272 limitations, 271 parallel tests, 265 precision and recall, 137, 268, 271 failures, 152 relevance in, 271 searching, 262 search strategy problem, 146 source documents, 274 summary of, 266 uniterms, 257, 264 variables, 259 Cranfield 2,2, 19,21,60, 136, 138, 143, 157, 225, 272-282 aims, 273 analysis, 20 basis of, 273, 275 conclusions, 278 co-ordination strategy, 238 criticisms, 279 defects of, 281 design, 274 failures, 280 impact of, 2 indexing, 274 index language, 139, 143, 144,233,272,278, 282 influence of, 282 methodology, 282 recall and precision, 223, 271, 273, 276, 278, 281,282 relevance measurement, 20 results, 228, 277 searching, 236, 275 search strategies, 276 source documents, 274 source of queries, 74 specificity, 144 statistics, 279, 281 test collection, 22, 163, 243 variables, 230, 275 Cranfield Western Reserve University test, 221, 222, 256, 263, 265, 269 comparisons, 225 criticisms, 270 index language, 224 methodology, 229 precision in, 271 recall, 223 results. 228 variables, 230 Cross-reference provision (linkage), 144 Currency, 15 Current awareness. 10