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. Searches (cont.) performance criteria and nieasures, 149 recording problems, 148 response to query, 64 rules for termination, 148 time, 150, 309 Searching,9, 116 controlling, 145 Cranfield, 1, 262 Cranfield 2, 275, 276 evaluation, 116, 236 formulation, 9, 132 logic, 236 strategy, 236 tests, 139, 232, 23[OCRerr]238 Search process, 63 Search statements, 63, 317 specificity and exhaustively, 64 type,64 Search strategy variationS, in SDI profiles see also INSPEC experiment, 285-315 Selective dissemination of information, 10, 15 Semantic analysis, 318 Sequence effects, fatigue factor, 81 magnitude of, 79 Set retrieval, 39 Significance levels, 94 Significance tests, 25, 29 Sign test, 95 Simulation, 119, 179-198 advantages of, 120 definitions, 179, 182, 183, 184 examples of, 185 goals of, 195 methodology, 196 previous work applied to retrieval, 192 random number, 191 reasons for, 184 virtual attribute technique, 196 SMART, 3, 72, 140, 161, 165, 175, 219, 233, 282, 3j[OCRerr]329 assumptions, 319, 327 automatic text analysis, 227 collections, 243, 244 design of, 318 document environment, 319 early results, 319 effectiveness, 327 extended, 321 feedback, 241 form, 240 methodology, 322, 323 processing options, 119 recall and' precision, 318, 320, 323 relevance, 231 assessments, 319, 322 relevance feedback, 242 results, 229, 321 theoretical insights, 323 users and, 322 Index 351 Software, 175 development of, 77 for INSPEC, 299 online, 77 Specificity, 139, 144 Cranfield, 2, 278 SPIRES/BALLOTS,77 STAIRS, 77, 163 Statistical inference, 91 Statistical Package for the Social Sciences, 185 Statistical packages, 99 Statistics, 23-30, 36, 225 approach to tests, 225, 227 averaging, 86 comparisons, 94 effect of bias, 26 estimation, 92 experimental design, 29 in Cranfield 2, 279, 281 laboratory tests, 28 limitation of, 30 parametric, 25 non-parametric, 25 reasons for, 23 samples, 25 sample size, 25, 29 significance levels, 94 techniques, 240 time and variables, 26 two populations, 24 value of, 239 variables, 27 STATUS, 163 Structured programs, 77 Subsearches, 146 Surrogates, 71 Syntactic analysis, 318, 320 Syntactic control, degree of, 63 Synthema, 48, 52 Syntol, 219, 221, 224 System restraints in testing, 130 Taube's uniterms, 219 Technical information intervention, 121 Terms, arrangement into groups (concepts), 295 automatic classification, 323 choice of, 148 in INSPEC, 295 Term associations, 232 Term clustering, 239 evaluation, 232 Term discrimination value, 62 Term groups, in INSPEC, 289, 313 weighting, 237 Term phrases, 325 Term precision, 325 ferm weighting, 157, 161, 226, 231, 237, 239, 243, 265 in INSPEC, 288, 290, 296, 313