IRS13 Search Matching Functions chapter E. M. Keen Harvard University Gerard Salton Use, reproduction, or publication, in whole or in part, is permitted for any purpose of the United States Government. 111-12 C[OCRerr]LLECTION INPUT AND DICTIONARY EVALUATION OVERLAP COSINE MEASURE CORRELATION CORRELATION Abstract, Suffix `5' Normed Recall .84o8 .8707 Normed Precision .5611 .6134 IRE:-3 Abstract, Stem Normed Recall .8725 .8777 34 Normed Precision .5829 .6167 Requests Abstract, Thesaurus Normed Recall .8974 .9067 -3 Normed Precision .6o4i .6574 Abstract, Stem Normed Recall .8237 .8397 Normed Precision .5830 .6377 CRAN-l Abstract, Thesaurus Normed Recall .8535 .8729 Requ2ests -3 Normed Precision .6251 .6936 Title, Stem Normed Recall .8082 .8120 Narmed Precision .5979 .621? Text, Suffix `s' Normed Recall .7546 .7768 Normed Precision .5097 .5462 Text, Stem Normed Recall .7434 .7695 Normed Precision .4978 .5248 ADI Text, Thesaurus-i Normed Recall .7386 .7819 Normed Precision .4350 .5092 35 Requests Test, Thesaurus-2 Normed Recall .6589 .6884 (Hastie) Normed Precision .3602 .4332 Abstract, Stem Normed Recall .7423 .7546 Normed Precision .4904 .5221 Abstract, Thesaurus Normed Recall .7830 .8043 -1 Normed Precision .5257 .5823 Performance results comparing cosine and overlap correlation coefficients, using logical vectors, for twelve options on three collections, using nor- malized recall and precision Figure 3.