ISR11 Relevance Feedback in an Information Retrieval System chapter W. Riddle T. Horwitz R. Dietz Harvard University Gerard Salton Use, reproduction, or publication, in whole or in part, is permitted for any purpose of the United States Government. Results for [OCRerr]ery No.9 QA9ANALYSIS Doc[OCRerr][OCRerr]ment No. Name Relevant Correlation 50 4l8ENGLISH-L x .12680691 32 902DoCUME[OCRerr][OCRerr] .09522233 18 11225TATE-OF .10680565 5[OCRerr] 90IA SELECTE .11023756 35 100[OCRerr]DENSITY .11934192 31 207F[OCRerr][OCRerr]ACTICAL .13022519 14 518[OCRerr]DUCAT10N .13494266 63 214ELECTRONI .13599739 29 21[OCRerr] STATIST .13606583 15 61OCHARACTER .13951793 28 [OCRerr] ROLE .14294774 44 903THE USE 0 .14716034 80 1OOlDESIGN 0 .15302047 25 20[OCRerr]HE REDUC .15309977 38 303C. M. HER .15538235 N Query(4) = [OCRerr]uery(3) + Alpha z 1=1 [OCRerr][OCRerr]EIG[OCRerr](I) * (1) Document of Rank I Document Razik 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Weight 1-1-1-1 0 0 0 0 0 0 0 0 0 0 0 Alpha = 2 (d) After Third Update (Cosine Correlation) Rigure 5 (continued)