ISR10 The Query-Document Matching Function chapter Joseph John Rocchio Harvard University Gerard Salton Use, reproduction, or publication, in whole or in part, is permitted for any purpose of the United States Government. Comparison Operation Definition Equality A=3 a [OCRerr] A[OCRerr]a 6 B Inclusion ACB a [OCRerr] A[OCRerr]a E 3 Overlap Correlation [OCRerr](A,B) = n(A[OCRerr]3) Metric Distance g(A,B) = 1[OCRerr](A:[OCRerr] B)/n(AU[OCRerr]3)[OCRerr] Comparison Operations on Set Represented Operands Table 4.1 Boolean al[OCRerr]ebra to the partially ordered system formed by the subsets of the keyword set and the set inclusion relation. This allows one to structure the representation of a search request in the form of a Boolean combination of [OCRerr]eywords, i.e. q =w(k[OCRerr]) as opposed to using an unordered keyword set representation. Let column i of the keyword document matrix (Fi[OCRerr]ure 4.1(a)) represent the document subset of the ith keyword. The retrieval,operation, then, consists in generating the retrieved subset R by replacing each keyword in the Boolean query polynomial by its keyword set and substituting set intersection for Boolean `and" and set union for Boolean "or". With this transformation the subset R is specified by: R =L[OCRerr]1(K.) 1