Appeal No. 2007-0429
Application 09/848,430
The integers are then used to represent the document corpus as a “small sparse
matrix in vector form,” which is depicted as follows in Figure 7:
This figure shows three vectors. The “ALLDATA” vector represents the order of
appearance of the terms in the document corpus, i.e., “name,” “fred,” “name,”
“scott,” “scott,” “live,” and “live” (Specification 9:12-15). The
“STARTMARKER” vector represents the starting point of each document in the
ALLDATA sequence (id. at 9:15-16). The STARTMARKER array in this
example indicates that the first terms of Documents 1, 2, and 3 correspond to the
first, third and fifth integers in the ALDATA array. The “MULT” vector consists
of normalization factors to be applied to the terms in Documents 1-3, respectively,
when floating point representation is required (id. at 9:16-18).
THE CLAIMS
The independent claims on appeal are claims 1, 5, 9, 13, and 15, of which
claim 1 reads:3
1. A method of converting a document corpus containing an
ordered plurality of documents into a compact representation in
memory of occurrence data, said method comprising:
3 A corrected Claims Appendix was filed on June 14, 2006.
3
Page: Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 Next
Last modified: September 9, 2013