Appeal 2007-0342
Application 10/323,932
INTRODUCTION
The claims are directed to methods for ranking documents according
to authoritativeness. Claim 1 is illustrative:
1. A method for creating, on a machine having an encoder, a
document textual authority model used to determine an authority of a
document having a plurality of document content features, the method
comprising:
determining, for each document in a set of documents, a set of
document classification attributes based on textured [sic] 1 contents of
documents;
applying a document attribute evaluation framework to each
document in the set of documents to determine, based on the set of
document classification attributes, a value of textual authoritativeness or a
class of textual authority for the document;
selecting a subset of document content features from the plurality of
document content features; and
encoding, using the encoder, the subset of document content features
into a feature vector x; and
determining a predictive model used to assign the feature vector x to
an authority rank or class based on the determined value of textured [sic]
authoritativeness or class of textual authority.
1 Although claim 1 recites “textured” contents of documents and “textured”
authoritativeness, we read the word as “textual,” consistent with both the
Examiner and Appellants’ reading. The obvious error was introduced by
Appellants’ amendment filed May 10, 2005.
2
Page: Previous 1 2 3 4 5 6 7 Next
Last modified: September 9, 2013