Back to results
US 2005/0091035 A1Published application
System and method for linguistic collation
AssigneeMicrosoft Corporation
InventorsMichael Kaplan, Catherine Wissink, Julie Bennett
Life of this patent
Estimating the term…
Loading patent…
Loading patent…
AssigneeMicrosoft Corporation
InventorsMichael Kaplan, Catherine Wissink, Julie Bennett
Estimating the term…
A system and method is provided for handling the collation of linguistic symbols of different languages that may have various types of compressions (e.g., from 2-to-1 to 8-to-1). A symbol table of the symbols identified as Unicode code points is generated, with each symbol tagged with a highest compression type of that symbol by sorting the compression tables of the various languages. During a sorting operation with respect to a given string, the tag of a symbol in the string is checked to identify the highest compression type of compressions beginning with that symbol, and the compression tables for the language with compression types equal or lower than the highest compression type of the symbol are searched using a binary search method to find a matching compression for the symbols in the string. A common search module is used to perform binary searches through compression tables of different compression types.
A computer-readable medium having computer-executable instructions for performing steps for building a symbol table for storing sort weights for a plurality of linguistic symbols used in a plurality of languages supported by a computer system, comprising:
A method of building a symbol table for storing sort weights for a plurality of linguistic symbols used in a plurality of languages supported by a computer system, comprising:
A computer-readable medium having computer-executable instructions for performing steps for a computer search program to carry out a linguistic sorting operation, comprising:
A method of performing a linguistic sorting operation, comprising:
Dependencies are inferred from the claim text.