Method and apparatus for constructing a compact similarity structure and for using the same in analyzing document relevance

• US 7,472,131 B2
• Filed: 12/12/2005
• Issued: 12/30/2008
• Est. Priority Date: 12/12/2005
• Status: Expired due to Fees

• Current Assignee: Justsystems Evans Research Incorporated
• Original Assignee: Justsystems Evans Research Incorporated
• Inventors: Evans, David A., Roma, Norbert, Shanahan, James G.
• Primary Examiner(s): Vy; Hung T
• Application Number:
  US11/298,500
• Publication Number:
  US 20070136336A1
• Time in Patent Office: 1,114 Days
• Field of Search: 707/5, 707/100, 707/101, 707/3, 707/10, 707/102, 704/10, 715/501.1, 715/229
• US Class Current: 1/1
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation

Abstract

A computer-readable medium comprises data structure for providing information about levels of similarity between pairs of N documents. The data structure comprises a plurality of entries of similarity values representing levels of similarity for a plurality of pairs of the documents. Each of the similarity values represents a level of similarity of one document of a given pair relative to the other document of the given pair. The similarity value of each entry is greater than a threshold similarity value that is greater than zero. The plurality of similarity-value entries are fewer than N²−N in number if the similarity values are asymmetric with regard to document pairing, and the plurality of similarity-value entries are fewer than

$\frac{N^2-N}{2}$
in number if the similarity values are symmetric with regard to document pairing. A method and apparatus for generating the data structure are described.

Claims

1. A method for constructing a data structure containing information about levels of similarity between pairs of documents of a set of documents, the method comprising:

obtaining similarity values for pairs of documents of the set of documents;

determining whether each of the similarity values is greater than or equal to a threshold similarity value; and

for each similarity value that is greater than the threshold similarity value, storing the similarity value in the data structure,

wherein the data structure comprises a plurality of entries of the similarity values, and wherein the plurality of entries of the similarity values are fewer than N²−

N in number if the similarity values are asymmetric with regard to document pairing and wherein the plurality of entries of the similarity values are fewer than

$\frac{N^2-N}{2}$

in number if the similarity values are symmetric with regard to document pairing, where N equals the number of documents.

2. The method of claim 1, wherein obtaining similarity values for pairs of documents comprises:

performing multiple queries on the set of documents, each query being based upon a given document of the set of documents; and

scoring documents responsive to the queries, said scoring providing the similarity values.

3. The method of claim 2, wherein the queries produce lists of documents ranked according to similarity values, and wherein obtaining similarity values for pairs of documents comprises extracting similarity values from the lists only for those document pairs whose similarity values are greater than or equal to the threshold similarity value.

4. The method of claim 1, further comprising:

storing an entry comprising the threshold similarity value in the data structure; and

storing an entry comprising a default similarity value in the data structure.

5. The method of claim 4, wherein the default similarity value is equal to the threshold similarity value.

6. The method of claim 4, wherein the default similarity value is less than the threshold similarity value.

7. A computer storage readable medium storing processing instructions adapted to cause a processing unit to execute a method comprising:

obtaining similarity values for pairs of documents of a set of documents;

determining whether each of the similarity values is greater than or equal to a threshold similarity value; and

for each similarity value that is greater than the threshold similarity value, storing the similarity value in a data structure,

wherein the data structure comprises a plurality of entries of the similarity values, and wherein the plurality of entries of the similarity values are fewer than N²−

N in number if the similarity values are asymmetric with regard to document pairing and wherein the plurality of entries of the similarity values are fewer than

$\frac{N^2-N}{2}$

in number if the similarity values are symmetric with regard to document pairing, where N equals the number of documents.

8. An apparatus for constructing a data structure containing information about levels of similarity between pairs of documents of a set of documents, comprising:

a memory; and

a processing unit coupled to the memory, wherein the processing unit is configured to execute the steps of;

obtaining similarity values for pairs of documents of the set of documents;

determining whether each of the similarity values is greater than or equal to a threshold similarity value; and

for each similarity value that is greater than the threshold similarity value, storing the similarity value in the data structure,

wherein the data structure comprises a plurality of entries of the similarity values, and wherein the plurality of entries of the similarity values are fewer than N²−

N in number if the similarity values are asymmetric with regard to document pairing and wherein the plurality of entries of the similarity values are fewer than

$\frac{N^2-N}{2}$

in number if the similarity values are symmetric with regard to document pairing, where N equals the number of documents.

9. The apparatus of claim 8, wherein the processing unit is configured to obtain similarity values for pairs of documents of the set of documents by:

performing multiple queries on the set of documents, each query being based upon a given document of the set of documents; and

scoring documents responsive to the queries, said scoring providing the similarity values.

10. The apparatus of claim 9, wherein the queries produce lists of documents ranked according to similarity values, and wherein the processing unit is configured to obtain similarity values for pairs of documents by extracting similarity values from the lists only for those document pairs whose similarity values are greater than or equal to the threshold similarity value.