System and method for improving the ranking of information retrieval results for short queries

• US 5,870,740 A
• Filed: 09/30/1996
• Issued: 02/09/1999
• Est. Priority Date: 09/30/1996
• Status: Expired due to Term

• Current Assignee: Apple Computer Incorporated (Apple Inc.)
• Original Assignee: Apple Computer Incorporated (Apple Inc.)
• Inventors: Rose, Daniel E., Cutting, Douglass R.
• Primary Examiner(s): Black, Thomas G.
• Assistant Examiner(s): Homere, Jean R.
• Application Number:
  US08/719,816
• Time in Patent Office: 862 Days
• Field of Search: 707/3, 707/4, 707/5, 707/7, 707/500
• US Class Current: 1/1
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99937 Sorting

Abstract

A method and system for retrieving information in response to a query by a user. The method includes the steps of receiving a signal s having a value corresponding to a relevance-ranking algorithm score of a retrieved document, receiving a signal q having a value corresponding to the number of words in the query and a signal v having a value corresponding to the coordination level of the retrieved document and query (i.e., the degree of overlap between the document terms and the query terms), and generating an adjusted score s1 dependent on the signal s, the signal q and the signal v. The adjusted score s1 takes the coordination level into account for small values of q and gradually decreases the importance of the coordination level as q increases. The system of this invention includes a computer-based system for carrying out the method of this invention.

Claims

1. A method for a computer system, having a CPU and RAM, to retrieve information in response to a query, comprising the steps of:

issuing a query on a database;

identifying a retrieved document based on the query;

receiving into said RAM said retrieved document and an accompanying variable s having a value corresponding to a relevance-ranking score of said retrieved document;

receiving into said RAM a variable q having a value corresponding to the number of words in the query and a variable v having a value corresponding to the overlap between the words in said retrieved document and in the query; and

using said CPU and said variables s, q and v to generate an adjusted score s1 corresponding to the value of said variable s increased by an amount proportional to the value of said variable v, said amount decreasing as the value of said variable q increases.

2. The method of claim 1, further comprising the steps of:

receiving into said RAM additional retrieved documents and accompanying variables s;

generating adjusted scores s1 for said additional retrieved documents; and

ordering said retrieved documents according to said adjusted scores s1.

3. The method of claim 1, further comprising the step of:

receiving into said RAM a variable ∂

chosen to control the increase to said variable s by said amount dependent on said variable v.

4. The method of claim 1, further comprising the steps of:

receiving into said RAM additional retrieved documents and accompanying variables s; and

generating adjusted scores s1 for said additional retrieved documents.

5. A computer system for assigning an adjusted relevancy score to information retrieved in response to a query, comprising:

a CPU, RAM and a database;

means for issuing a query on the database;

means for identifying a retrieved document based on the query;

means for receiving into said RAM a retrieved document and an accompanying variable s having a value corresponding to a relevance-ranking score of said retrieved document;

means for receiving into said RAM a variable q having a value corresponding to the number of words in the query and a variable v having a value corresponding to the overlap between the words in said retrieved document and in the query; and

means for generating, dependent on said variables s, q and v, an adjusted score s1 equal to the value of said variable s increased by an amount dependent on the value of said variable v, said amount decreasing as the value of said variable q increases.

6. The computer system of claim 5 wherein:

said means for receiving a retrieved document comprises means for receiving additional retrieved documents and accompanying variables s; and

said means for generating generates adjusted scores s1 for said additional retrieved documents; and

further comprising means for ordering said retrieved documents according to said adjusted scores s1.

7. The computer system of claim 5, further comprising:

means for receiving a variable ∂

chosen to control the increase to said variable s by said amount dependent on said variable v.

8. The computer system of claim 5 wherein:

said means for receiving document comprises means for receiving additional retrieved documents and accompanying variables s; and

said means for generating generates adjusted scores s1 for said additional retrieved documents.

9. A computer system for assigning an adjusted relevancy score to information retrieved in response to a query, comprising:

a CPU, RAM and a database;

means for issuing a query on the database;

means for identifying a retrieved document based on the query;

means for receiving into said RAM said retrieved document;

means responsive to a variable s having a value corresponding to a relevance-ranking score of said retrieved document;

means responsive to a variable q having a value corresponding to the number of words in the query and to a variable v having a value corresponding to the overlap between the words in said retrieved document and in the query; and

a function generator for receiving said variables s, q and v and responsively generating an adjusted score s1 equal to the value of said variable s increased by an amount dependent on the value of said variable v, said amount decreasing as the value of said variable q increases.

10. The computer system of claim 9 wherein said means responsive to a variable s is responsive to additional variables s having values corresponding to relevance-ranking scores of accompanying retrieved documents;

and

said function generator receives said additional variables s and responsively generates corresponding adjusted scores s1 for said additional variables s; and

further comprising

means for ordering said retrieved documents according to adjusted scores s1.

11. The computer system of claim 9, further comprising:

means for receiving a variable ∂

chosen to control the increase to said variable s by said amount dependent on said variable v.

12. The computer system of claim 9 wherein:

said means responsive to a variable s is responsive to additional variables s having values corresponding to relevance-ranking scores of additional retrieved documents; and

said function generator receives said additional variables s and responsively generates corresponding adjusted scores s1 for said additional variables s.

13. A method for a computer system having a CPU, RAM, and a database to assign a relevancy score to information retrieved in response to a query, comprising the steps of:

issuing a query on said database;

identifying a retrieved document based on the query;

receiving into said RAM said retrieved document and an accompanying variable s having a value corresponding to a relevance-ranking score of said retrieved document;

receiving into said RAM said variable f1 having a value dependent on the number of words in the query and on a value corresponding to the overlap between words in said retrieved document and in the query; and

using said CPU to add said variable f1 to said variable s to generate a function B equal to the value of said variable s increased by an amount dependent on the value of said variable f1, said amount decreasing as the number of words in the query increases.

14. The method of claim 13, further comprising the steps of:

receiving into said RAM a variable f2 having a value dependent on the number of words in the query; and

using said CPU to multiply said variable f2 with said function B to produce a scaled function f3 having a value in the range from 0 to 1 and corresponding to the value of said variable s increased by an amount dependent on the value of said variable f1, said amount decreasing as the number of words in the query increases.

15. The method of claim 14, further comprising the step of:

adjusting said variable f1 to control the increase to said variable s by said amount dependent on said variable f1.

16. The method of claim 14, further comprising the steps of:

receiving into said RAM additional retrieved documents and accompanying variables s;

generating scaled functions f3 for said additional retrieved documents; and

ordering said retrieved documents according to said scaled functions f3.

17. The method of claim 13, further comprising the steps of:

receiving additional retrieved documents and accompanying variables s; and

generating additional functions B for said additional retrieved documents.

18. A computer system for assigning an adjusted relevancy score to information retrieved in response to a query, comprising:

means for issuing a query on a database;

means for identifying a retrieved document based on said query;

means for receiving into a retrieved document and an accompanying variable s having a value corresponding to a relevance-ranking score of said retrieved document;

means for receiving a variable f1 having a value dependent on the number of words in the query and on a value corresponding to the overlap between the words in said retrieved document and in the query; and

means for adding said variable s to said variable f1 to generate a function B corresponding to the value of said variable s increased by an amount dependent on the value of said variable f1.

19. The computer system of claim 18 further comprising:

means for receiving a variable f2 having a value dependent on said number of words in the query; and

means for multiplying said variable f2 with said function B to produce a scaled function f3 having a range from 0 to 1 and corresponding to the value of said variable s increased by an amount dependent on the value of said variable f1.

20. The computer system of claim 19, further comprising:

means for adjusting said variable f1 to control the increase to said variable s by said amount dependent on said variable f1.

21. The computer system of claim 18, further comprising:

means for determining values of said scaled function f3 corresponding to other retrieved documents; and

means for ordering said retrieved documents based on the corresponding values of said scaled function f3.

22. The computer system of claim 18, further comprising:

means for adjusting said variable f1 to control the increase to said variable s by said amount dependent on said variable f1.

23. The computer system of claim 18 wherein:

said means for receiving a retrieved document and an accompanying variable s is set to further receive additional retrieved documents and accompanying variables s; and

said means for adding is set to generate additional functions B corresponding to the values of said accompanying variables s increased by said amount dependent on the value of said variable f1.

24. A computer system for assigning an adjusted relevancy score to information retrieved in response to a query, comprising:

a function generator for producing a function s having a value corresponding to a relevance-ranking score of a retrieved document;

a variable f1 having a value dependent on the number of words in the query and on a value corresponding to the overlap between the words in said retrieved document and in the query; and

an adder, coupled to said function generator, for adding said variable f1 to said variable s to generate a function B corresponding to the value of said variable s increased by an amount dependent on the value of said variable f1, said amount decreasing as the number of words in the query increases.

25. The computer system of claim 24, further comprising:

means for receiving a variable f2 having a value dependent on said number of words in the query; and

a multiplier for multiplying said variable f2 with said functions B to produce a scaled function f3 having a range from 0 to 1 and corresponding to the value of said variable s increased by an amount dependent on the value of said variable f1, said amount decreasing as the number of words in the query increases.

26. The computer system of claim 25, further comprising:

an adjusting unit for adjusting said variable f1 to control the increase to said variable s by said amount dependent on said variable f1.

27. A method for a computer system having a CPU and RAM to assign an adjusted relevancy score to information identified in response to a query, comprising the steps of:

receiving into said RAM a variable s having a value corresponding to a relevance-ranking score of an identified document;

receiving into said RAM a variable q having a value corresponding to the number of words in the query and a variable v having a value corresponding to the overlap between the number of words in said identified document and in the query; and

using said CPU and said variables s, q and v to generate an adjusted score s1 corresponding to the value of said variable s increased by an amount proportional to the value of said variable v, said amount decreasing as the value of said variable q increases.

28. The method of claim 27, further comprising the steps of:

receiving into said RAM additional variables s having values corresponding to relevance-ranking scores of additional identified documents; and

generating adjusted scores s1 for said additional identified documents.

29. The method of claim 28 further comprising the step of:

ordering said adjusted scores s1 for ranking said identified documents in response to the query.

30. The method of claim 27, further comprising the step of:

receiving into said RAM a variable ∂

chosen to control the increase to said variable s by said amount dependent on said variable v.

31. A computer system for identifying information in response to a query, comprising:

means for receiving a variable s having a value corresponding to a relevance-ranking score of an identified document;

means for receiving a variable q having a value corresponding to the number of words in the query and a variable v having a value corresponding to the overlap between the words in said identified document and in the query; and

means for generating, dependent on said variables s, q and v, an adjusted score s1 corresponding to the value of said variable s increased by an amount dependent on the value of said variable v, said amount decreasing as the value of said variable q increases.

32. The computer system of claim 31 wherein:

said means for receiving a variable s receives additional variables s corresponding to relevance ranking scores of additional identified documents; and

said means for generating generates adjusted scores s1 for said additional identified documents.

33. The computer system of claim 32 further comprising:

means for ordering said adjusted scores s1 for ranking said identified documents in response to the query.

34. The computer system of claim 31, further comprising:

means for receiving a variable ∂

chosen to control the increase to said variable s by said amount dependent on said variable v.

35. A method for a computer system to identify information in response to a query, comprising the steps of:

receiving a variable s having a value corresponding to a relevance-ranking score of an identified document;

receiving a variable f1 having a value dependent on the number of words in the query and on a value corresponding to the overlap between words in said identified document and in the query; and

adding said variable f1 to said variable s to generate a function B corresponding the value of said variable s increased by an amount dependent on the value of said variable f1, said amount decreasing as the number of words in the query increases.

36. The method of claim 35, further comprising the steps of:

receiving a variable f2 having a value dependent on the number of words in the query; and

multiplying said variable f2 with said function B to produce a scaled function f3 having a value in the range from 0 to 1 and corresponding to the value of said variable s increased by an amount dependent on the value of said variable f1, said amount decreasing as the number of words in the query increases.

37. The method of claim 35, further comprising the step of:

adjusting said variable f1 to control the increase to said variable s by said amount dependent on said variable f1.

38. The method of claim 35, further comprising the steps of:

receiving additional variables s corresponding to relevance-ranking scores of additional identified documents; and

generating additional functions B for said additional identified documents.

39. A computer system for identifying information in response to a query, comprising:

means for receiving a variable s having a value corresponding to a relevance-ranking score of an identified document;

means for receiving a variable f1 having a value dependent on the number of words in the query and on a value corresponding to the overlap between the words in said identified document and in the query; and

means for adding said variable s to said variable f1 to generate a function B corresponding to the value of said variable s increased by an amount dependent on the value of said variable f1.

40. The computer system of claim 39 further comprising:

means for receiving a variable f2 having a value dependent on said number of words in the query; and

means for multiplying said variable f2 with said function B to produce a scaled function f3 having a value in the range from 0 to 1 and corresponding to the value of said variable s increased by an amount dependent on the value of said variable f1.

41. The computer system of claim 39, further comprising:

means for adjusting said variable f1 to control the increase to said variable s by said amount dependent on said variable f1.

42. The computer system of claim 39 wherein said means for adding generates additional functions B corresponding to other identified documents.

43. The computer system of claim 42 further comprising:

means for ordering said identified documents based on the corresponding values of said functions B.

44. A program recorded in a computer-readable medium for causing a computer to perform the steps of:

receiving a variable s having a value corresponding to a relevance-ranking score of an identified document;

receiving a variable q having a value corresponding to the number of words in the query and a variable v having a value corresponding to the overlap between the words in said identified document and in the query; and

using said variables s, q and v to generate an adjusted score s1 corresponding to the value of said variable s increased by an amount dependent on the value of said variable v, said amount decreasing as the value of said variable q increases.

45. A program recorded in a computer-readable medium for causing a computer to perform the steps of:

receiving a variable s having a value corresponding to a relevance-ranking score of an identified document;

receiving a variable f1 having a value dependent on the number of words in the query and on a value corresponding to the overlap between words in said identified document and in the query; and

adding said variable f1 to said variable s to generate a function B corresponding the value of said variable s increased by an amount dependent on the value of said variable f1, said amount decreasing as the number of words in the query increases.