Computer system for identifying storylines that emerge from highly ranked web search results

• US 7,895,196 B2
• Filed: 04/30/2004
• Issued: 02/22/2011
• Est. Priority Date: 04/30/2004
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Sivakumar, Dandapani, Mahadevan, Uma, Ravikumar, Shanmugasundaram
• Primary Examiner(s): Jalil; Neveen Abel
• Assistant Examiner(s): Syed; Farhan M
• Application Number:
  US10/835,983
• Publication Number:
  US 20050246321A1
• Time in Patent Office: 2,489 Days
• Field of Search: 707/2, 707/3, 707722-730
• US Class Current: 707/728
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9538 Presentation of query results

Abstract

This disclosure provides a method of and service for searching for locations (e.g., web sites) on a network (e.g., the internet). The method begins by inputting a query from a user and searching the internet based on the query to produce search results (comprising web sites) in an order of relevance, where more relevant results are ordered ahead of less relevant results. However, rather than directly reporting the search results to the user as conventional search methodologies do, the invention reorganizes the search results into storylines. More specifically, the invention identifies focused vocabularies from the search results, where each focused vocabulary forms a separate storyline. In addition, the invention maximizes the number of search results that are included in each storyline, minimizes the number of search results that are included in multiple storylines; and also minimizes the number of storylines. Then the invention reports the storylines to the user.

Claims

1. A method of searching locations on a computer network using a computer, said method comprising:

searching, using said computer, said network based on a query to produce search results; and

reorganizing, using said computer, said search results into storylines by;

identifying, using said computer, focused vocabularies from said search results, wherein each focused vocabulary forms a separate storyline;

maximizing, using said computer, the number of search results that are included in each storyline;

minimizing, using said computer, the number of search results that are included in multiple storylines; and

minimizing, using said computer, the number of storylines,

wherein said process of identifying focused vocabularies comprises evaluating distributions of terms within said search results and grouping search results with similar distributions together into said storylines,

wherein said process of evaluating distributions of said terms creates cliques of terms, and

wherein search results that have more than a predetermined percentage of terms of a given clique are grouped together.

2. The method in claim 1, wherein said process of minimizing the number of storylines comprises adjusting the level of uniformity of said distributions.

3. The method in claim 1, wherein said process of minimizing the number of storylines comprises adjusting the number of cliques produced by said process of evaluating distributions of terms.

4. The method in claim 1, wherein the processes of maximizing the number of search results included within each storyline and minimizing the number of storylines are given a higher priority than:

the uniformity of said focused vocabularies and minimizing the number of search results included within a storyline.

5. The method in claim 1, wherein search results are listed in said storylines in the order of relevance of said search results.

6. A method of searching locations on a computer network using a computer, said method comprising:

searching, using said computer, said network based on a query to produce search results; and

reorganizing, using said computer, said search results into storylines, by;

evaluating distributions of terms within said search results;

grouping search results with similar distributions together into said storylines;

maximizing, using said computer, the number of search results that are included in each storyline; and

minimizing, using said computer, the number of storylines,

wherein said process of evaluating distributions of said terms creates cliques of terms, and

wherein search results that have more than a predetermined percentage of terms of a given clique are grouped together.

7. The method in claim 6, wherein said process of minimizing the number of storylines comprises adjusting the level of uniformity of said distributions.

8. The method in claim 6, wherein said process of minimizing the number of storylines comprises adjusting the number of cliques produced by said process of evaluating distributions of terms.

9. The method in claim 6, wherein search results are listed in said storylines in the order of relevance of said search results.

10. A method of searching for web sites on the internet using a computer, said method comprising:

inputting, using said computer, a query from a user;

searching, using said computer, the internet based on said query to produce search results comprising said web sites in an order of relevance, where more relevant search results are ordered ahead of less relevant search results;

reorganizing, using said computer, said search results into storylines by;

identifying, using said computer, focused vocabularies from said search results, wherein each focused vocabulary forms a separate storyline;

maximizing, using said computer, the number of search results that are included in each storyline;

minimizing, using said computer, the number of search results that are included in multiple storylines; and

minimizing, using said computer, the number of storylines; and

reporting said storylines to said user,

wherein said process of identifying focused vocabularies comprises evaluating distributions of terms within said search results and grouping search results with similar distributions together into said storylines,

wherein said process of evaluating distributions of said terms creates cliques of terms, and

wherein search results that have more than a predetermined percentage of terms of a given clique are grouped together.

11. The method in claim 10, wherein said process of minimizing the number of storylines comprises adjusting the level of uniformity of said distributions.

12. The method in claim 10, wherein said process of minimizing the number of storylines comprises adjusting the number of cliques produced by said process of evaluating distributions of terms.

13. The method in claim 10, wherein the processes of maximizing the number of search results included within each storyline and minimizing the number of storylines are given a higher priority than:

the uniformity of said focused vocabularies and minimizing the number of search results included within a storyline.

14. The method in claim 10, wherein search results are listed in said storylines in the order of relevance.

15. A program storage device readable by machine, tangibly embodying a program of instructions executable by the machine to perform a method of searching locations on a computer network using a computer, said method comprising:

searching, using said computer, said network based on a query to produce search results; and

reorganizing, using said computer, said search results into storylines by;

identifying, using said computer, focused vocabularies from said search results, wherein each focused vocabulary forms a separate storyline;

maximizing, using said computer, the number of search results that are included in each storyline;

minimizing, using said computer, the number of search results that are included in multiple storylines; and

minimizing, using said computer, the number of storylines,

wherein said process of identifying focused vocabularies comprises evaluating distributions of terms within said search results and grouping search results with similar distributions together into said storylines,

wherein said process of evaluating distributions of said terms creates cliques of terms, and

wherein search results that have more than a predetermined percentage of terms of a given clique are grouped together.

16. The program storage device in claim 15, wherein said process of minimizing the number of storylines comprises adjusting the level of uniformity of said distributions.

17. The program storage device in claim 15, wherein said process of minimizing the number of storylines comprises adjusting the number of cliques produced by said process of evaluating distributions of terms.

18. The program storage device in claim 15, wherein the processes of maximizing the number of search results included within each storyline and minimizing the number of storylines are given a higher priority than:

the uniformity of said focused vocabularies and minimizing the number of search results included within a storyline.

19. The program storage device in claim 15, wherein search results are listed in said storylines in the order of relevance of said search results.

20. A service for searching locations on a computer network using a computer, said service comprising:

searching, using said computer, said network based on a query to produce search results; and

reorganizing, using said computer, said search results into storylines by;

identifying, using said computer, focused vocabularies from said search results, wherein each focused vocabulary forms a separate storyline;

maximizing, using said computer, the number of search results that are included in each storyline;

minimizing, using said computer, the number of search results that are included in multiple storylines; and

minimizing, using said computer, the number of storylines,

wherein said process of identifying focused vocabularies comprises evaluating distributions of terms within said search results and grouping search results with similar distributions together into said storylines,

wherein said process of evaluating distributions of said terms creates cliques of terms, and

wherein search results that have more than a predetermined percentage of terms of a given clique are grouped together.

21. The service in claim 20, wherein said process of minimizing the number of storylines comprises adjusting the level of uniformity of said distributions.

22. The service in claim 20, wherein said process of minimizing the number of storylines comprises adjusting the number of cliques produced by said process of evaluating distributions of terms.

23. The service in claim 20, wherein the processes of maximizing the number of search results included within each storyline and minimizing the number of storylines are given a higher priority than:

the uniformity of said focused vocabularies and minimizing the number of search results included within a storyline.

24. The service in claim 20, wherein search results are listed in said storylines in the order of relevance of said search results.