System and method for generating a relationship network

• US 9,026,543 B2
• Filed: 07/25/2011
• Issued: 05/05/2015
• Est. Priority Date: 06/06/2005
• Status: Active Grant

• Current Assignee: Regents of the University of California (University of California)
• Original Assignee: Regents of the University of California (University of California)
• Inventors: Franks, Kasian, Myers, Cornelia A., Podowski, Raf M.
• Primary Examiner(s): Saeed, Usmaan
• Assistant Examiner(s): Kim, Paul
• Application Number:
  US13/190,365
• Publication Number:
  US 20120054206A1
• Time in Patent Office: 1,380 Days
• Field of Search: 707/736, 707/748, 707/750
• US Class Current: 707/750
• CPC Class Codes: G06F 16/3347 using vector based model

Abstract

A computer-implemented system and process for generating a relationship network is disclosed. The system provides a set of data items to be related and generates variable length data vectors to represent the relationships between the terms within each data item. The system can be used to generate a relationship network for documents, images, or any other type of file. This relationship network can then be queried to discover the relationships between terms within the set of data items.

Claims

1. A system for discovering data items related to a search query, comprising:

a processor;

a relationship database stored in a computer memory and comprising a set of data items to be related, wherein each data item comprises a plurality of terms; and

a network generation engine executing on the processor and configured to;

receive a query object;

based on the query object, extract all direct and associated relationships from the relationship database for the query object;

score and rank the extracted direct and associated relationships;

measure the similarity score of any intersecting objects by taking a sum of the relationship scores from any intersecting terms and multiplying the sum by the length of a vector composed only of the intersecting terms;

build a relationship network comprising the relationship of the query object to other objects, wherein the strength of the relationship with the other objects is based on the measured similarity score;

execute a filter to provide thematic context for controlling the type of relationships extracted within the relationship network; and

an output for outputting data items related to the query object and within the thematic context.

2. The system of claim 1, wherein the relationship network is comprised of a network of relationship vectors.

3. The system of claim 2, wherein the relationship vectors comprise connections to each other, and the strength of those connections are based on shared unique attributes within the thematic context.

4. The system of claim 1, wherein the relationship database has had common language words removed from the data items.

5. The system of claim 1, wherein the network generation engine is configured to extract all direct and associated relationships from the relationship database by retrieving a query object vector (“

QOV”

) for the query for the query object.

6. The system of claim 5, wherein the relationship database comprises associated object vectors configured to represent the relationships between data items in the database.

7. The system of claim 6, wherein the network generation engine is configured to use the QOV to generate expanded associated object vectors.

8. The system of claim 1, wherein the filter is based on a word, symbol or object provided to the network generation engine.

9. The system of claim 1, wherein the network generation engine is configured to receive filter information along with the query object to focus the query results.

10. The system of claim 1, wherein the filter is an exclusion filter that is configured to remove terms and vectors which do not match the filter.

11. The system of claim 1, wherein the filter is a selection filter that is configured to select terms and vectors which match the filter.

12. The system of claim 1, wherein the filter is a weighing filter that is configured to adjust relationship scores of specific terms or vectors to cause the terms or vectors to be reordered in any search results.

13. An electronic method performed by a processor of discovering data items related to a search query, comprising:

providing a relationship database comprising a set of data items to be related, wherein each data item comprises a plurality of terms;

receiving a query object;

extracting all direct and associated relationships from the relationship database for the query object based on the query object;

scoring and ranking the extracted direct and associated relationships;

measuring the similarity score of any intersecting objects by taking a sum of the relationship scores from any intersecting terms and multiplying the sum by the length of a vector composed only of the intersecting terms;

building a relationship network showing the relationship of the query object to other objects, wherein the strength of the relationship with the other objects is based on the measured similarity score;

executing a filter to provide thematic context for controlling the type of relationships extracted within the relationship network; and

outputting data items related to the query object and within the thematic context.

14. The method of claim 13, wherein the relationship network is comprised of a network of relationship vectors.

15. The method of claim 14, wherein building a relationship network comprises building relationship vectors comprising connections to each other, and the strength of those connections are based on shared unique attributes within the thematic context.

16. The method of claim 13, wherein the relationship database is provided with common language words removed from the data items.

17. The method of claim 13, wherein extracting all direct and associated relationships from the relationship database comprises retrieving a query object vector (“

QOV”

) for the query for the query object.

18. The method of claim 17, wherein the relationship database comprises associated object vectors configured to represent the relationships between data items in the database.

19. The method of claim 18, further comprising using the QOV to generate expanded associated object vectors.

20. The method of claim 13, wherein the filter is based on a word, symbol or object provided to the network generation engine.

21. The method of claim 13, wherein receiving a query object comprises receiving filter information to focus the query results.

22. The method of claim 13, wherein executing the filter comprises executing an exclusion filter that is configured to remove terms and vectors which do not match the filter.

23. The method of claim 13, wherein executing the filter comprises executing a selection filter that is configured to select terms and vectors which match the filter.

24. The method of claim 13, wherein executing the filter comprises executing a weighing filter configured to adjust relationship scores of specific terms or vectors to cause the terms or vectors to be reordered in any search results.