Method and system for information retrieval with clustering

• US 8,676,802 B2
• Filed: 11/30/2006
• Issued: 03/18/2014
• Est. Priority Date: 11/30/2006
• Status: Active Grant

• Current Assignee: Oracle OTC Subsidiary LLC (Oracle Corporation)
• Original Assignee: Oracle OTC Subsidiary LLC (Oracle Corporation)
• Inventors: Zelevinsky, Vladimir V., Tunkelang, Daniel, Knabe, Frederick C., Saji, Michael Y., Tzanov, Velin Krassimirov
• Primary Examiner(s): Ehichioya, Fred I
• Assistant Examiner(s): Meng, Thomas
• Application Number:
  US11/606,469
• Publication Number:
  US 20080133479A1
• Time in Patent Office: 2,665 Days
• Field of Search: None
• US Class Current: 707/737
• CPC Class Codes: G06F 16/3325 Reformulation based on results of preceding query
• CPC Class Codes: G06F 16/338 Presentation of query results
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification

Abstract

Methods and systems that enable searching with clustering in information access systems are described. The methods of clustering operate on a collection of materials wherein each item in the collection may be associated with one or more properties. An original subset of materials is selected from the collection and relevant properties associated with the subset of materials are clustered into property clusters. Each property cluster generally contains properties that are more similar to each other than to properties in a different property cluster. The property clusters can be used to respond to the query. A mapping function can be used to identify a set of materials that correspond to each property cluster based on the associations between individual items and properties. The property clusters can also be used for iterative query refinement.

Claims

1. A non-transitory computer readable medium having instructions stored thereon that cause a processor to retrieve documents in response to at least one search term from a user, the retrieving the documents comprising:

receiving the search term for searching a plurality of text documents, wherein each text document is associated with one or more salient terms extracted from the document and each text document is associated with one or more properties that represent the one or more extracted salient terms;

retrieving a first set of retrieved documents from a query of the plurality of text documents, wherein each of the retrieved documents comprises the search term;

retrieving the associated salient terms for each of the retrieved documents and the associated properties;

grouping based on a distance metric the retrieved salient terms into one or more clusters of salient terms and providing the clusters of salient terms to the user, wherein each of the cluster of salient terms corresponds to one of the properties associated with the retrieved documents and each cluster displays the associated salient terms;

receiving a selection of a first cluster of the clusters of salient terms from the user, wherein the first cluster comprises first salient terms;

selecting a second set of retrieved documents from the first set of retrieved documents, wherein each second set document of the second set includes at least one of the first salient terms of the first cluster of salient terms;

retrieving associated second salient terms for each of the second set documents; and

grouping the second salient terms into one or more second clusters of salient terms and providing the second clusters of salient terms to the user.

2. The non-transitory computer readable medium of claim 1, wherein each second set document includes two or more of the first salient terms of the first cluster.

3. The non-transitory computer readable medium of claim 1, the retrieving the documents further comprising:

receiving a selection of a second cluster from one of the second clusters from the user, wherein the second cluster comprises second salient terms; and

selecting a third set of retrieved documents from the second set of retrieved documents, wherein each third set document includes at least one of the second salient terms of the second cluster.

4. The non-transitory computer readable medium of claim 1, wherein each of the clusters provided to the user corresponds to a synonym of the search term.

5. The non-transitory computer readable medium of claim 1, the retrieving the documents further comprising, after selecting the second set of retrieved documents, iteratively repeating the retrieving the associated salient terms, grouping the retrieved salient terms into one or more clusters of salient terms, receiving the selection of a first cluster, and selecting a second set of retrieved documents in response to additional selections of a cluster by the user.

6. The non-transitory computer readable medium of claim 1, wherein the providing the clusters to the user comprises displaying on a user interface for each cluster provided to the user one or more salient terms for the cluster.

7. The non-transitory computer readable medium of claim 6, wherein a selection of one of the displayed clusters causes a corresponding set of documents to be displayed on the user interface.

8. The non-transitory computer readable medium of claim 6, further displaying on the user interface a list of related terms associated with the first set of retrieved documents.

9. The non-transitory computer readable medium of claim 6, wherein a selection of one of the displayed clusters triggers a new query based on one or more of corresponding terms of the displayed cluster.

10. The non-transitory computer readable medium of claim 1, the retrieving the documents further comprising mapping each of the first set of retrieved documents to at least one of the clusters provided to the user.

11. The non-transitory computer readable medium of claim 10, wherein the mapping comprises at least one of:

match-any, match-all, or match-partial.

12. A computer implemented method for retrieving text documents, the method comprising:

receiving at least one search term for searching a plurality of text documents, wherein each text document is associated with one or more salient terms extracted from the document and each text document is associated with one or more properties that represent the one or more extracted salient terms;

retrieving a first set of retrieved documents from a query of the plurality of text documents, wherein each of the retrieved documents comprises the search term;

retrieving the associated salient terms for each of the retrieved documents and the associated properties;

grouping based on a distance metric the retrieved salient terms into one or more clusters of salient terms and providing the clusters of salient terms to the user, wherein each of the cluster of salient terms corresponds to one of the properties associated with the retrieved documents and each cluster displays the associated salient terms;

receiving a selection of a first cluster of the clusters of salient terms from the user, wherein the first cluster comprises first salient terms;

selecting a second set of retrieved documents from the first set of retrieved documents, wherein each second set document of the second set includes at least one of the first salient terms of the first cluster of salient terms;

retrieving associated second salient terms for each of the second set documents; and

grouping the second salient terms into one or more second clusters of salient terms and providing the second clusters of salient terms to the user.

13. The method of claim 12, wherein each second set document includes two or more of the first salient terms of the first cluster.

14. The method of claim 12, the method further comprising:

receiving a selection of a second cluster from one of the second clusters from the user, wherein the second cluster comprises second salient terms; and

selecting a third set of retrieved documents from the second set of retrieved documents, wherein each third set document includes at least one of the second salient terms of the second cluster.

15. The method of claim 12, wherein each of the clusters provided to the user corresponds to a synonym of the search term.

16. The method of claim 12, the method further comprising, after selecting the second set of retrieved documents, iteratively repeating the retrieving the associated salient terms, grouping the retrieved salient terms into one or more clusters of salient terms, receiving the selection of a first cluster, and selecting a second set of retrieved documents in response to additional selections of a cluster by the user.

17. The method of claim 12, wherein the providing the clusters to the user comprises displaying on a user interface for each cluster provided to the user one or more salient terms for the cluster.

18. The method of claim 17, wherein a selection of one of the displayed clusters causes a corresponding set of documents to be displayed on the user interface.

19. The method of claim 17, further displaying on the user interface a list of related terms associated with the first set of retrieved documents.

20. The method of claim 17, wherein a selection of one of the displayed clusters triggers a new query based on one or more of corresponding terms of the displayed cluster.

21. The method of claim 12, the method further comprising mapping each of the first set of retrieved documents to at least one of the clusters provided to the user.

22. The method of claim 21, wherein the mapping comprises at least one of:

match-any, match-all, or match-partial.

23. An information retrieval system comprising:

a processor coupled to a computer readable medium having instructions stored thereon, wherein the processor executing the instructions implements modules comprising;

an interface module that receives at least one search term and searches a plurality of text documents, wherein each text document is associated with one or more salient terms extracted from the document and each text document is associated with one or more properties that represent the one or more extracted salient terms;

a database access module that retrieve a first set of retrieved documents from a query of the plurality of text documents, wherein each of the retrieved documents comprises the search term;

a matching module that retrieves the associated salient terms for each of the retrieved documents and the associated properties;

a clustering module that groups based on a distance metric the retrieved salient terms into one or more clusters of salient terms and provides the clusters of salient terms to the user, wherein each of the cluster of salient terms corresponds to one of the properties associated with the retrieved documents and each cluster displays the associated salient terms;

the interface module that receives a selection of a first cluster of the clusters of salient terms from the user, wherein the first cluster comprises first salient terms;

the matching module selecting a second set of retrieved documents from the first set of retrieved documents, wherein each second set document of the second set includes at least one of the first salient terms of the first cluster of salient terms;

further comprising retrieving associated second salient terms for each of the second set documents; and

grouping the second salient terms into one or more second clusters of salient terms and providing the second clusters of salient terms to the user.

24. The system of claim 23, wherein each second set document includes two or more of the first salient terms of the first cluster.

25. The system of claim 23, further comprising:

receiving a selection of a second cluster from one of the second clusters from the user, wherein the second cluster comprises second salient terms; and

selecting a third set of retrieved documents from the second set of retrieved documents, wherein each third set document includes at least one of the second salient terms of the second cluster.

26. The system of claim 23, wherein each of the clusters corresponds to a synonym of the search term.

27. The system of claim 23, further comprising, after selecting the second set of retrieved documents, iteratively repeating the retrieving the associated salient terms, grouping the retrieved salient terms into one or more clusters of salient terms, receiving the selection of a first cluster, and selecting a second set of retrieved documents in response to additional selections of a cluster by the user.

28. The system of claim 23, wherein the providing the clusters to the user comprises displaying on a user interface for each cluster provided to the user one or more salient terms for the cluster.

29. The system of claim 23, wherein a selection of one of the displayed clusters causes a corresponding set of documents to be displayed on the user interface.

30. The system of claim 28, further displaying on the user interface a list of related terms associated with the first set of retrieved documents.

31. The system of claim 23, wherein a selection of one of the displayed clusters triggers a new query based on one or more of corresponding terms of the displayed cluster.

32. The system of claim 23, further comprising mapping each of the first set of retrieved documents to at least one of the clusters provided to the user.

33. The system of claim 32, wherein the mapping comprises at least one of:

match-any, match-all, or match-partial.