Fixed phrase detection for search

Patent metadata

Field	Value
Publication number	US-10055476-B2
Application number	US-201414297474-A
Country	US
Kind code	B2
Filing date	Jun 5, 2014
Priority date	Feb 24, 2010
Publication date	Aug 21, 2018
Grant date	Aug 21, 2018

Abstract

A set of search requests may be analyzed to detect fixed phrases suitable for inclusion in a search index. Sets of candidate phrases may be identified among the search requests. Fixed phrases may be detected among the candidate phrases using statistical techniques, for example, by identifying phrases having a relatively high pointwise mutual information (PMI) with respect to component keywords. Fixed phrase detection may include keyword and/or phrase clustering. Clusters may correspond to topics defined using a latent Dirichlet allocation (LDA) procedure. Fixed phrase detection may include identifying phrases having relatively high PMI within particular clusters.

First claim

What is claimed is:

1. A computer-implemented method for searching a collection of content, comprising:
detecting at least one interaction with search result content, the search result content part of a collection of content provided through an electronic marketplace;
detecting at least one fixed phrase based at least in part on the detected at least one interaction with the search result content, the search result content associated with a column in a relevance matrix, the relevance matrix maintained in a data store in a distributed environment that utilizes a plurality of network resources available over a network, the at least one fixed phrase associated with a row in the relevance matrix, the relevance matrix corresponding to a data structure of a content index that indexes the collection of content;
determining a position in the relevance matrix corresponding to an intersection of the column and the row;
retrieving at least one relevance score at the position to identify at least one determined relevance score for facilitating a search; and
updating the content index based at least in part on the at least one interaction and the at least one determined relevance score, the at least one determined relevance score being included in another search of the content index upon updating the relevance matrix.

2. The computer-implemented method of claim 1 , further comprising:
presenting at least one search result, the at least one search result referencing content selected from the content index.

3. The computer-implemented method of claim 1 , wherein detecting the at least one interaction with the search result content further comprises at least one of moving a display pointer to a search result portion or selecting a search result portion.

4. The computer-implemented method of claim 1 , wherein detecting the at least one fixed phrase further comprises utilizing a parsing interface.

5. The computer-implemented method of claim 1 , wherein determining the at least one relevance score associated with the detected at least one fixed phrase further comprises analyzing a search history.

6. The computer-implemented method of claim 1 , wherein updating the content index further comprises at least one of:
adding a new row to the relevance matrix;
periodic updating;
scheduled updating;
updating based on at least one interaction with the search result content since a previous content index update;
updating during a build mode of a search engine operation;
updating during a live mode of the search engine operation;
updating a search history; and
updating based on interaction statistics.

7. A computerized system for searching a collection of content, comprising:
at least one processor; and
memory including instructions that, when executed by the at least one processor, cause the computerized system to:
detect at least one interaction with search result content, the search result content part of a collection of content provided through an electronic marketplace;
detect at least one fixed phrase based at least in part on the detected at least one interaction with the search result content, the search result content associated with a column in a relevance matrix, the relevance matrix maintained in a data store in a distributed environment that utilizes a plurality of network resources available over a network, the at least one fixed phrase associated with a row in the relevance matrix, the relevance matrix corresponding to a data structure of a content index that indexes the collection of content;
determine a position in the relevance matrix corresponding to an intersection of the column and the row;
retrieve at least one relevance score at the position to identify at least one determined relevance score for facilitating a search; and
update the content index based at least in part on the at least one interaction and the at least one determined relevance score, the at least one determined relevance score being included in another search of the content index upon updating the relevance matrix.

8. The computerized system of claim 7 , the instructions further causing the computerized system to:
present at least one search result, the at least one search result referencing content selected from the content index.

9. The computerized system of claim 7 , wherein detecting the at least one interaction with the search result content further comprises at least one of moving a display pointer to a search result portion or selecting a search result portion.

10. The computerized system of claim 7 , wherein detecting the at least one fixed phrase further comprises utilizing a parsing interface.

11. The computerized system of claim 7 , wherein determining the at least one relevance score associated with the detected at least one fixed phrase further comprises analyzing a search history.

12. The computerized system of claim 7 , wherein updating the content index further comprises at least one of:
adding a new row to the relevance matrix;
periodic updating;
scheduled updating;
updating based on at least one interaction with the search result content since a previous content index update;
updating during a build mode of a search engine operation;
updating during a live mode of the search engine operation;
updating a search history; and
updating based on interaction statistics.

13. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to:
detect at least one interaction with search result content, the search result content part of a collection of content provided through an electronic marketplace;
detect at least one fixed phrase based at least in part on the detected at least one interaction with the search result content, the search result content associated with a column in a relevance matrix, the relevance matrix maintained in a data store in a distributed environment that utilizes a plurality of network resources available over a network, the at least one fixed phrase associated with a row in the relevance matrix, the relevance matrix corresponding to a data structure of a content index that indexes the collection of content;
determine a position in the relevance matrix corresponding to an intersection of the column and the rows;
retrieve at least one relevance score at the position to identify at least one determined relevance score for facilitating a search; and
update the content index based at least in part on the at least one interaction and the at least one determined relevance score, the at least one determined relevance score being included in another search of the content index upon updating the relevance matrix.

14. The non-transitory computer-readable storage medium of claim 13 , the instructions further causing the processor to:
present at least one search result, the at least one search result referencing content selected from the content index.

15. The non-transitory computer-readable storage medium of claim 13 , wherein detecting the at least one interaction with the search result content further comprises at least one of moving a display pointer to a search result portion or selecting a search result portion.

16. The non-transitory computer-readable storage medium of claim 13 , wherein detecting the at least one fixed phrase further comprises utilizing a parsing interface.

17. The non-transitory computer-readable storage medium of claim 13 , wherein determining the at least one relevance score associated with the detected at least one fixed phrase further comprises analyz

Assignees

• A9 Com Inc

Inventors

• Ahmad Waseem
• Jain Deepak

Classifications

• G06F17/30598 Primary

  Physics · mapped topic

• G06F17/30864

  Physics · mapped topic

• G06F17/30616

  Physics · mapped topic

• G06F16/951 Primary

  Indexing; Web crawling techniques · CPC title

• G06F16/313

  Selection or weighting of terms for indexing · CPC title