Contextual display of query refinements

• US 7,917,528 B1
• Filed: 04/02/2007
• Issued: 03/29/2011
• Est. Priority Date: 04/02/2007
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Whitelaw, Casey, Dave, Kushal, Battle, Alexis J.
• Primary Examiner(s): Cottingham; John R.
• Assistant Examiner(s): Allen; Nicholas E
• Application Number:
  US11/695,177
• Time in Patent Office: 1,457 Days
• Field of Search: 707/999.005, 707/766
• US Class Current: 707/766
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence

Abstract

Systems and methods for providing refinements to queries based upon the context of a refinement indication. Refinements, for example, can be related to a previous search query and can be provided to the user based upon receipt of a refinement indication.

Claims

1. A method executed by a data processing apparatus, comprising:

detecting a query refinement indication associated with a first search query displayed in a search field in a user interface, the first search query comprising two or more query tokens, and the query refinement indication being an indication of a user selection, the user selection being in the search field and being a selection of at least one and less than all of the query tokens;

detecting a context associated with the query refinement indication based on the selected query tokens and one or more unselected query tokens of the query;

identifying one or more refinements to the first search query based upon the query refinement indication and the context associated with the query refinement indication, the one or more refinements being refinements to the selected query tokens of the first search query; and

ordering the one or more refinements based on recorded votes on each of the one or more refinements, the votes corresponding to each word break in a region of overlap between the query tokens in the search query and refinement tokens in the refinement; and

providing the refinements according to the order of the refinements for display in the user interface.

2. The method of claim 1, comprising presenting a list of at least one of the one or more refinements to the first search query.

3. The method of claim 2, comprising:

ordering the list according to a quality measure of results associated with the one or more refinements.

4. The method of claim 2, comprising:

ordering the list according to a popularity measure of refinements, derived from a popularity of search queries associated with the one or more refinements.

5. The method of claim 1, wherein identifying one or more refinements comprises:

mapping the selected query tokens to one or more refinement tokens.

6. The method of claim 1, where ordering the one or more refinements comprises:

comparing each of the one or more refinements to the search query, including determining whether the refinement includes a common prefix or common suffix with the search query.

7. The method of claim 6, where each of the votes is recorded for a word break dividing the search query into portions, at least one of the portions of the search query is within the region of overlap.

8. A search system, comprising:

a search engine to provide a search interface to a client, the search interface to facilitate entry of a search query, to display the search query in a search field in a user interface, and to receive a refinement request comprising context associated with the refinement request;

a refinement store to store a plurality of refinements based upon the search query;

a tokenization engine to tokenize the search query into two or more query tokens, wherein the refinement request is determined based on an indication of a user selection, the user selection being in the search field and being a selection of at least one and less than all of the query tokens, and the context is determined based on the selected query tokens and one or more unselected query tokens of the query;

a refinement engine to retrieve one or more refinements from the refinement store and to process the retrieved refinements to prepare the retrieved refinements for filtering or reordering, the retrieved refinements for the selected query tokens, where the refinement engine is configured to perform operations comprising;

recording votes on each of the one or more refinements, the votes corresponding to each word break in a region of overlap between the query tokens in the search query and refinement tokens in the refinement; and

ordering the one or more refinements based on recorded votes; and

the search engine to serve the retrieved refinements.

9. The system of claim 8, wherein the search engine provides results from the search query to a user, the results comprising an agent embedded within the results, wherein the agent receives the refinement request through the search interface and filters or reorders the retrieved refinements for presentation to the user using the search interface.

10. The system of claim 8, wherein the refinement engine retrieves a plurality of refinements responsive to the refinement request, and the search engine serves the retrieved refinements to a user responsive to the refinement request.

11. The system of claim 10, wherein the refinement engine reorders or filters the retrieved one or more refinements based on the context.

12. The system of claim 11, wherein the one or more refinements are reordered based on quality of results associated with the retrieved refinements, respectively.

13. The system of claim 11, wherein the one or more refinements are reordered based on popularity of search queries associated with the retrieved refinements, respectively.

14. The system of claim 8, wherein the refinement engine defines a map from the two or more query tokens in the search query to retrieved refinements preferred for the two or more query tokens, respectively, and iterates over the refinements to provide mappings to the two or more query tokens in the search query.

15. The system of claim 8, wherein the refinement engine identifies the region of overlap between a set of query tokens comprising the search query and a set of refinement tokens comprising a selected retrieved refinement, records one vote for each word break in the overlapping region, defines mappings between the query tokens and the refinement tokens in the non-overlapping region, and sorts the refinements based upon the votes.

16. One or more computer readable media storing instructions that are executable by a processing device, and upon such execution cause the processing device to perform operations comprising:

receiving a search query comprising two or more query tokens, the search query displayed in a search field of a user interface;

detecting a query refinement indication associated with the search query, the query refinement indication being an indication of a user selection, the user selection being in the search field and being a selection of at least one and less than all of the query tokens;

detecting a context associated with the query refinement indication based on the selected query tokens and one or more unselected query tokens of the search query;

retrieving one or more refinements associated with the search query based upon the query refinement indication and the context, the retrieved refinements being refinements to the selected query tokens of the search query;

processing the one or more refinements, including ordering the one or more refinements based on recorded votes on each of the one or more refinements, the votes corresponding to each word break in a region of overlap between the query tokens in the search query and refinement tokens in the refinement;

embedding the ordered refinements within a web page comprising search results associated with the received search query; and

providing the web page for display.

17. The computer readable media of claim 16, wherein the web page spawns an agent upon being loaded by a browser.

18. The computer readable media of claim 17, the operations further comprising:

filtering the embedded refinements responsive to the context; and

reordering the filtered refinements responsive to the context.