Performing application searches

• US 9,152,674 B2
• Filed: 03/15/2013
• Issued: 10/06/2015
• Est. Priority Date: 04/27/2012
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Quixey, Inc.
• Inventors: Glover, Eric, Shapira, Liron, Tarleton, Nicolas
• Primary Examiner(s): Ortiz Ditren, Belix M
• Application Number:
  US13/842,169
• Publication Number:
  US 20130290319A1
• Time in Patent Office: 935 Days
• Field of Search: 707/695, 707/711, 707/696, 707/706, 707/723
• US Class Current: 1/1
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/31 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/90 Details of database functions independent of the retrieved data types
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

According to some implementations, a technique for processing an application search query is disclosed. The techniques include receiving a search query from a partner, determining a set of subqueries based on the search query, and extracting query features of the search query from the search query. The techniques further include determining an initial result set of applications based on the set of subqueries and the query features. The initial result set includes a plurality of application representations each defining features of a different application. The techniques also include determining a score for each application in the initial result set of applications based on the set of query features and one or more scoring models, determining a result set based on the scores of the applications in the initial result set, and providing the result set to the partner.

Claims

1. A method for processing a search query, the method comprising:

receiving, at a processing device, the search query from a partner;

determining, at the processing device, a set of subqueries based on the search query;

extracting, at the processing device, query features of the search query from the search query;

determining, at the processing device, an initial result set of application representations based on the set of subqueries, the initial result set including a set of one or more application representations, each application representation being a data structure representing a different application and defining features of the application, wherein the features of the application are extracted from one or more documents obtained from one or more respective sources, each document relating to the application;

determining, at the processing device, a score for each application representation in the initial result set of application representations based on the query features and one or more machine-learned scoring models;

determining, at the processing device, a ranked result set based on the scores for the application representations of the initial result set, the ranked result set indicating one or more applications that correspond to the search query; and

providing, at the processing device, the ranked result set to the partner.

2. The method of claim 1, wherein determining the initial result set includes:

obtaining, at the processing device, a pre-consideration set of application representations from a search index based on the set of subqueries, wherein the set of one or more application representations is a subset of the pre-consideration set; and

deriving, at the processing device, the initial result set from the pre-consideration set based on the query features of the search query.

3. The method of claim 2, wherein obtaining the pre-consideration set includes identifying, at the processing device, a group of application representations by querying the search index with the set of subqueries, each identified application representation having a feature associated with at least one of the set of subqueries, wherein the set of one or more application representations is a subset of the identified group of application representations.

4. The method of claim 3, wherein deriving the initial result set includes:

determining, at the processing device, a consideration set of application representations from the pre-consideration set of application representations based on one or more features of the identified group of application representations, the consideration set including the set of one or more application representations; and

for each application representation in the consideration set of application representations;

determining, at the processing device, one or more query/result features, each query/result feature defining a feature of the application representation with respect to the search query; and

augmenting, at the processing device, features of the application representation with the one or more query/result features.

5. The method of claim 4, wherein determining the score of an application representation includes inputting, at the processing device, the features of the application representation, the query/result features of the application representation, and the query features into the one or more scoring models.

6. The method of claim 1:

wherein the search query includes one or more query terms, and

wherein the method further comprises generating a query wrapper that contains the search query in response to receiving the search query, wherein the query wrapper further contains contextual data relating to the partner.

7. The method of claim 6, wherein the query features are based on the query terms and the contextual data.

8. A server for processing a search query, the server comprising:

a storage device that stores;

a plurality of application representations, each application representation being a data structure representing a different application and including one or more features of the application, the features of the application being extracted from one or more documents obtained from one or more respective sources, each document relating to the application; and

a search index that indexes the plurality of application representations, each application representation representing a different application and including one or more features of the application;

a processing device that executes a query processing module, an initial set determination module, a set reduction module, a result processing module, a scoring module, and a post processing module;

the query processing module being configured to;

receive the search query from a partner;

determine a set of subqueries based on the search query; and

extract query features of the search query from the search query;

the initial set determination module being configured to identify a pre-consideration set of application representations from the search index based on the set of subqueries;

the set reduction module being configured to determine a consideration set of application representations from the pre-consideration set based on one or more result features of the identified pre-consideration set of application representations; and

the result processing module being configured to determine an initial result set of application representations based on the query features and result features of the application representations in the consideration set, the initial result set including a set of application representations, each application representation including features of a respective application and query/result features of the respective application, the query/result features defining features of the respective application in view of the query features;

the scoring module being configured to;

determine a score for each application representation in the initial result set of application representations based on the query features, the result features, the query/result features, and one or more machine-learned scoring models; and

determine a ranked result set based on the scores of the application representations in the initial result set; and

the post-processing module being configured to provide the ranked result set to the partner.

9. The server of claim 8, wherein the initial set determination module is configured to identify an application representation from a search index that has a feature implicated by at least one of the set of subqueries.

10. The server of claim 8, wherein the result processing module is configured to determine the initial result set by, for each application representation in the consideration set:

determining the one or more query/result features of the application representation; and

augmenting the features of the application representation with the one or more query/result features.

11. The server of claim 10, wherein the scoring module is configured to determine the score for each application representation in the initial result set by, for each application representation, inputting the features of the application representation, the query/result features of the application representation, and the query features into the one or more scoring models.

12. The server of claim 8, wherein the search query includes one or more query terms and is received in a query wrapper that further contains contextual data relating to the partner.

13. The server of claim 12, wherein the query features are based on the query terms and the contextual data.

14. A server for processing an application search query, the server comprising:

a storage device that stores;

a plurality of application representations, each application representation being a data structure representing a different application and including one or more features of the application, the features of the application being extracted from one or more documents obtained from one or more respective sources, each document relating to the application; and

a search index that indexes the plurality of application representations, each application representation representing a different application and including one or more features of the application;

a processing device that executes computer-readable instructions, the computer-executable instructions causing the processing device to;

receive a search query from a partner device;

determine a set of subqueries based on the search query;

extract query features of the search query from the search query;

determine an initial result set of application representations based on the set of subqueries and the search index, the initial result set including a set of one or more application representations from the plurality of application representations;

determine a score for each application representation in the initial result set of application representations based on the set of query features and one or more machine-learned scoring models;

determine a ranked result set based on the scores for the application representations of the initial result set, the ranked result set indicating one or more applications that correspond to the search query; and

provide the ranked result set to the partner.

15. The server of claim 14, wherein determining the initial result set includes:

obtaining a pre-consideration set of application representations from the search index based on the set of subqueries, wherein the set of one or more application representations is a subset of the pre-consideration set; and

deriving the initial result set from the pre-consideration set based on the query features of the search query.

16. The server of claim 15, wherein obtaining the pre-consideration set includes identifying a group of application representations by querying the search index with the set of subqueries, each identified application representation having a feature associated with at least one of the set of subqueries, wherein the set of one or more application representations is a subset of the identified group of application representations.

17. The server of claim 16, wherein deriving the initial result set includes:

determining a consideration set of application representations from the pre-consideration set of application representations based on one or more features of the identified group of application representations, the consideration set including the set of one or more application representations; and

for each application representation in the consideration set of application representations;

determining one or more query/result features, each query/result feature defining a feature of the application representation with respect to the search query; and

augmenting features of the application representation with the one or more query/result features.

18. The server of claim 17, wherein determining the score of an application representation includes inputting the features of the application representation, the query/result features of the application representation, and the query features into the one or more scoring models.

19. The server of claim 14:

wherein the search query includes one or more query terms, and

wherein, the computer-readable instructions further cause to the processing device to generate a query wrapper that contains the search query in response to receiving the search query, wherein the query wrapper further contains contextual data relating to the partner.

20. The server of claim 14 where the query features are based on the query terms and the contextual data.