Identifying similar applications

• US 9,569,536 B2
• Filed: 12/17/2013
• Issued: 02/14/2017
• Est. Priority Date: 12/17/2013
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Ronen, Royi, Lavee, Gal
• Primary Examiner(s): Pham, Hung Q
• Application Number:
  US14/109,782
• Publication Number:
  US 20150169759A1
• Time in Patent Office: 1,155 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/26 Visual data mining; Browsing structured data
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

Systems, methods, and media for proposing similar applications to a candidate application are provided. A similar results server may surface similar applications to a user of an application store. The applications can be selected based on similarity between features of the application. The features may be extracted from graphical user interface information rendered by a device that executes the applications. The applications that are identified as similar (e.g., based on the features) are displayed to the user.

Claims

1. A computer-implemented method for detecting similar applications, the method comprising:

executing applications in at least one operating environment to determine features available to the applications rendered during application processing;

extracting feature information from graphical user interface (GUI) elements of the applications, the GUI elements being rendered by the applications during execution of the applications;

storing the feature information in an application database;

creating feature vectors for each application, the feature vectors being based on the extracted feature information from the GUI elements;

assigning similarity scores to each application based on a distance between feature vectors, each similarity score of the similarity scores indicating a similarity of the GUI elements; and

returning suggestions that include results for a query application and one or more applications that are close in similarity score to the query application.

2. The method of claim 1, wherein the feature information is provided by publishers of the applications.

3. The method of claim 2, wherein the feature information is provided during a review period before the application enters an application store.

4. The method of claim 1, wherein the feature information is parsed by a screen reader.

5. The method of claim 1, wherein the feature information comprises binary information and weighted information.

6. The method of claim 5, wherein the binary information includes access to a video interface;

existence of button graphics, existence of menu graphics, existence of input elements, existence of links, access to microphone, access to network interfaces, access to other applications;

existence of battery consumption indicator, or existence of description text.

7. The method of claim 5, wherein the weighted information includes number of accesses to a video interface, number of button graphics, number of menu graphics, number of screens, number of input elements, number of links, number of accesses to microphone, number of accesses to network interfaces, number of accesses to other applications, percent content formatted in each media type, number of languages, number of daily launches, application version, number of levels, or number of co-purchased applications.

8. The method of claim 1, wherein the applications are stored in an applications store.

9. The method of claim 1, wherein the applications include game applications or productivity applications.

10. One or more computer systems for recommending similar applications, the systems comprising:

a database to store feature information from graphical user interface (GUI) elements of applications; and

a similar results server to provide one or more similar applications that are close in similarity score to a query application identified by a user, the similar results server including;

an extraction engine to;

execute applications in a plurality of operating environments to determine features available to the one or more similar applications rendered during application processing;

extract the feature information from the GUI elements of the one or more similar applications,

the GUI elements being rendered by the one or more similar applications during execution of the one or more similar applications; and

create feature vectors for each application, the feature vectors being based on the extracted feature information from the GUI elements; and

a similarity engine to assign similarity scores to each application based on a distance between feature vectors, each similarity score of the similarity scores indicating a similarity of the GUI elements.

11. The systems of claim 10, wherein the applications are part of an application store.

12. The systems of claim 10, wherein the applications include game applications.

13. The systems of claim 10, wherein the applications include news applications.

14. The systems of claim 10, wherein the extraction engine stores the identified features in feature vectors that are stored in the database.

15. The systems of claim 10, wherein the query application is identified by the user or the query application is identified by a recent installation on a device accessed by the user.

16. The systems of claim 10, wherein the similarity scores are based on similarity between feature vectors based on the identified features.

17. One or more computer-readable media having embodied thereon computer-executable instructions that, when executed, detect similar applications, comprising:

executing applications in a plurality of operating environments to determine features available to the applications rendered during application processing;

extracting feature information from graphical user interface (GUI) control elements of the applications, the GUI control elements being static elements and rendered by the applications during execution of the applications;

storing the feature information in an application database;

creating feature vectors for each application, the feature vectors being based on the extracted feature information from the GUI control elements; and

assigning similarity scores to each application based on a distance between feature vectors, each similarity score of the similarity scores indicating a similarity of the GUI control elements; and

returning suggestions that include results for a query application and one or more applications that are close in similarity score to the query application.

18. The media of claim 17, wherein the feature information is parsed by a screen reader.

19. The media of claim 17, wherein the feature information includes access to a video interface;

existence of button graphic, existence of menu graphics;

existence of input elements, existence of links, access to microphone, access to network interfaces, access to other applications;

battery consumption indicator, existence of description text;

number of accesses to a video interface;

number of button graphics, number of menu graphics;

number of screens, number of input elements, number of links, number of accesses to microphone, number of accesses to network interfaces, number of accesses to other applications;

percent content formatted in each media type, number of languages, number of daily launches, application version, number of levels, or number of co-purchased applications.

20. The media of claim 17, wherein at least one of the GUI control elements is a button.