Inferring user intent based on network navigation paths

• US 8,775,275 B1
• Filed: 03/30/2011
• Issued: 07/08/2014
• Est. Priority Date: 03/30/2011
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Pope, Elmore Eugene
• Primary Examiner(s): Rosen, Nicholas D
• Application Number:
  US13/076,237
• Time in Patent Office: 1,196 Days
• Field of Search: 705/26.1, 705/26.7, 705/27.1, 705/14.49, 705/14.5, 705/14.51, 705/14.52, 705/14.53, 705/14.54, 705/14.55, 705/14.58, 705/14.64
• US Class Current: 705/27.1
• CPC Class Codes: G06Q 30/0251 Targeted advertisements
• CPC Class Codes: G06Q 30/0255 based on user history
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: G06Q 30/0641 Shopping interfaces

Abstract

Paths followed by a plurality of devices are recorded. Devices of the plurality have sent content requests similar to a current content request. Behaviors exhibited by respective ones of the plurality of devices. The respective ones of the paths are grouped into intent groupings. A path followed by a device is assembled. The assembling the path comprises recording a plurality of content requests generated by the device prior to the current content request. An intent grouping matching the path is identified. The intent grouping is associated with an expected behavior. Content calculated to facilitate the expected behavior is identified.

Claims

1. A system, comprising:

one or more computers configured to implement a transaction-related communication system, wherein the transaction-related communication system comprises;

a network navigation monitoring module configured to assemble a current navigation path followed by a particular user, wherein

the current navigation path comprises a description of a series of URLs requested by the particular user prior to requesting a URL of a current requested page and forming a navigation path leading the particular user to the current requested page;

an intent database configured to

record network navigation paths followed by a plurality of other users who have requested the URL of the same requested page requested by the particular user, wherein

the navigation paths comprise descriptions of respective pluralities of URLs forming navigation paths leading the respective users to the same requested page requested by the particular user,

record behaviors exhibited by respective ones of the plurality of users after requesting the URL of the requested page, and

group the respective ones of the plurality of users into intent groupings based on comparison of respective ones of the behaviors, wherein

the respective ones of the behaviors comprise requests for content; and

a content delivery module configured to

perform a comparison of the current navigation path of the particular user to the navigation paths followed by the other users to the same requested page,

based on the comparison, identify an intent grouping having navigation paths within a degree of similarity to the current navigation path for the particular user,

identify content appropriate to a group behavior associated with the intent grouping, and

communicate the content appropriate to the group behavior to the user.

2. The system of claim 1, wherein the content delivery module is further configured to insert the content into a portion of the requested page.

3. The system of claim 1, wherein the requests for content further comprise requests for electronic commerce transactions.

4. The system of claim 1, wherein the content delivery module is further configured to communicate the content in place of the requested page.

5. The system of claim 1, wherein

the navigation path further comprises a plurality of spatial locations visited by the user prior to requesting a URL of a requested page;

and

the paths followed by the plurality of users who have requested the URL of the requested page further comprise a plurality of spatial locations visited by the plurality of users who have requested the URL of the requested page.

6. A computer-implemented method, comprising:

recording navigation paths followed by a plurality of devices, wherein

devices of the plurality have sent content requests similar to a current content request from a particular device, and

the navigation paths comprise descriptions of respective pluralities of content requests forming navigation paths leading the respective devices to same requested content requested by the particular device;

recording behaviors exhibited by respective ones of the plurality of devices after sending the content requests;

grouping the respective ones of the plurality of devices into intent groupings, wherein

the grouping the respective ones of the plurality of devices into intent groupings further comprises grouping the respective ones of the plurality of devices into intent groupings based at least in part on comparison of respective ones of the behaviors;

assembling a current navigation path followed by the particular device, wherein

the assembling the current navigation path comprises recording a plurality description of a series of content requests generated by the particular device prior to the current content request and forming a navigation path leading the particular device to the current content request;

identifying an intent grouping matching the current navigation path, each intent grouping is associated with an expected behavior, and

the identifying comprises performing a comparison of the current navigation path to the navigation paths followed by the respective devices to the same requested content requested by the particular device to identify an intent grouping having navigation paths within a degree of similarity to the current navigation path for the particular device; and

identifying content calculated to facilitate the expected behavior.

7. The computer-implemented method of claim 6, wherein the grouping the respective ones of the plurality of devices into the intent groupings further comprises grouping the respective ones of the plurality of devices into the intent groupings based on a statistical measure of path similarity.

8. The computer-implemented method of claim 7, wherein the grouping the respective ones of the plurality of devices into the intent groupings based on a statistical measure of path similarity further comprises grouping the respective ones of the plurality of devices into the intent groupings based on an analysis of content similarity of sites in the paths.

9. The computer-implemented method of claim 7, wherein the grouping the respective ones of the plurality of devices into the intent groupings based on a statistical measure of path similarity further comprises grouping the respective ones of the plurality of devices into the intent groupings based on an analysis of common sites in the paths.

10. The computer-implemented method of claim 7, wherein the grouping the respective ones of the plurality of devices into the intent groupings based on a statistical measure of path similarity further comprises grouping the respective ones of the plurality of devices into the intent groupings based on an analysis of similarity of search queries in the paths.

11. The computer-implemented method of claim 7, wherein the grouping the respective ones of the plurality of devices into the intent groupings based on a statistical measure of path similarity further comprises grouping the respective ones of the plurality of devices into the intent groupings based on an analysis of similarity of timestamps in the paths.

12. The computer-implemented method of claim 6, wherein the grouping the respective ones of the plurality of devices into the intent groupings further comprises grouping the respective ones of the plurality of devices into the intent groupings based on comparison of respective ones of the behaviors.

13. The computer-implemented method of claim 12, wherein the respective ones of the behaviors comprise requests for content.

14. A non-transitory computer-readable storage medium storing program instructions that when executed cause a server to implement a content selection system, wherein the content selection system comprises:

a path recording function for

assembling a current navigation path followed by a particular device, wherein

the current navigation path comprises a description of a series of network locations requested by the particular device prior to a current content request and forming a navigation path leading the particular device to the current content request, and

recording network navigation paths followed by a plurality of other devices that have sent content requests similar to the current content request from the particular device, wherein

the network navigation paths comprise descriptions of respective pluralities of network locations forming navigation paths leading the respective other devices to a same network location as for the current content request by the particular device;

a behavior recording function recording behaviors exhibited by respective ones of the plurality of other devices after sending the content requests similar to the current content request;

a grouping function for grouping the respective ones of the plurality of other devices into intent groupings, wherein

the grouping the respective ones of the plurality of other devices into intent groupings further comprises

grouping the respective ones of the plurality of other devices into intent groupings based at least in part on comparison of respective ones of the behaviors;

an intent group identification function for identifying an intent grouping having navigation paths statistically similar to the current navigation path for the particular device; and

a content identification function for identifying content appropriate to a group behavior associated with the intent grouping.

15. The non-transitory computer-readable storage medium of claim 14, wherein the content selection system further comprises a transaction information communication function for communicating the content to the device.

16. The non-transitory computer-readable storage medium of claim 14, wherein the content identification function further comprises a recommendation generation engine for recommending products for purchase by a user of the device.

17. The non-transitory computer-readable storage medium of claim 14, wherein the content identification function further comprises a content ordering function for placing items of content for display in an order reflecting a measure of appropriateness to the group behavior associated with the intent grouping.

18. The non-transitory computer-readable storage medium of claim 14, wherein the content selection system further comprises a paths reporting function for reporting sites visited by respective ones of the plurality of devices.

19. The non-transitory computer-readable storage medium of claim 14, wherein the path recording function comprises a path monitor for recording sites visited by the device.

20. The non-transitory computer-readable storage medium of claim 19, wherein the path recording function comprises an integration function for reporting sites visited by the device in the current content request.

21. A system, comprising:

one or more computers configured to implement a content-selection system, wherein the content selection system comprises;

an intent database configured to

record network navigation paths followed by a plurality of users who have requested content similar to content that is specified in a current content request from a particular user, wherein

the network navigation paths comprise descriptions of respective pluralities of network locations forming navigation paths leading the respective users to a same network location as for the current content request by the particular user, and

record behaviors exhibited by respective ones of the plurality of users after requesting the content similar to content that is specified in a current content request, and

group the respective ones of the plurality of users into intent groupings based on comparison of respective ones of the behaviors; and

a content identification module configured to

perform a comparison of the current navigation path of the particular user to the network navigation paths to identify an intent grouping matching the current navigation path, wherein

the current navigation path comprises a description of a series of network locations requested by the particular user prior to the current content request and forming a navigation path leading the particular user to the current content request, and

identify content appropriate to a group behavior associated with the intent grouping.

22. The system of claim 21, wherein the paths are compared to the path using a calculation giving greater weight to more recent ones of the paths.

23. The system of claim 21, wherein the paths are compared to the path using a calculation giving greater weight to ones of the paths that are repeated.

24. The system of claim 21, wherein the paths are compared to the path using a calculation giving less weight to ones of the paths for which the group behavior failed to predict behavior of a previous device.

25. The system of claim 21, wherein the path comprises a series of physical locations visited by the device and the paths comprise physical locations visited by respective ones of the plurality of users.