Performance-based content delivery

• US 10,027,739 B1
• Filed: 12/16/2014
• Issued: 07/17/2018
• Est. Priority Date: 12/16/2014
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Krishnan, Prasanth Krishnan Krishnasamy Navaneetha, Srinivasan, Vengadanathan, Tibrewal, Saharsh, Pandey, Rajeev Kumar
• Primary Examiner(s): Chea, Philip J
• Assistant Examiner(s): Chen, Wuji
• Application Number:
  US14/572,710
• Time in Patent Office: 1,309 Days
• CPC Class Codes: G06F 11/00 Error detection; Error correction; Monitoring error detection, correction or monitoring in information storage based on relative movement between record carrier and transducer G11B20/18; monitoring, i.e. supervising the progress of recording or reproducing G11B27/36; in static stores G11C29/00
• CPC Class Codes: G06F 11/3006 where the computing system is distributed, e.g. networked systems, clusters, multiprocessor systems multiprogramming arrangements G06F9/46; allocation of resources G06F9/50
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 16/906 Clustering; Classification
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: H04L 43/0852 Delays
• CPC Class Codes: H04L 67/60 Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources admission control or resource allocation H04L47/70

Abstract

Systems and methods for performance-based content delivery are disclosed. A performance management service can define client performance categories based on performance data regarding content requesting, delivery and rendering, and thereby enable content providers to generate or update content based on characteristics of different performance categories in order to improve user experience. The performance management service may also predict performance categories for clients with respect to their currently submitted content requests based on applicable client classification criteria. The performance management service can provide the category prediction to content providers so that a version of the requested content appropriate for the predicted category is transmitted to the client.

Claims

1. A computer-implemented method for facilitating content delivery based on performance categories, the computer-implemented method comprising:

under control of a hardware computing device configured with specific computer executable instructions,

obtaining performance data from a plurality of client computing devices, wherein the performance data includes latency information regarding processing of requests for network resources submitted by individual client computing devices of the plurality of client computing devices over a specified period of time;

obtaining a definition of a first performance category in a plurality of performance categories, wherein the definition of the first performance category corresponds to a percentile cutoff associated with a list of the plurality of client computing devices that are sorted based on the performance data;

determining, based at least in part on the definition of the first performance category, at least one criterion that represents a threshold value that defines when a target client computing device in the plurality of client computing devices belongs to the first performance category when a request to a content provider for a target network resource is submitted by the target client computing device;

receiving an indication that a first client computing device transmitted the request for the target network resource to the content provider; and

in response to the received indication;

generating a prediction that the first client computing device belongs to the first performance category based, at least in part, on the at least one criterion;

transmitting the prediction to the content provider; and

causing the content provider to transmit a version of the target network resource to the first client computing device based, at least in part, on the prediction that the first client computing device belongs to the first target performance category.

2. The computer-implemented method of claim 1, wherein the definition of the first performance category is generated based on the performance data.

3. The computer-implemented method of claim 1, wherein the at least one criterion includes latency requirements associated with a specified number of most recent requests for network resources transmitted by the target client computing device.

4. The computer-implemented method of claim 1, wherein the first computing device is not a member of the plurality of client computing devices.

5. The computer-implemented method of claim 1, wherein the version of the target network resource transmitted to the first client computing device is generated based, at least in part, on a performance characteristic of the first performance category.

6. A computer-implemented method comprising:

under control of a hardware computing device configured with specific computer executable instructions,

obtaining a definition of a first performance category in a plurality of performance categories for a plurality of client computing devices, wherein the definition of the first performance category corresponds to a percentile cutoff associated with a list of the plurality of client computing devices that are sorted based on performance data;

determining, based at least in part on the definition of the first performance category, at least one criterion for determining that a target client computing device belongs to the first performance category when a request for a target network resource is submitted by the target client computing device; and

in response to receiving an indication that a first client computing device transmitted the request for the target network resource;

generating a prediction that the first client computing device belongs to the first performance category based, at least in part, on the at least one criterion, wherein a content provider transmits one of a plurality of versions of the target network resource to the first client computing device based, at least in part, on the prediction that the first client computing device belongs to the first target performance category.

7. The computer-implemented method of claim 6, wherein generating the prediction comprises analyzing at least one portion of the performance data associated with the first client computing device in accordance with the at least one criterion.

8. The computer-implemented method of claim 7, wherein generating the prediction further comprises obtaining the at least one portion of the performance data from at least one of the first client computing device or a second content provider that fulfilled at least one network resource request submitted by the first client computing device.

9. The computer-implemented method of claim 6, wherein the at least one criterion is related to performance category information associated with a specified number of most recent requests for network resources transmitted by the target computing device.

10. The computer-implemented method of claim 9, wherein the specified number of most recent requests for network resources are fulfilled by the content provider.

11. A system comprising:

a data store configured to at least store computer-executable instructions; and

a hardware processor in communication with the data store, the hardware processor configured to execute the computer-executable instructions to at least;

obtain a definition of a first performance category in a plurality of performance categories for a plurality of client computing devices, wherein the definition of the first performance category corresponds to a percentile cutoff associated with a list of the plurality of client computing devices that are sorted based on performance data;

determine, based at least in part on the definition of the first performance category, at least one criterion for determining that a target client computing device belongs to the first performance category when a request for a target network resource is submitted by the target client computing device; and

in response to receiving an indication that a first client computing device transmitted the request for the target network resource;

generate a prediction that the first client computing device belongs to the first performance category based, at least in part, on the at least one criterion, wherein a content provider transmits one of a plurality of versions of the target network resource to the first client computing device based, at least in part, on the prediction that the first client computing device belongs to the first target performance category.

12. The system of claim 11, wherein the at least one criterion is related to the performance data associated with the target client computing device and wherein the performance data comprises at least one of process information, memory information, network data, resource data, or client computing component information.

13. The system of claim 11, wherein the indication that the first client computing device transmitted the request for the target network resource is received from at least one of the first client computing device or the content provider.

14. The system of claim 11, wherein the hardware processor is further configured to execute the computer-executable instructions to at least cause the content provider to select the one of the plurality of versions of the target network resource in accordance with the first performance category.

15. The system of claim 11, wherein the hardware processor is further configured to execute the computer-executable instructions to at least generate the prediction based, at least in part, on performance data associated with a second client computing device.

16. A non-transitory computer readable storage medium storing computer executable instructions that when executed by a processor perform operations comprising:

obtaining a definition of a first performance category in a plurality of performance categories for a plurality of client computing devices, wherein the definition of the first performance category corresponds to a percentile cutoff associated with a list of the plurality of client computing devices that are sorted based on the performance data;

determining, based at least in part on the definition of the first performance category, at least one criterion for determining that a target client computing device belongs to the first performance category when a request for a target network resource is submitted by the target client computing device; and

in response to receiving an indication that a first client computing device transmitted the request for the target network resource;

generating a first prediction that the first client computing device belongs to the first performance category based, at least in part, on the at least one criterion, wherein a content provider transmits one of a plurality of versions of the target network resource to the first client computing device based, at least in part, on the first prediction that the first client computing device belongs to the first target performance category.

17. The non-transitory computer-readable storage medium of claim 16, wherein the operations further comprise generating a second prediction that the first client computing device belongs to a second performance category of the plurality of performance categories.

18. The non-transitory computer-readable storage medium of claim 17, wherein the content provider transmits a first version in the plurality of versions of the target network resource based further on the second prediction.

19. The non-transitory computer-readable storage medium of claim 17, wherein the first and second predictions each are associated with a respective confidence level.

20. The non-transitory computer-readable storage medium of claim 19, wherein the operations further comprise transmitting the first prediction to the content provider based, at least in part, on confidence levels associated with the first and second predictions.