Ranking video delivery problems

Abstract

Systems and methods for determining video infrastructure delivery problems using machine learning are presented. In an aspect, a system includes a reception component configured to receive information regarding videos streamed by the system to devices, wherein the information identifies video playback events at the devices and rebuffer events respectively associated with the video playback events. The system further includes a quality component configured to identify features related to quality of the playback events at the devices based on the information, and an analysis component configured to determine probabilities of occurrence of the re-buffer events based on different combinations of the features, and determine weighted values for each of the features that reflect their contribution to the probabilities of occurrence of the re-buffer events based on the different combinations of the features.

What is claimed: 
     
         1 . A system, comprising:
 a memory that stores computer executable instructions;   a processor that executes the computer executable instructions stored in the memory to perform operations comprising:
 receiving playback information regarding videos streamed by the system to devices, wherein the playback information identifies playback events at the devices and re-buffer events respectively associated with the playback events; 
 identifying features related to quality of the playback events at the devices based on the playback information; 
 determining for each of a plurality of different combinations of the identified features, a probability of a re-buffer event based re-buffer events associated with a respective combination; 
 determining a weighted value for each of the identified features based on the determined probabilities by applying regularization function to the determined probabilities; 
 determining a recommendation of changes to one or more features from the identified features based on the weighted value determined for each of the one or more features, the recommendation for reducing a probability of a re-buffer event; and 
 transmitting, by the system, the recommendation to a client device. 
   
     
     
         2 . The system of  claim 1 , wherein the processor further performs operations comprising analyzing effects changes to the identified features have on a probability of a re-buffer event based on the weighted values for each of the identified features. 
     
     
         3 . The system of  claim 2 , wherein analyzing the effects includes identifying a number of re-buffer events that would not have occurred based on the changes and rank the changes based on the number of the re-buffer events that would not have occurred. 
     
     
         4 . The system of  claim 2 , wherein the processor further performs operations comprising generating a report that identifies the effects the changes to the identified features have on the probability of a re-buffer event. 
     
     
         5 . The system of  claim 1 , wherein determining a recommendation of changes includes recommending one of the changes to the one or more features that results in a lowest probability of occurrence of a re-buffer event. 
     
     
         6 . The system of  claim 1 , wherein determining the weighted value for each of the identified features includes applying machine learning to the determined probabilities. 
     
     
         7 . The system of  claim 1 , wherein determining the weighted value for each of the identified features includes applying a logistic regression function to the determined probabilities. 
     
     
         8 . The system of  claim 1 , wherein the features related to the quality of the playback events include at least one of: internet protocol (IP) addresses respectively associated with each of the playback events, internet service providers (ISPs) respectively associated with each of the playback events, racks respectively associated with each of the playback events, type of the racks, protocols employed for each of the playback events, video format of the videos respectively associated with each of the playback events. 
     
     
         9 . The system of  claim 1 , wherein the features related to the quality of the playback events include at least one of type of the devices or operating system of the devices. 
     
     
         10 . A method comprising:
 receiving playback information regarding videos streamed to devices, wherein the playback information identifies playback events at the devices and re-buffer events respectively associated with the playback events;   identifying features related to quality of the playback events at the devices based on the playback information;   determining for each of a plurality of different combinations of the identified features, a probability of a re-buffer event based re-buffer events associated with a respective combination;   determining a weighted value for each of the identified features based on the determined probabilities by applying regularization function to the determined probabilities;   determining a recommendation of changes to one or more features from the identified features based on the weighted value determined for each of the one or more features, the recommendation for reducing a probability of a re-buffer event; and   transmitting the recommendation to a client device.   
     
     
         11 . The method of  claim 10 , further comprising analyzing effects changes to the identified features have on a probability of a re-buffer event based on the weighted values for each of the identified features. 
     
     
         12 . The method of  claim 11 , wherein analyzing the effects includes identifying a number of re-buffer events that would not have occurred based on the changes and rank the changes based on the number of the re-buffer events that would not have occurred. 
     
     
         13 . The method of  claim 11 , further comprising generating a report that identifies the effects the changes to the identified features have on the probability of a re-buffer event. 
     
     
         14 . The method of  claim 10 , wherein determining a recommendation of changes includes recommending one of the changes to the one or more features that results in a lowest probability of occurrence of a re-buffer event. 
     
     
         15 . The method of  claim 10 , wherein determining the weighted value for each of the identified features includes applying machine learning to the determined probabilities. 
     
     
         16 . The method of  claim 10 , wherein determining the weighted value for each of the identified features includes applying a logistic regression function to the determined probabilities. 
     
     
         17 . The method of  claim 10 , wherein the features related to the quality of the playback events include at least one of: internet protocol (IP) addresses respectively associated with each of the playback events, internet service providers (ISPs) respectively associated with each of the playback events, racks respectively associated with each of the playback events, type of the racks, protocols employed for each of the playback events, video format of the videos respectively associated with each of the playback events. 
     
     
         18 . The method of  claim 10 , wherein the features related to the quality of the playback events include at least one of type of the devices or operating system of the devices. 
     
     
         19 . A non-transitory computer-readable storage medium comprising computer-readable instructions that, in response to execution, cause a computing system to perform operations, comprising:
 receiving playback information regarding videos streamed to devices, wherein the playback information identifies playback events at the devices and re-buffer events respectively associated with the playback events;   identifying features related to quality of the playback events at the devices based on the playback information;   determining for each of a plurality of different combinations of the identified features, a probability of a re-buffer event based re-buffer events associated with a respective combination;   determining a weighted value for each of the identified features based on the determined probabilities by applying regularization function to the determined probabilities;   determining a recommendation of changes to one or more features from the identified features based on the weighted value determined for each of the one or more features, the recommendation for reducing a probability of a re-buffer event; and   transmitting the recommendation to a client device.   
     
     
         20 . The non-transitory computer-readable storage medium of  claim 19 , wherein determining the weighted value for each of the identified features includes applying a logistic regression function to the determined probabilities.