Neural Network-based Content Inferencing Method and System

Abstract

A neural network-based content inferencing method and system automatically performs interpretive inferences of content such as video that is associated with a media instance through the application of computer-implemented neural networks. Recommended objects are generated based, at least in part, on the interpretative inferences and are delivered to users. The recommended objects may be further generated based upon inferences of preferences from usage behaviors. User behaviors associated with users interacting with the recommended objects are accessed and elements of the media instance are selected for delivery to users based on an automatic analysis of the user behaviors.

What is claimed is: 
     
         1 . A computer-implemented method, comprising:
 accessing automatically content that is associated with a media instance;   performing automatically one or more inferences that are based on analyzing the content by application of a computer-implemented neural network, wherein the neural network is a convolutional neural network;   generating automatically a plurality of recommended objects based, at least in part, upon the one or more inferences;   delivering automatically the plurality of recommended objects to one or more users;   accessing automatically one or more usage behaviors associated with at least one user of the one or more users interacting with at least one recommended object of the plurality of recommended objects;   selecting automatically an element of the media instance based on the one or more usage behaviors; and   delivering automatically the element of the media instance to a user of the one or more users.   
     
     
         2 . The method of  claim 1 , further comprising:
 accessing automatically the content that is associated with the media instance, wherein the content comprises video.   
     
     
         3 . The method of  claim 2 , further comprising:
 performing automatically the one or more inferences, wherein the one or more inferences are further based on automatically analyzing audio content that is associated with the video by application of a second computer-implemented neural network, wherein the second neural network comprises a long short-term memory neural network.   
     
     
         4 . The method of  claim 1 , further comprising:
 performing automatically the one or more inferences, wherein each of the one or more inferences is associated with a weight;   generating automatically the plurality of recommended objects based, at least in part, upon the one or more inferences and each of the associated weights.   
     
     
         5 . The method of  claim 1 , further comprising:
 performing automatically the one or more inferences, wherein the one or more inferences each comprise a theme that is inferred from the content.   
     
     
         6 . The method of  claim 1 , further comprising:
 generating automatically the plurality of recommended objects, wherein the recommended objects are further generated in accordance with an inference of a preference that is based on a plurality of usage behaviors associated with one or more users.   
     
     
         7 . The method of  claim 6 , further comprising:
 generating automatically the plurality of recommended objects, wherein the recommended objects are further generated in accordance with the inference of the preference, wherein the inference of the preference is based upon the application of a plurality of inference weightings that are determined in accordance with usage behavior priorities that are applied to the plurality of usage behaviors that are associated with the one or more users.   
     
     
         8 . A computer-implemented system comprising one or more processors configured to:
 access automatically content that is associated with a media instance;   perform automatically one or more inferences that are based on analyzing the content by application of a computer-implemented neural network wherein the computer-implemented neural network comprises a recurrent neural network;   generate automatically a plurality of recommended objects based, at least in part, upon the one or more inferences;   deliver automatically the plurality of recommended objects to one or more users;   access automatically one or more usage behaviors associated with at least one user of the one or more users interacting with at least one recommended object of the plurality of recommended objects;   select automatically an element of the media instance based on the one or more usage behaviors; and   deliver automatically the element of the media instance to a user of the one or more users.   
     
     
         9 . The computer-implemented system of  claim 8  further comprising the one or more processors configured to:
 access automatically the content that is associated with the media instance, wherein the content comprises video and associated audio; 
 perform automatically the one or more inferences that are based on analyzing the associated audio by the computer-implemented neural network. 
 
     
     
         10 . The computer-implemented system of  claim 8  further comprising the one or more processors configured to:
 perform automatically the one or more inferences, wherein the one or more inferences are based on automatically analyzing text that is contained within the content, wherein the recurrent neural network comprises a long short-term memory neural network. 
 
     
     
         11 . The computer-implemented system of  claim 8  further comprising the one or more processors configured to:
 perform automatically the one or more inferences, wherein each of the one or more inferences is associated with a weight; 
 generate automatically the plurality of recommended objects based, at least in part, upon the one or more inferences and each of the associated weights. 
 
     
     
         12 . The computer-implemented system of  claim 8  further comprising the one or more processors configured to:
 perform automatically the one or more inferences, wherein the one or more inferences each comprise a theme that is inferred from the content. 
 
     
     
         13 . The computer-implemented system of  claim 8  further comprising the one or more processors configured to:
 perform automatically the one or more inferences, wherein the one or more inferences are performed in accordance with an inference tuning control. 
 
     
     
         14 . The computer-implemented system of  claim 8  further comprising the one or more processors configured to:
 generate automatically the plurality of recommended objects, wherein the recommended objects are further generated in accordance with a computer-implemented structural aspect comprising a plurality of affinities among a plurality of computer-implemented objects, wherein the plurality of affinities are generated from a plurality of usage behaviors. 
 
     
     
         15 . A computer-implemented system comprising one or more processors configured to:
 access automatically content that is associated with a media instance;   perform automatically one or more inferences that are based on analyzing the content by application of a computer-implemented neural network, wherein the computer-implemented neural network comprises a convolutional neural network;   generate a plurality of recommended objects based, at least in part, upon the one or more inferences;   deliver automatically the plurality of recommended objects to one or more users;   access automatically a plurality of usage behaviors associated with at least one user of the one or more users, wherein the plurality of usage behaviors comprise behaviors associated with the at least one of the plurality of users navigating the media instance;   generate automatically a recommendation that is based upon an inference of a preference that is derived from the plurality of usage behaviors; and   deliver automatically the recommendation to a user.   
     
     
         16 . The computer-implemented system of  claim 15  further comprising the one or more processors configured to:
 perform automatically the one or more inferences, wherein the one or more inferences are based on automatically analyzing video by the computer-implemented neural network. 
 
     
     
         17 . The computer-implemented system of  claim 15  further comprising the one or more processors configured to:
 generate automatically the plurality of recommended objects, wherein the recommended objects are further generated in accordance with a computer-implemented structural aspect comprising a plurality of affinities among a plurality of computer-implemented objects, wherein the plurality of affinities are generated from a plurality of usage behaviors. 
 
     
     
         18 . The computer-implemented system of  claim 15  further comprising the one or more processors configured to:
 deliver automatically the plurality of recommended objects to the one or more users, wherein the plurality of recommended objects are arranged in a temporal-based sequence. 
 
     
     
         19 . The computer-implemented system of  claim 15  further comprising the one or more processors configured to:
 generate automatically the recommendation that is based upon the inference of a preference that is derived from the plurality of usage behaviors, wherein the inference of the preference is based upon the application of a plurality of inference weightings that are determined in accordance with usage behavior priorities that are applied to the plurality of usage behaviors. 
 
     
     
         20 . The computer-implemented system of  claim 15  further comprising the one or more processors configured to:
 deliver automatically an explanation for delivering the recommendation, wherein the explanation is in a natural language format and comprises reasoning for the delivery of the recommendation.