Dynamic Placement of Advertisements in a Video Streaming Platform

Abstract

Aspects of the subject disclosure may include, for example, a device that includes a processing system including a processor; and a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations such as collecting historical data on viewership of media content; gathering metadata of events in the media content; training a scoring model on the metadata and historical data, wherein the scoring model ranks placement choices and provide a score for each placement choice for advertisements within the media content; and evaluating a placement choice for ad completion rate, thereby generating an evaluation of the placement choice. Other embodiments are disclosed.

What is claimed is: 
     
         1 . A device, comprising:
 a processing system including a processor; and   a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, the operations comprising:   extracting metadata from video and audio feeds of media content;   identifying events, entities and statistics from the metadata;   determining a placement point for an advertisement within the media content in real time based on the events, entities and statistics identified; and   measuring a completion rate for the advertisement placed at the placement point.   
     
     
         2 . The device of  claim 1 , wherein the operations further comprise simulating future ad placements based on time points, intervals, ad length or a combination thereof. 
     
     
         3 . The device of  claim 1 , wherein the operations further comprise generating the metadata by analyzing the media content with a metadata extraction model. 
     
     
         4 . The device of  claim 1 , wherein the determining of the placement point is made by a model trained on the metadata. 
     
     
         5 . The device of  claim 4 , wherein the model is restricted to genre of media content. 
     
     
         6 . The device of  claim 4 , wherein historical data used to train the model is restricted by time. 
     
     
         7 . The device of  claim 6 , wherein the operations further comprise applying a differential weighting to the historical data and the metadata for training the model to smooth over gaps in the historical data and the metadata. 
     
     
         8 . The device of  claim 7 , wherein the gaps comprise missing data across time, viewer demographics, content segments, or a combination thereof. 
     
     
         9 . The device of  claim 1 , wherein the operations further comprise predicting future placement points based on the completion rate measured. 
     
     
         10 . The device of  claim 7 , wherein the processing system comprises a plurality of processors operating in a distributed computing environment. 
     
     
         11 . A non-transitory, machine-readable medium, comprising executable instructions that, when executed by a processing system including a processor, facilitate performance of operations, the operations comprising:
 extracting metadata from video and audio feeds of an audiovisual presentation;   identifying events, entities and statistics from the metadata;   determining a placement choice for an advertisement within the audiovisual presentation based on the events, entities and statistics identified; and   measuring a completion rate for the advertisement placed at the placement choice.   
     
     
         12 . The non-transitory, machine-readable medium of  claim 11 , wherein the operations further comprise selecting the placement choice for the advertisement within the audiovisual presentation based on past completion rates. 
     
     
         13 . The non-transitory, machine-readable medium of  claim 11 , wherein the placement choice is relative in time to a current viewing point of the audiovisual presentation. 
     
     
         14 . The non-transitory, machine-readable medium of  claim 11 , wherein the processing system comprises a plurality of processors operating in a distributed computing environment. 
     
     
         15 . The non-transitory, machine-readable medium of  claim 11 , wherein the operations further comprise creating a simulator model that predicts placement choices based on the completion rate measured. 
     
     
         16 . The non-transitory, machine-readable medium of  claim 15 , wherein the simulator model predicts the placement choices based on intervals and ad length. 
     
     
         17 . A method, comprising:
 extracting, by a processing system including a processor, metadata from video and audio feeds of media content;   identifying, by the processing system, events, entities and statistics from the metadata;   determining, by the processing system, in real-time a placement point for an advertisement within the media content based on the events, entities and statistics identified; and   measuring, by the processing system, a completion rate for the advertisement placed at the placement point.   
     
     
         18 . The method of  claim 17 , wherein the determining of the placement point is made by a model trained on historical metadata. 
     
     
         19 . The method of  claim 17 , further comprising: predicting, by the processing system, future placement points based on the completion rate measured. 
     
     
         20 . The method of  claim 17 , further comprising: simulating, by the processing system, future ad placements based on time points, intervals, ad length or a combination thereof.