Show-Specific View Prediction Using Decay Curve Regularization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video delivery services face challenges in accurately predicting future video views for shows with episodic releases, leading to mismatches in ad impressions and potential losses in contracts with advertisers.
Innovation Solution
A method is developed to determine show-specific models that account for historical records and decay curves with regularizing terms, allowing for precise prediction of future video views by modeling the decay speed and outputting these predictions to an ad system for targeted advertising.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a general decay curve model is used to predict future video views, then the prediction process is simple, but the prediction accuracy is insufficient due to ignoring show-specific viewing patterns
Solution Approach 1:
The patent segments the general prediction model into show-specific models, where each show has its own decay curve parameters (alpha and beta) that capture unique viewing patterns. This segmentation allows the system to account for variations in how different shows are consumed while maintaining the overall prediction framework.
Solution Approach 2:
The patent applies local quality by allowing different shows to have different decay curve characteristics (local parameters) rather than using a single global model. Each show's viewing behavior is modeled with its own alpha and beta parameters, enabling the system to adapt to local viewing patterns while maintaining a unified prediction approach.
2Measurement precision
If show-specific models with regularizing terms are used, then prediction accuracy improves by accounting for decay speed, but computational complexity increases
Solution Approach 1:
The patent transforms the prediction problem by changing parameters from raw view counts to decay curve parameters (alpha and beta) that directly represent viewing behavior characteristics. This parameter transformation simplifies the modeling process and improves prediction accuracy while making the computational requirements more manageable through efficient parameter estimation.
Solution Approach 2:
The patent incorporates feedback through regularizing terms in the objective function that use historical prediction errors to adjust and refine the decay curve parameters. This feedback mechanism allows the model to learn from past performance and continuously improve prediction accuracy without requiring excessive computational resources.
3Productivity
If future video views are not predicted accurately, then ad slot sales may be affected, but advertisers may reduce future contracts leading to revenue loss
Solution Approach 1:
The patent applies preliminary action by predicting future video views in advance before ad slots are sold. This allows the video delivery service to provide advertisers with reliable forecast data ahead of time, enabling informed bidding decisions and securing ad contracts before the actual viewing period begins, thus preventing revenue loss from contract cancellations.
Data Source
AI summary
In one embodiment, a method includes sending videos to users that use a video delivery service. The videos include shows that have episodes released sequentially. The method records historical records of video views for the video based on the sending of the videos to the users. For a show, a show-specific model is determined to predict future video views by performing: determining historical records of video views for different episodes of the show; training the show-specific model with the historical records, wherein the show-specific model models a decay curve with a regularizing term to regularize a decay speed; using the show-specific model to predict future video views for a future time range for episodes of the show; and outputting the future video views to an ad system configured to sell ads for the show.