Cross-Media Viewership Forecasting Using Time Series Data Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for forecasting media viewership, particularly across television, websites, and consumer electronic devices, face challenges in accuracy due to the increasing number of channels and varied programming, leading to unreliable audience samples and inability to predict viewership spanning multiple media platforms effectively.
Innovation Solution
The method involves organizing viewership data from various sources into time series, comparing and cross-validating these data sets to identify patterns and eliminate untrustworthy data, then using forecasting models like neural networks or Bayesian multivariate analysis to combine and predict future viewership, accounting for missing data and different data capture intervals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sample-based estimation methods are used to forecast viewership, then the forecasting process remains simple and cost-effective, but accuracy deteriorates as the number of media channels and programming variety increase
Solution Approach 1:
The patent combines multiple data sets from different sources (television viewership data, website traffic data, consumer electronic device usage data) into a unified forecasting model. This merging of diverse data sources enables accurate cross-media viewership forecasting that traditional single-source methods cannot achieve, directly resolving the contradiction between maintaining simplicity and improving accuracy in the face of increasing media variety.
Solution Approach 2:
The forecasting system is designed to handle multiple media types (television, websites, consumer electronic devices, software applications) through a single unified model. This multi-functional approach allows the system to accurately predict viewership across diverse platforms without requiring separate forecasting methods for each medium, addressing the challenge of increasing media channel complexity.
2Reliability
If direct viewing activity data is collected from all media sources, then complete data coverage is achieved, but data reliability deteriorates due to unavailable or insufficient data for many channel-program-advertisement combinations
Solution Approach 1:
The patent introduces cross-validation as an intermediary process that mediates between multiple data sources. By comparing and validating data across different sources, the system identifies and eliminates untrustworthy data while filling gaps using information from complementary sources. This intermediary validation mechanism resolves the contradiction between achieving complete data coverage and maintaining data reliability.
Solution Approach 2:
The system transforms raw viewership data from multiple sources into standardized time series with consistent parameters and time intervals. This parameter standardization enables reliable comparison and combination of data from diverse sources, converting unreliable heterogeneous data into reliable homogeneous data suitable for accurate forecasting.
3Measurement precision
If existing television viewership prediction methods are applied to multi-media forecasting, then the forecasting approach remains straightforward, but prediction accuracy deteriorates when spanning television, websites, consumer electronic devices and software applications
Solution Approach 1:
The patent creates a universal forecasting model that adapts to multiple media types through a common data structure and validation framework. The system processes television, website, and consumer electronic device data through the same analytical pipeline, enabling accurate cross-media predictions without requiring media-specific methodologies, thus resolving the contradiction between method simplicity and multi-media adaptability.
Solution Approach 2:
The forecasting model dynamically adjusts to different media characteristics while maintaining a unified analytical approach. The system adapts its processing based on the specific properties of each data source (time intervals, data formats, measurement methods) while applying consistent validation and forecasting principles, enabling versatile multi-media forecasting without sacrificing methodological coherence.
Data Source
AI summary
Future media viewership is forecast based on time ordered analysis of historical viewership information from an individual or combination of a plurality of data sets. Forecast models having coefficients derived from comparisons of time series representations of data sets across a plurality of time periods and data sources join together disparate data sets. Individual data sets from disparate data sources may be compared to identify possible untrustworthy data or data that requires further investigation. Organizing viewership information in a time series allows for imputing missing data in a respective data set.