Forecasting User Impressions via Deterministic Probabilistic Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Producers of online videos and other data elements face challenges in allocating resources effectively due to slow and subjective forecasting methods, which fail to integrate varying degrees of certainty about user identity.
Innovation Solution
The method involves determining an estimate of available user impressions on a network by receiving requests with demographic limitations, distinguishing between deterministic and probabilistic users, and calculating the estimate based on query results from respective user data sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional forecasting methods are used, then the process is simple to implement, but the forecasting speed is slow and lacks objectivity
Solution Approach 1:
The system segments users into deterministic users (with confirmed identity) and probabilistic users (with inferred identity based on device fingerprints and behavior patterns). This segmentation allows the forecasting system to process different user types through separate data sets and methodologies, improving computational efficiency and forecasting speed while maintaining manageable system complexity through modular architecture
Solution Approach 2:
The system performs preliminary actions by pre-processing user data, creating device fingerprints, and establishing baseline behavior patterns before actual forecasting occurs. This pre-computation of user profiles and impression estimates enables faster real-time forecasting decisions without requiring complex real-time analysis during the actual forecasting operation
2Measurement precision
If subjective decision-making is used for data element selection, then the process requires less data processing, but the accuracy of resource allocation is reduced
Solution Approach 1:
The system enables self-service forecasting by automatically processing user data, calculating impression estimates, and generating forecasts without requiring manual analysis or subjective judgment. The automated algorithms process deterministic and probabilistic user data independently, producing accurate forecasts efficiently and eliminating the time-consuming manual review process while maintaining high measurement precision
Solution Approach 2:
The system changes the parameters used for forecasting by transitioning from subjective evaluation metrics to objective quantitative parameters such as user impression counts, engagement rates, and conversion probabilities. This parameter transformation enables precise measurement of forecasting accuracy while reducing data processing time through automated computational methods rather than manual analysis
3Quantity of substance
If only deterministic users are counted, then the user base is smaller and data processing is faster, but the completeness of impression forecasting is reduced
Solution Approach 1:
The system segments the user base into deterministic users (with confirmed identity through login credentials or account data) and probabilistic users (with inferred identity through device fingerprints, browser characteristics, and behavior patterns). This segmentation allows the system to process both user types efficiently using optimized algorithms for each segment, increasing the total user base size while maintaining fast data processing through parallel computation and selective querying
Solution Approach 2:
The system applies partial action by processing probabilistic user data with appropriate uncertainty weighting rather than treating all users equally. This allows the system to include a larger user base while managing computational resources effectively, processing only the necessary probabilistic user attributes required for the specific forecast rather than all possible user data, thus balancing quantity with processing time
Data Source
AI summary
Systems and methods are disclosed for determining an estimate of available user impressions on a network, comprising receiving a request for an estimate of available user impressions for viewing one or more media elements on a network, the request comprising one or more viewer demographic group limitations. A request may be received to include deterministic users and probabilistic users in the estimate of available user impressions. A number of deterministic users may be determined based on query results from a deterministic user data set. A number of probabilistic users may be determined based on query results from a probabilistic user data set, and the estimate of available user impressions may be determined based on the number of deterministic users and the number of probabilistic users.