Session Building from Ping-Level Data Using Parallel Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audience measurement systems face challenges in accurately determining viewing session durations from ping-level data, as they fail to account for changes in media content, leading to incorrect session classification when the time difference between pings is less than a threshold but the media content has changed.
Innovation Solution
The system determines viewing session durations by comparing time differences between pings to a threshold and detecting changes in payload data, using multiple session building procedures that operate in parallel to evaluate different partitioning schemes based on media characteristics like program names and genre identifiers, and selects the scheme that best aligns with demographic models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the system uses a simple threshold-based approach to determine session duration, then the processing speed is improved, but the accuracy of session classification deteriorates when media content changes
Solution Approach 1:
The patent segments the session building process into multiple parallel procedures, each handling different partitioning schemes (e.g., by program name, by genre). This allows the system to process session data through multiple specialized pathways simultaneously, maintaining speed while improving accuracy through comprehensive content change detection.
Solution Approach 2:
The system dynamically adapts its session building approach by selecting from multiple partitioning schemes based on the specific characteristics of the media content and ping data. This dynamic selection allows the system to optimize both processing speed and accuracy for different content types without using a single fixed threshold approach.
2Measurement precision
If the system implements multiple session building procedures to detect content changes, then the accuracy of viewing session determination is improved, but the device complexity increases
Solution Approach 1:
The complex task of detecting content changes is segmented into multiple independent session building procedures, each responsible for a specific partitioning scheme. This segmentation allows complexity to be distributed across parallel processes rather than concentrated in a single complex system, making the overall system more manageable while maintaining high accuracy.
Solution Approach 2:
The patent creates a universal session building framework that can handle multiple partitioning schemes through a common architectural pattern. Each procedure follows the same basic structure of comparing ping data against thresholds and detecting content changes, allowing the system to achieve multi-functionality without proportionally increasing complexity.
3Adaptability or versatility
If the system uses multiple partitioning schemes to evaluate different media characteristics, then the flexibility of session building is improved, but the processing time increases
Solution Approach 1:
The patent divides the session building process into separate parallel procedures that each handle specific partitioning schemes independently. This segmentation enables the system to process multiple media characteristics simultaneously rather than sequentially, maintaining flexibility while reducing overall processing time through parallel execution.
Solution Approach 2:
The system dynamically selects which partitioning schemes to apply based on the specific media characteristics and data available, rather than always processing all possible schemes. This dynamic approach allows the system to maintain high flexibility for different content types while minimizing processing time by avoiding unnecessary computations.
Data Source
AI summary
Example apparatus disclosed herein determine a first candidate duration of a session based on a first difference between a first time associated with a first start ping and a second time associated with a first end ping, the first start ping and the first end ping identified from a plurality of pings based on a first partitioning scheme. Disclosed example apparatus also determine a second candidate duration of the session based on a second difference between a third time associated with a second start ping and a fourth time associated with a second end ping, the second start ping and the second end ping identified from the plurality of pings based on a second partitioning scheme. Disclosed example apparatus further select at least one of the first candidate duration or the second candidate duration to represent the session based on a demographics model.