Media Impression Misattribution Correction via Matrix Inversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for monitoring media impressions, such as server logs and beaconing, are prone to over-counting and under-counting errors due to tampering and caching issues, leading to misattribution of impressions to incorrect demographics.
Innovation Solution
The proposed solution involves creating a matrix of misattribution correction factors to redistribute impressions accurately among demographics. This is achieved by collecting and analyzing impression data from both audience measurement entities and database proprietors, using techniques such as beaconing and cookie-based identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional media measurement methods (server logs, beaconing) are used to collect impression data, then media exposure metrics can be obtained, but misattribution errors occur leading to incorrect demographic information
Solution Approach 1:
The patent applies feedback by using panel member survey responses to correct and refine impression attribution data. The system continuously compares actual panel member demographics with attributed demographics and uses this feedback to adjust measurement models, thereby reducing misattribution errors and improving demographic accuracy over time.
Solution Approach 2:
The patent introduces an intermediary correction model that mediates between raw impression data and final demographic attributes. This intermediary layer processes and adjusts the data using panel member information and measurement models, serving as a buffer that transforms inaccurate raw data into more accurate demographic attributions.
2Measurement precision
If panel members are monitored to determine media exposure, then audience engagement levels can be measured, but the system becomes complex requiring coordinated data collection from multiple sources
Solution Approach 1:
The patent merges multiple data collection approaches into a unified system. It combines server log data, beaconing technology, and panel member surveys into a single coordinated measurement framework, integrating these diverse data sources through a common processing architecture that simplifies the overall system while maintaining measurement precision.
Solution Approach 2:
The patent creates a universal measurement platform that performs multiple functions: collecting impression data, verifying panel member identity, obtaining demographic information, and generating corrected attribution metrics. This multi-functional system reduces the need for separate specialized systems and simplifies operational complexity.
3Productivity
If server logs are used to track media requests, then impression counts can be obtained, but over-counting and under-counting errors occur due to tampering and caching
Solution Approach 1:
The patent introduces an intermediary verification layer that sits between server logs and final impression counts. This intermediary component uses panel member identification and survey data to validate and correct impression counts, filtering out errors caused by server log tampering, caching, and other artifacts.
Solution Approach 2:
The patent replaces reliance on mechanical server log counting with a more sophisticated system that uses beaconing technology and panel member verification. This substitution moves from simple request counting to intelligent verification based on panel member behavior patterns and demographic data, eliminating counting errors.
Data Source
AI summary
Methods, apparatus, and articles of manufacture to correct misattributions of media impressions are disclosed. An example method includes obtaining first demographic-based impressions transmitted via a network in response to access to media by a first set of individuals, obtaining, from a database, second demographic-based impressions of the media accessed by a second set of individuals, and multiplying a vector of database proprietor impression data from the database by a pseudo-inverse matrix representative of misattribution data among a first and second set of individuals to calculate corrected demographic-based impression values.