Modulo 2 Bloom Filter for Privacy-Preserving Audience Deduplication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current audience measurement methods face challenges in accurately estimating unique audience sizes across multiple platforms and databases while preserving user privacy, as they rely on third-party cookies and suffer from duplicate impression counting and data sharing limitations.
Innovation Solution
The use of modulo 2 Bloom filter arrays allows for the estimation of total cardinality of users across multiple datasets without revealing individual identities, enabling accurate deduplication of audience members and providing a privacy-preserving solution for audience measurement by flipping bit values alternately between 0 and 1 based on user allocations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional Bloom filter arrays are used to track user presence, then user privacy is preserved, but accurate deduplication of audience members across multiple datasets cannot be achieved
Solution Approach 1:
The patent changes the parameter of the Bloom filter from traditional binary state (0 or 1) to modulo 2 arithmetic state (alternating between 0 and 1). This parameter change enables the filter to encode count information while maintaining privacy, allowing accurate audience measurement through cardinality estimation without revealing individual user identities. The modulo 2 approach transforms the filter's mathematical properties to support both precision and privacy goals.
Solution Approach 2:
The patent introduces a new dimension of analysis by using multiple hash functions to map users to different positions in the Bloom filter array. This dimensional approach allows the system to estimate total cardinality across datasets by analyzing the distribution and interaction of hash mappings, enabling accurate deduplication while maintaining the privacy-preserving nature of the original Bloom filter structure.
2Measurement precision
If census-wide tracking is implemented across multiple platforms, then complete audience exposure metrics can be obtained, but user privacy cannot be preserved
Solution Approach 1:
The patent introduces the modulo 2 Bloom filter as an intermediary mechanism between census-wide tracking and user privacy. This intermediary structure allows the system to collect and analyze exposure data across multiple platforms without directly exposing or storing identifiable user information. The filter acts as a privacy-preserving layer that enables complete audience measurement while protecting individual user identities through its mathematical properties and aggregation approach.
3Measurement precision
If database proprietors share detailed exposure data, then accurate cross-platform audience metrics can be calculated, but data sharing limitations and privacy concerns prevent this
Solution Approach 1:
The patent uses copying by creating simplified representations (Bloom filter arrays) of the actual user exposure data. Instead of sharing detailed exposure information, database proprietors share these compact filter copies that preserve the essential cardinality information needed for cross-platform metrics while eliminating sensitive user details. This copying approach enables data sharing capability while maintaining privacy protections and overcoming data sharing limitations.
Data Source
AI summary
Methods and systems to estimate audience measurement metrics based on users represented in Bloom filter arrays are disclosed. An apparatus includes a communications interface to receive a first Bloom filter array from a first computer of a first database proprietor. The first Bloom filter array is representative of first users who accessed media. The first users are registered with the first database proprietor. The first Bloom filter array includes a first array of first elements. Values of respective ones of the first elements are either a 0 or a 1 based on whether quantities of the first users allocated to the respective ones of the first elements are even or odd. The apparatus further includes a Bloom filter array analyzer to estimate a first cardinality for the first Bloom filter array. The first cardinality is indicative of a total number of the first users who accessed the media.