Unified Matching Engine for Cross-Channel Identity Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The phase-out of third-party cookies has hindered the ability of digital marketers to effectively track consumer interactions across devices and channels, leading to fragmented and incomplete consumer profiles, which impede personalized advertising and CRM efforts.
Innovation Solution
A unified matching engine that analyzes consumer data from various sources, including digital and non-digital interactions, to create an identity graph by associating and matching attributes, assigning confidence and combination scores, and generating an identity graph to enhance matching rates and provide a more comprehensive consumer profile.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If third-party cookies are phased out, then consumer privacy and security are improved, but the ability to track consumer interactions across devices and channels deteriorates
Solution Approach 1:
The patent introduces an intermediary matching engine that operates between data sources and consumer profiles. This engine receives data from multiple sources (web, mobile, offline), matches attributes across sources using probabilistic algorithms, and creates unified consumer profiles without relying on third-party cookies. The intermediary layer enables cross-device tracking while respecting privacy constraints.
Solution Approach 2:
The matching engine is designed to handle multiple data types and sources universally - it can process both digital data (clickstream, device identifiers) and non-digital data (offline purchase records, public records). This multi-functional approach allows the system to maintain comprehensive consumer profiling capability across all channels despite the removal of traditional tracking mechanisms.
2Quantity of substance
If consumer data is collected from multiple data sources, then the comprehensiveness of consumer profiles is improved, but the complexity of data processing and attribute matching deteriorates
Solution Approach 1:
The patent segments the data processing into distinct modules: data reception from multiple sources, attribute extraction, attribute matching, and profile generation. Each module handles specific tasks independently, making the overall complex process manageable. The segmentation allows specialized algorithms to be applied to each data type while maintaining a unified output structure.
Solution Approach 2:
The system dynamically adjusts matching parameters and algorithms based on data source characteristics and consumer context. Different probabilistic thresholds and matching criteria are applied depending on the data type and confidence levels, allowing the system to optimize performance across diverse data sources without requiring a single complex universal algorithm.
3Measurement precision
If attributes from multiple combinations are matched, then the accuracy of consumer identification is improved, but the computational resources and processing time required deteriorates
Solution Approach 1:
The system performs preliminary attribute extraction and pre-processing of data from multiple sources before the actual matching process. Attributes are standardized, validated, and organized in advance, which reduces the computational burden during the matching phase. This preliminary action enables more accurate matching without proportionally increasing processing time.
Solution Approach 2:
The patent replaces traditional mechanical/brute-force matching approaches with probabilistic algorithms and statistical models. Instead of exhaustive comparison of all attribute combinations, the system uses probabilistic inference to identify likely matches, significantly reducing computational requirements while maintaining or improving identification accuracy through statistical confidence scoring.
Data Source
AI summary
A computer-implemented method for implementing a uniform matching engine. In aspects, the method includes receiving consumer data from a plurality of data sources, the consumer data include a plurality of attributes that include personally identified information, non-personally identifiable information, or a combination thereof. In some aspects, the method associates a first attribute of the plurality of attributes with a second attribute of the plurality of attributes and creates combinations of attributes based upon the associations. In some aspects, the method includes matching first and second combinations of the multiple combinations and associating one or more attributes included in the first combination with one or more attributes included in the second combination based upon the matching. In addition, the method includes outputting the association of the one or more attributes included in the first combination with the one or more attributes included in the second combination.