Cross-Network Event Attribution via Selective Data Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Complex data exchanges between multiple computing systems controlled by different parties lead to conflicting views of event history and attribution, often requiring excessive data sharing, resulting in increased network utilization and latency.
Innovation Solution
An attribution server and coordination servers are configured to perform real-time, efficient attribution of online events across content networks, with the attribution server correcting duplication of conversion attribution by selecting only one event as attributable based on time stamps, while conserving network bandwidth by exchanging only necessary data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If excessive data sharing is performed to resolve attribution disagreement, then attribution accuracy is improved, but network utilization increases and latency is added
Solution Approach 1:
The patent extracts and shares only the specific event data elements necessary for attribution resolution (event timestamps, event types, and participant identifiers) rather than sharing complete datasets. This selective extraction approach resolves attribution disagreements while minimizing network bandwidth consumption and data transfer overhead.
Solution Approach 2:
The system performs partial data sharing by exchanging only the minimum necessary event information required for attribution determination. Rather than sharing all available data, the system transmits a subset of critical event parameters, achieving sufficient attribution accuracy while reducing network utilization and latency.
2Measurement precision
If excessive data sharing is performed to resolve attribution disagreement, then attribution accuracy is improved, but latency is added to data exchanges
Solution Approach 1:
The patent extracts and shares only the specific event data elements necessary for attribution resolution (event timestamps, event types, and participant identifiers) rather than sharing complete datasets. This selective extraction approach resolves attribution disagreements while minimizing network bandwidth consumption and data transfer overhead.
Solution Approach 2:
The system performs partial data sharing by exchanging only the minimum necessary event information required for attribution determination. Rather than sharing all available data, the system transmits a subset of critical event parameters, achieving sufficient attribution accuracy while reducing network utilization and latency.
3Loss of information
If multiple computing systems exchange data to determine event attribution, then attribution completeness is improved, but conflicting views of event history occur
Solution Approach 1:
The patent introduces an intermediary attribution resolution mechanism that receives event data from multiple computing systems, applies consistent attribution rules, and produces a unified attribution determination. This intermediary process reconciles conflicting event history views by establishing a standardized method for evaluating and resolving attribution disputes across distributed systems.
Solution Approach 2:
The system changes the parameters of data exchange by transforming complete event datasets into standardized attribution requests containing only essential parameters (timestamps, event types, participant IDs). This parameter transformation enables multiple systems to exchange attribution information efficiently while maintaining consistent event history views through standardized data formats and resolution rules.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
Systems and methods that may be used to provide cross content network event attribution are provided. One method includes receiving, by a coordination server, an event attribution request for a second event from an attribution server, the second event is an online activity performed by a user via one of a plurality of user devices after a first event, the first event is another online activity performed by the user via one of the plurality of devices. The method includes determining, by the coordination server, whether the second event is attributable to the first event.