Permutation Matching for Data Reconciliation Across Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Reconciling data between different systems is challenging due to varying taxonomies and data models, including different levels of granularity, which makes it difficult to identify discrepancies and inconsistencies.
Innovation Solution
The system employs permutation matching to reconcile data by using matching rules that include permutation keys. These keys identify subsets of data to be grouped together, allowing for the comparison of characteristics between different data sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data from different systems with varying taxonomies and data models is directly compared, then the reconciliation process becomes complex and time-consuming, but skipping grouping would miss important data relationships and increase discrepancy identification errors
Solution Approach 1:
The patent applies segmentation by dividing data into grouped subsets based on permutation keys before comparison. This groups related data elements together, reducing the complexity of direct system-to-system comparison while maintaining reconciliation accuracy. The segmentation allows the system to process data in manageable units rather than handling entire data sets simultaneously.
Solution Approach 2:
The patent introduces permutation keys as intermediary elements that mediate between different data systems with varying taxonomies. These keys serve as a common language or bridge that enables comparison across systems without requiring direct mapping of all data elements, thereby reducing reconciliation complexity while maintaining productivity.
2Measurement precision
If permutation keys are used to group data subsets, then discrepancy identification accuracy improves, but the matching rule complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-defining permutation keys and grouping data subsets before the actual discrepancy identification process. This preliminary grouping establishes a structured framework that improves measurement precision during comparison, while the grouping rules are defined once and reused, preventing exponential growth in matching rule complexity.
Solution Approach 2:
The patent uses parameter changes by transforming data into grouped subsets based on permutation keys, effectively changing the organizational parameters of the data. This transformation enables more precise discrepancy identification by comparing like-with-like data elements, while the parameter changes are systematic and rule-based rather than ad-hoc.
3Ease of operation
If data sets with different levels of granularity are compared without grouping, then the comparison process is simpler, but the ability to locate discrepancies quickly deteriorates
Solution Approach 1:
The patent applies segmentation by dividing heterogeneous data sets into grouped subsets based on permutation keys. This segmentation maintains operational simplicity by providing a systematic grouping approach, while simultaneously reducing discrepancy location time by organizing data for targeted comparison rather than exhaustive search across all data elements.
Solution Approach 2:
The patent introduces another dimension by adding the permutation key grouping layer to the data comparison process. This additional organizational dimension allows the system to maintain simplicity in the comparison logic while dramatically improving discrepancy location speed through structured data organization that enables focused comparisons.
Data Source
AI summary
A method includes obtaining first and second data sets to be reconciled and, using matching rules, identifying discrepancies between the data sets. The matching rules include at least one permutation key, where each permutation key identifies a subset of data to be grouped together in one of the data sets. Identifying the discrepancies includes attempting to match one or more first characteristics associated with the grouped subset of data in one of the data sets to one or more second characteristics associated with another of the data sets. The matching rules could involve multiple matching characteristics, and the matching rules could be generated using a metric to select the matching characteristics of the matching rules. The metric could be based on a combination of a number of matched data items and a number of matched groups of data items.