Entity Resolution Using Relationship Scores for Master Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Master data management systems face challenges in accurately resolving entity records due to the presence of duplicate or redundant data, which often requires manual review and can lead to false positives and negatives, resulting in increased costs and reduced data quality.
Innovation Solution
A method that combines a matching score with a relationship score, generated by analyzing categorical and demographic data and additional relationship data, to determine whether entity records are duplicates or distinct, thereby automating the resolution of unresolved records and reducing the need for manual intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual review is used to resolve entity records, then data accuracy is improved, but productivity deteriorates due to increased time and cost
Solution Approach 1:
The patent introduces a relationship score as an intermediary mechanism that bridges the gap between automated matching and manual review. The relationship score analyzes connections between entity records (such as shared contacts, communication patterns, and organizational relationships) to provide additional evidence for resolution decisions, reducing the burden on manual reviewers while maintaining accuracy
Solution Approach 2:
The system changes the parameter of resolution by introducing a new scoring dimension (relationship score) alongside the traditional matching score. This creates a composite evaluation framework that transforms the resolution process from purely manual judgment to a hybrid automated-assisted process, improving both accuracy and efficiency
2Productivity
If automated matching is used to resolve entity records, then productivity is improved, but measurement precision deteriorates due to false positives and negatives
Solution Approach 1:
The relationship score serves as an intermediary validation layer that reduces false positives and negatives in automated matching. By analyzing relationship data such as shared contacts, communication frequency, and organizational hierarchy, the system provides additional confidence signals that help automated systems make more accurate decisions
Solution Approach 2:
The patent adds another dimension to the matching evaluation by incorporating relationship analysis alongside traditional attribute matching. This multi-dimensional approach (matching score + relationship score) enables the system to distinguish between true duplicates and false matches more effectively, improving precision while maintaining high productivity
3Measurement precision
If relationship data analysis is added to improve resolution accuracy, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The relationship score mechanism is designed to be universal and multi-functional, serving multiple purposes: validating matches, detecting duplicates, analyzing entity connections, and providing explanatory evidence for resolutions. This single addition performs multiple functions that would otherwise require separate systems, minimizing the increase in overall complexity
Data Source
AI summary
A first score associated with matching between entity records of a plurality of entities of master data of an MDM system is received. A set of entity records with a first score above a lower threshold score and below an upper threshold score is identified as unresolved; neither confirmed as matched or unmatched. A second score associated with relationships between entity records is generated. Overall scores for pairs of the set of entity records are determined by combining the first matching score with the second relationship score. The overall score of respective pairs of the set of entities is compared to the upper threshold, and if the upper threshold is exceeded, then the information of the pair of entity records of the set of entity records are combined into a single record, and redundant entity records are removed from the MDM system.