Rule Graph Traversal for Record Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional rule engines face significant computational challenges when processing large rule sets, requiring numerous comparisons and failing to identify overlapping rule sets, which can lead to inaccurate rule application.
Innovation Solution
The implementation of a rule graph that represents rule sets as nodes connected by edges, allowing for traversal to determine matching rule sets and identify overlaps, thereby reducing computational requirements and enabling automatic pruning of unattainable rule sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional rule engines process large rule sets using traditional comparison methods, then complete rule evaluation is achieved, but computational resources and processing time are excessively consumed
Solution Approach 1:
The patent segments the rule set into a hierarchical tree structure where rules are organized into parent-child relationships based on their conditions and overlaps. This segmentation allows the system to evaluate rules in a structured manner from root to leaves, avoiding exhaustive pairwise comparisons while maintaining complete rule evaluation. The tree structure divides the computational task into smaller sub-tasks that can be processed more efficiently.
2Measurement precision
If traditional rule engines evaluate all rule sets without optimization, then accurate classification is maintained, but computational complexity increases significantly
Solution Approach 1:
The patent performs preliminary actions by pre-processing the rule set to identify overlapping rules and construct the hierarchical tree structure before actual data classification. During this preliminary phase, the system analyzes rule conditions, determines overlaps, and organizes rules into parent-child relationships. This preliminary organization enables faster and more accurate classification during runtime without requiring complex real-time computations.
Solution Approach 2:
The patent introduces dynamic elements by allowing the rule evaluation process to adapt based on the hierarchical structure. The system dynamically traverses the tree, selecting which rules to evaluate based on the input data characteristics and the tree structure. This dynamic approach optimizes the evaluation path, maintaining classification accuracy while reducing computational complexity compared to static exhaustive evaluation.
3Productivity
If rule engines process overlapping rule sets without identification, then all rules are applied, but rule application accuracy decreases due to conflicts
Solution Approach 1:
The patent implements feedback mechanisms by continuously monitoring rule overlaps during the tree construction and traversal process. When overlapping rules are detected, the system provides feedback to adjust the evaluation strategy, such as prioritizing certain rules over others based on their position in the hierarchy or their specificity. This feedback loop ensures that overlapping rules are handled systematically, maintaining both high rule application volume and accuracy by resolving conflicts through the hierarchical structure.
Data Source
AI summary
An approach for rule engine, filtering, and rule management is described herein. A computing device can access a rule graph that represents rule sets and includes nodes connected by edges. The computing device can receive record data associated with a record and determine at least one of the rule sets that matches the record by at least traversing the rule graph using the record data. The computing device can generate rule information based on the at least one rule set and associate the rule information with the record.