Rule Selectivity Estimation for Fraud Detection Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for estimating validation time and selectivity of rules on data sets are inaccurate, leading to high error rates and service unavailability, particularly in fraud detection systems where quick and effective rule validation is crucial.
Innovation Solution
A method that estimates rule selectivity by using distribution steps with equal height buckets to calculate validation time for each rule, allowing for more accurate prediction of the time required to execute a rule file against historical data, thereby providing visibility into the validation process and enabling better decision-making.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional validation time estimation methods are used (multiplying sample validation time by total data amount), then the calculation is simple and quick, but the estimation accuracy is poor with high error rates
Solution Approach 1:
The patent segments the data into equal-height buckets based on distribution steps, where each bucket represents a range of values with the same frequency count. This segmentation allows for more accurate estimation by considering the actual distribution characteristics rather than treating all data uniformly. The rule validation time is then estimated by summing the products of bucket counts and corresponding validation times, providing a segmented approach that improves accuracy while maintaining computational feasibility.
2Reliability
If rule validation is performed on large datasets, then comprehensive validation results are obtained, but service unavailability and downtime increase
Solution Approach 1:
The patent performs preliminary actions by first estimating the validation time using the distribution-based method before actually executing the rule validation on the large dataset. This preliminary estimation allows users to understand the expected validation duration and make informed decisions about resource allocation and scheduling. The system can then plan validation executions during off-peak hours or allocate appropriate resources in advance, reducing the impact of service downtime while ensuring comprehensive validation results.
3Measurement precision
If histogram method with equal width buckets is used for data estimation, then the implementation is straightforward, but the error rate is high and performance is poor
Solution Approach 1:
The patent changes the key parameter from equal-width buckets to equal-height buckets based on distribution steps. Instead of dividing the data range into uniform intervals, the system creates buckets where each bucket contains the same number of records (equal height). This parameter change fundamentally improves estimation accuracy by accounting for the actual data distribution characteristics. The implementation calculates distribution steps to determine bucket boundaries, ensuring that each bucket represents an equal frequency count rather than an equal value range.
Data Source
AI summary
Some embodiments of the present invention relate to systems and methods for estimating validation time that use an estimation of the amount of data fulfilling a rule or a set of rules. This estimation method may be more accurate because it considers how much data will match the rule versus how much data can be discarded. The disclosed systems and methods give visibility into how much time a rule set validation will take and how long the computer system will be unavailable due to the rule set being evaluated. This allows users to make better decisions about whether and when a rule set should be evaluated, and/or whether and how to modify implementation of the rule set.