Anomaly Detection Scoring for Auditor Review
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated anomaly detection systems in various industries face inefficiencies due to high rates of false positive results, leading to wasted resources and reduced effectiveness over time, as they struggle to accurately identify anomalous activities amidst vast transaction data.
Innovation Solution
A method that calculates weighted scores for anomaly detection strategies based on effectiveness metrics such as normalized discounted cumulative gain and hit rate, prioritizing the review of potentially anomalous activities identified by more effective strategies, and continuously updates and replaces less effective strategies with new ones to optimize resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If automated anomaly detection systems analyze each individual activity, then detection accuracy is improved, but resource consumption and time cost increase significantly
Solution Approach 1:
The patent segments the anomaly detection process into two stages: (1) automated detection strategies that flag potentially anomalous activities using predefined criteria, and (2) human auditor review of only those flagged activities. This segmentation allows the system to maintain high detection accuracy while improving resource efficiency by eliminating the need for human reviewers to examine every single activity.
Solution Approach 2:
The system performs partial analysis by applying automated detection strategies to all activities to generate flags, then performs complete human review only on a subset of activities that meet the anomaly criteria. This partial action approach ensures thorough review of suspicious cases while avoiding wasteful expenditure of human resources on clearly normal activities.
2Reliability
If multiple anomaly detection strategies are used to improve detection coverage, then the number of true positives increases, but the false positive rate also increases
Solution Approach 1:
The patent implements feedback mechanisms where auditor review results are fed back into the system to refine and recalibrate the detection strategies. This feedback loop allows the system to learn from actual cases, adjust sensitivity thresholds, and reduce false positives over time while maintaining comprehensive detection coverage through multiple strategies.
Solution Approach 2:
The system dynamically adjusts detection parameters such as sensitivity thresholds and weighting factors based on performance metrics and feedback from auditor reviews. By changing these parameters, the system can optimize the balance between detection coverage and false positive rate, allowing multiple detection strategies to work effectively without being overwhelmed by false alarms.
3Measurement precision
If human auditors review all flagged activities, then false positives are reduced, but auditor time and operational costs increase
Solution Approach 1:
The patent segments the review workload by having human auditors examine only activities that are flagged by the automated detection strategies, rather than reviewing all activities. This segmentation dramatically reduces auditor time requirements while maintaining false positive reduction, as auditors focus their expertise on the small subset of potentially anomalous cases identified by the automated system.
Solution Approach 2:
The automated detection strategy acts as an intermediary that filters and pre-screens activities before they reach human auditors. This intermediary layer reduces the volume of work for auditors by eliminating clearly normal activities, allowing them to concentrate their time and expertise on cases that truly require human judgment.
4Reliability
If detection strategies are continuously updated to maintain effectiveness, then detection efficacy is improved, but system complexity increases
Solution Approach 1:
The patent manages system complexity by focusing updates on adjusting detection parameters such as sensitivity thresholds, weighting factors, and criteria thresholds rather than fundamentally redesigning the detection strategies. This parameter-based approach allows the system to adapt to changing conditions and maintain effectiveness while avoiding the complexity of complete strategy redesigns.
Data Source
AI summary
A method is disclosed for detecting anomalous activity including receiving alert statistics associated with at least one anomaly detection strategy configured to identify anomalous activity in transaction records, determining for each anomaly detection strategy at least one effectiveness metric based on the alert statistics, determining a weighted score for each anomaly detection strategy based on the at least one effectiveness metric, receiving anomalous activity data indicative of one or more activities identified as potentially anomalous by each of the anomaly detection strategies, and determining a portion of the anomalous activity data identified by each of the anomaly detection strategies to be transmitted to an auditor for review based on the weighted score. The portion of potentially anomalous activities to be transmitted is substantially proportional to the weighted score for each of the anomaly detection strategies.