Anomaly Alert Accreditation Using Direction and Depth Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing anomaly detection systems generate excessive alerts without considering the direction and depth of anomalies, leading to unnecessary alerts and resource wastage, as they do not differentiate between normalizing and broadening trends, thus requiring a method to assess the significance of detected anomalies based on their direction and depth.
Innovation Solution
A method and system that detect anomalies, determine their direction (normalizing or broadening) and depth, and generate alerts only when the trend is broadening, minimizing unnecessary alerts and optimizing resource usage by focusing on significant anomalies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If anomaly alerts are generated for all detected anomalies, then comprehensive monitoring is achieved, but unnecessary alerts and resource wastage increase
Solution Approach 1:
The patent applies parameter changes by introducing direction (positive/negative) and depth parameters to characterize anomalies. By evaluating these parameters, the system determines whether to generate alerts, transforming the alert generation decision from a binary approach to a multi-parameter assessment that filters out unnecessary alerts while maintaining comprehensive monitoring of significant anomalies.
2Loss of information
If anomaly alerts are generated for all detected anomalies, then no relevant anomalies are missed, but alert inundation occurs
Solution Approach 1:
The patent applies local quality by differentiating anomalies based on their specific characteristics (direction and depth) rather than treating all anomalies uniformly. This allows the system to provide targeted alerting for anomalies with negative direction and significant depth, while excluding those with positive direction or minimal impact, thereby maintaining information completeness while improving user manageability.
3Measurement precision
If direction and depth analysis is added to anomaly detection, then alert accuracy improves, but system complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-defining the direction and depth parameters and their interpretation rules before anomaly detection. This allows the system to enhance measurement precision through structured parameter analysis without proportionally increasing complexity, as the evaluation framework is established in advance rather than computed dynamically for each anomaly.
Data Source
AI summary
The technology is generally directed to generating an alert based on a direction and depth of a detected anomaly. The direction of the anomaly may provide an indication as to whether the anomaly is broadening or normalizing, and the depth of the anomaly may provide an indication of the magnitude of the direction as compared to the detected anomaly. The direction may be positive, indicating that the anomaly is normalizing or moving towards an expected value, or negative, indicating that the anomaly is broadening or moving further from the expected value. Based on the determined direction and depth, a trend of the anomaly may be determined. A normalizing trend may indicate that the anomaly is likely to normalize and, therefore, an alert is not necessary. A broadening trend may indicate that the anomaly is likely to broaden and, therefore, an alert should be generated.