Anomaly Detection Analytics System for Metric Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying contributing factors and audiences associated with metric anomalies in network user actions are complex, time-consuming, and costly, requiring administrators to run numerous reports and queries to determine the causes of anomalies, which can take days or weeks.
Innovation Solution
An analytics system that identifies anomalies by comparing current data against historical or training data, allowing users to select anomalies for further analysis, and utilizing statistical calculations and machine learning to identify contributing factors and audience segments, generating reports and visualizations to understand the anomalies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If administrators manually run reports and queries to identify contributing factors and audiences, then they can obtain detailed information about anomalies, but the process becomes extremely time-consuming and complex
Solution Approach 1:
The patent replaces the manual mechanical process of running reports and queries with an automated computer-implemented system. The anomaly detection system automatically identifies anomalies, determines contributing factors, identifies affected audiences, and generates explanations without requiring administrators to manually execute multiple reports and queries, thereby resolving the contradiction between information completeness and analysis time
Solution Approach 2:
The system performs self-service by automatically detecting anomalies and independently analyzing their causes and affected audiences. Rather than requiring administrators to manually investigate each anomaly through multiple reports, the system autonomously completes the entire analysis process, providing both comprehensive information and rapid results
2Reliability
If administrators run numerous reports and queries to identify contributing factors, then they can understand the causes of anomalies, but the complexity and cost of the process increases significantly
Solution Approach 1:
The patent merges multiple separate analytical functions into a single integrated system. Instead of requiring administrators to run separate reports for anomaly detection, factor analysis, and audience identification, the system combines these functions into one unified process that simultaneously performs all tasks, thereby maintaining high reliability while reducing system complexity
Solution Approach 2:
The anomaly detection system is designed as a multi-functional platform that can detect anomalies, identify contributing factors, determine affected audiences, and generate explanations within a single system. This universal approach eliminates the need for multiple separate reporting tools and complex query processes, resolving the contradiction between reliability and complexity
3Loss of information
If the system analyzes large quantities of data to identify contributing factors and audience segments, then it provides comprehensive insights, but the processing requirements and resource consumption increase
Solution Approach 1:
The system extracts only the essential and most relevant contributing factors and audience segments from the large quantity of analyzed data, rather than processing and presenting all available information. This extraction approach maintains data completeness for critical insights while reducing unnecessary processing resources by focusing on the most significant findings
Data Source
AI summary
The present disclosure is directed toward systems and methods for identifying contributing audience segments associated with a metric anomaly. One or more embodiments described herein identify contributing factors based on statistical analysis and machine learning. Additionally, one or more embodiments identify audience segments associated with each contributing factor. In one or more embodiments, the systems and methods provide an interactive display that enables a user to select a particular anomaly for further analysis. The interactive display also provides additional interfaces through which the user can view informational displays that illustrate the factors and segments that caused the particular anomaly and how those factors correlate with each other.