Irregular Event Detection in Push Notification Analytics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The transition from pull reporting to push notifications in automated analytics systems poses challenges in automation and integration, particularly in managing and reviewing frequent, uneventful results, leading to inefficiencies in detecting irregular events.
Innovation Solution
A method and system for detecting irregular events involve extracting values from push notifications, calculating distributions, comparing them to established norms, and generating alerts when irregularities are detected, utilizing a measures database, distribution analyzer, and alerter to notify users of significant deviations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If push notifications are sent frequently to report analytics results, then the automation and integration of reporting is improved, but the efficiency of detecting irregular events deteriorates because users do not review frequent normal results
Solution Approach 1:
The patent extracts only the irregular events from the stream of push notifications and sends them as separate alerts to users. This separates the normal reporting function from the alerting function, allowing users to receive comprehensive reports automatically while only paying attention to exceptional cases. The system extracts irregular events by comparing notification values against historical distributions and threshold criteria.
Solution Approach 2:
The patent segments the notification content into two types: regular analytics reports (sent periodically) and irregular event alerts (sent only when anomalies are detected). This segmentation allows the system to maintain frequent automated reporting while reducing user burden by separating routine information from actionable alerts that require immediate attention.
2Loss of information
If users review all pushed results, then the completeness of information review is improved, but the time and effort required deteriorates
Solution Approach 1:
The system implements feedback by automatically comparing current notification values against historical distributions and user-defined thresholds. This feedback mechanism identifies irregular events without requiring user review of all data, allowing users to receive summarized insights about exceptional cases while maintaining awareness of overall system performance.
Solution Approach 2:
The patent performs preliminary analysis by calculating statistical distributions from historical notification data and pre-establishing threshold criteria before actual review occurs. This preliminary action enables the system to automatically identify and filter irregular events, so users receive only the information that requires their attention rather than reviewing all historical data manually.
3Loss of information
If the system sends detailed analytics reports, then the information provided is improved, but the number of notifications and potential spam perception increases
Solution Approach 1:
The patent extracts only the most critical information (irregular events) from detailed analytics reports and presents them as concise alerts. This extraction approach maintains comprehensive information availability through the full report system while reducing the immediate impact of frequent notifications by highlighting only exceptional cases that require user action.
Solution Approach 2:
The patent applies local quality by customizing the notification content based on the specific context and user preferences. Instead of sending uniform detailed reports, the system tailors alerts to local conditions (specific irregular events detected in particular dimensions) while maintaining overall report comprehensiveness through the full analytics system.
Data Source
AI summary
Systems and methods of detecting irregular events include the extraction of values for measure in each of a plurality of notifications. The extracted values are stored in a measures database and a distribution is calculated for the values of each of the measures. The extracted values are compared to the calculated distributions to determine if an irregular event has occurred. An irregularity alert is produced if an irregular event has occurred.