Cognitive Context Modeling for Multi-Actor Behavior Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computing systems face challenges in translating qualitative behavior criteria into numerical values, adapting to changing circumstances, and processing large amounts of data, particularly in dynamic and infinite data environments, where existing systems struggle to scale performance across a large actor population.
Innovation Solution
A cognitive modeling system employing periodic execution components, including a peer-to-peer analyzer, actor behavior analyzer, rate of change predictor, and semantic rule analyzer, to detect and adjust qualitative contexts across multiple dimensions, enabling dynamic assessment and alteration of system parameters based on changing circumstances and handling large data sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If existing analysis systems use fixed numerical thresholds for qualitative criteria, then analysis can be performed, but the system cannot adapt to changing contexts and the criteria definitions vary considerably among different contexts
Solution Approach 1:
The patent implements dynamic threshold adjustment by continuously monitoring contextual parameters (time of day, user behavior patterns, system load) and automatically adapting numerical thresholds accordingly. Instead of fixed thresholds, the system uses historical data and contextual information to dynamically determine appropriate threshold values, enabling the system to adapt to changing contexts without manual reconfiguration.
Solution Approach 2:
The system changes numerical parameters (thresholds) based on contextual conditions. Different contexts (peak usage time vs. low usage time, different user roles, different system states) trigger different parameter sets, allowing the same qualitative criteria to be evaluated appropriately across varying conditions without increasing fundamental system complexity.
2Quantity of substance
If the system processes large amounts of data from multiple actors, then comprehensive analysis is achieved, but scaling performance becomes difficult particularly when the actor population is exceedingly large or apparently infinite
Solution Approach 1:
The patent segments the large actor population into smaller groups or clusters based on similarity metrics (behavioral patterns, role, department, etc.). Each segment is processed independently through distributed analysis components, allowing the system to handle large data volumes by dividing the processing workload across multiple processors or nodes, thereby maintaining processing speed despite increasing data quantity.
Solution Approach 2:
The system implements incremental processing and sampling techniques where not all data needs to be processed in full detail simultaneously. Priority is given to processing critical or high-impact data first, while less critical data is processed asynchronously or in summary form, enabling the system to maintain productivity while handling large volumes of actor data.
Data Source
AI summary
The present design is directed to a system for detecting and adjusting qualitative contexts across multiple dimensions for multiple actors with cognitive computing techniques including a series of periodic execution components configured to operate over full or partial sets of received data, the series of periodic components comprising a peer to peer analyzer configured to detect anomalous behaviors among work-specific peer actors sharing similar types tasks, an actor behavior analyzer configured to examine change in an actor's behavior over time by comparing the similarity of past behavior and current behavior, a rate of change predictor configured to study changes in behavior over time for peer to peer performance according to the peer to peer analyzer, actor behavior change according to the actor behavior analyzer, and actor correlation analysis, and a semantic rule analyzer configured to encode conditional, provisional, cognitive, operational, and functional knowledge, and a plurality of signal managers.