Task Graph Pattern Mining for Variable User Action Sequences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional task mining approaches struggle to identify repetitive patterns in user interactions due to variations in how different users perform tasks, such as handling invoices through different actions in computing systems.
Innovation Solution
A system and method for identifying patterns in task execution data using a task graph and a large language model, which includes filtering sequences of actions based on start and end actions and applying similarity measures to enhance pattern recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional task mining approaches are used to capture user interactions, then task data can be collected, but the ability to identify repetitive patterns deteriorates due to variations in user actions
Solution Approach 1:
The patent introduces an intermediary representation layer (task graphs, action sequences, normalized event structures) that mediates between raw user interactions and pattern identification. This intermediary structure standardizes diverse user actions into comparable formats, enabling pattern recognition despite variability in how users perform tasks
Solution Approach 2:
The patent transforms task execution data by changing parameters such as normalizing action sequences, extracting key features, and representing tasks in standardized formats. These parameter transformations convert variable user behaviors into consistent data structures that reveal underlying patterns
2Productivity
If different users perform different actions to complete the same task, then user flexibility is maintained, but conventional task mining approaches fail to identify the underlying repetitive patterns
Solution Approach 1:
The patent segments task execution data into discrete, analyzable components such as individual actions, events, and sequence elements. By breaking down complex task variations into segmentable units, the system can identify recurring patterns across different user approaches without being overwhelmed by overall complexity
3Loss of information
If detailed user interaction data is collected to understand task performance, then comprehensive analysis is enabled, but the complexity of identifying consistent patterns increases
Solution Approach 1:
The patent extracts essential pattern-carrying information from detailed user interaction data while discarding irrelevant variations. By taking out only the critical elements that define task patterns (key actions, sequence relationships, task boundaries), the system maintains comprehensive analysis capability while reducing detection complexity
Data Source
AI summary
Systems and methods for identifying patterns of sequences of actions for performing a task from task execution data are provided. The task execution data of user interaction with a computing system for performing the task is received. A task graph is generated based on the task execution data. Patterns of sequences of actions for performing the task are identified based on the task graph. The identified patterns are output.