Duplicate Action Request Detection Using ML Execution Queues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Unintended duplicate execution of action requests consumes significant computing and user resources, affecting memory, processor, and human resources, and can adversely impact accounts or other interests.
Innovation Solution
A machine-learning-based system that utilizes a trained model to detect duplicate action execution requests by recognizing patterns in historical action execution information, temporarily placing requests in a future execution queue, and generating outputs to remove duplicates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If action execution requests are processed without duplicate detection, then processing speed is maintained, but computing resources (memory, processor, human resources) are wasted due to unintended duplicate execution
Solution Approach 1:
The system performs preliminary duplicate detection by training a machine-learning model on historical action execution information before processing new requests. The trained model proactively identifies potential duplicate requests in the future execution queue before they are executed, preventing wasted computing resources while maintaining efficient processing of unique requests
Solution Approach 2:
The patent replaces traditional mechanical duplicate detection methods (exact matching, rule-based systems) with a machine-learning model that recognizes patterns in historical data. This substitution enables more accurate identification of duplicates including near-duplicates and variations, reducing false negatives while maintaining processing efficiency
2Measurement precision
If a machine-learning model is trained on historical action execution information, then duplicate detection accuracy is improved, but system complexity and training time increase
Solution Approach 1:
The machine-learning model is trained on historical action execution information generated by the system's own operations. The system uses its accumulated historical data to train the model, making the system self-improving without requiring external datasets or complex external training infrastructure
Solution Approach 2:
The patent introduces a future execution queue as an intermediary between request reception and execution. This queue allows the system to pause and analyze requests with the trained model before final execution, enabling accurate duplicate detection without blocking the overall processing flow
Data Source
AI summary
A computing system for determining the presence of one or more duplicate action execution requests among multiple received action execution requests is disclosed. The system can train a machine-learning model on training data including historical action execution information associated with a plurality of historical actions executed by one or more action execution applications, to generate a trained a machine-learning model. The trained machine-learning model can subsequently identify duplicate requests among newly received action execution requests by recognizing patterns in associated action execution information received with the action execution requests. An action execution request that is determined to be a duplicate action execution request by the trained a machine-learning model can then be identified for removal or automatically removed from the system.