Automated Request Flow Log Retrieval via ML Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual processes for identifying specific request flows in resource-provisioning systems are cumbersome and time-consuming for support engineers, requiring significant training and increasing the time needed to assist users.
Innovation Solution
A method and system that extract keywords from natural language descriptions using machine learning to classify actions and identify corresponding request flows, automatically presenting relevant log entries to support engineers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual processes are used to identify request flows and log entries, then support engineers can access system information, but the process is time-consuming and requires significant training
Solution Approach 1:
The system automatically extracts keywords from natural language descriptions, classifies actions using machine learning models, and identifies request flows without human intervention. This self-service approach eliminates the need for support engineers to manually search and analyze logs, significantly reducing assistance time while maintaining high accuracy through automated keyword extraction and classification.
Solution Approach 2:
The patent replaces manual mechanical processes (support engineers manually searching logs and analyzing request flows) with an automated computer-based system that uses natural language processing and machine learning classification. This substitution transforms a labor-intensive manual process into an automated intelligent system, reducing both time consumption and training requirements.
2Ease of operation
If manual log analysis processes are used, then support engineers can retrieve system information, but the complexity and training requirements increase
Solution Approach 1:
The system introduces an intermediary automated processing layer between the support engineer and the complex log data. This intermediary automatically extracts keywords, classifies actions, and identifies request flows, shielding support engineers from system complexity while providing easy-to-use natural language query interfaces and automated results.
Solution Approach 2:
The automated system performs all complex analysis tasks independently, including keyword extraction, action classification, and request flow identification. This self-service capability eliminates the need for support engineers to understand or navigate complex system internals, making log retrieval effortless while the system handles the complexity behind the scenes.
3Productivity
If automated keyword extraction and classification is implemented, then request flow identification is automated, but machine learning model training is required
Solution Approach 1:
The system performs preliminary action by training machine learning classification models in advance using historical action data and keywords. This pre-training phase, while complex, is done once or periodically, enabling the automated high-speed request flow identification to occur without real-time training complexity. The pre-trained models are then reused for rapid classification during actual support operations.
Data Source
AI summary
Request flow log retrieval can include extracting one or more keywords from a natural language description of an action, the action being a system response to a user request submitted to a resource-provisioning system during a user session. Request flow log retrieval can also include determining a classification of the action based on a correlation value generated by a classifier model trained using machine learning to classify actions performed by the resource-provisioning system, the classification based on the one or more keywords. Additionally, request flow log retrieval can include automatically identifying a request flow associated with the action based on the classification of the action and returning at least one system log entry corresponding to the request flow.