Computing Environment Change Requests with NLP Risk Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the risk of software and hardware updates are complex, specific to programming language differences, and not applicable to non-incremental changes, leading to potential bugs and server downtime.
Innovation Solution
A computer-implemented method using natural language processing and a trained classification model to analyze text descriptions of change requests, extracting features and applying them to a machine learning model to validate the risk of updates, preventing risky changes from being implemented.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If machine learning models are used to analyze code differences for security risk detection, then security threat detection capability is improved, but device complexity and processing difficulty increase significantly
Solution Approach 1:
The patent extracts only the essential risk indicators from code changes rather than analyzing the entire codebase. The system identifies and extracts specific risk features (such as sensitive function calls, security-critical code patterns) from the code differences, feeding only these extracted features into the machine learning model. This extraction approach maintains high detection capability while significantly reducing model complexity and processing requirements.
Solution Approach 2:
The patent applies different analysis depths and model complexities to different parts of the codebase. Instead of uniformly analyzing all code changes with high complexity, the system identifies local regions of interest (such as security-critical functions, recently modified modules) and applies targeted analysis only to those areas. This local quality approach optimizes the balance between detection accuracy and processing complexity.
2Reliability
If comprehensive regression testing is performed to ensure software update safety, then reliability is improved, but loss of time and productivity decrease
Solution Approach 1:
The patent performs preliminary risk assessment and change analysis before full regression testing is initiated. The machine learning model evaluates the risk level of proposed changes in advance, and based on this preliminary assessment, the system determines the scope and depth of required regression testing. Low-risk changes undergo minimal or no regression testing, while high-risk changes trigger comprehensive testing. This preliminary action approach ensures reliability for critical changes while minimizing time loss for low-risk updates.
Solution Approach 2:
The patent applies partial regression testing based on the risk assessment results. Instead of always performing complete regression suites, the system selects and executes only the necessary subset of tests corresponding to the risk level and change scope. For low-risk changes, minimal testing is performed; for high-risk changes, more extensive testing is applied. This partial action approach optimizes the balance between ensuring software safety and minimizing testing time.
3Measurement precision
If machine learning models are trained on historical software change data to predict update outcomes, then prediction accuracy is improved for known patterns, but adaptability to new and unique change scenarios deteriorates
Solution Approach 1:
The patent designs the machine learning model with universal features and patterns that can generalize across different types of software changes. Rather than training separate models for specific programming languages or change types, the system identifies universal risk indicators and change patterns that apply across diverse scenarios. The model is trained on multi-language, multi-project historical data to learn transferable risk assessment capabilities, enabling it to handle both known patterns and new scenarios effectively.
Solution Approach 2:
The patent employs adjustable model parameters and dynamic feature weighting that can be adapted to new scenarios. The system includes mechanisms to update model parameters based on feedback from new change outcomes, allowing the model to evolve and adapt to emerging patterns. Feature importance weights are dynamically adjusted based on the specific change context, enabling the model to prioritize relevant indicators for novel scenarios while maintaining accuracy for well-known patterns.
Data Source
AI summary
There is provided a computer-implemented method for handling updates to a computing environment, the method comprising: receiving, from a user, a change request comprising a text description justifying a change in hardware and/or software within the computing environment; applying natural language processing to the text description to extract one or more text features from the text description; performing a validation check of the received change request by applying a trained classification model to the extracted one or more text features, the classification model having been trained on a dataset of historical change requests; and if the received change request fails the validation check, preventing submission of the change request into a queue of change requests and/or implementation of the change associated with the change request.