Code Score Determination Using Machine Learning Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining application reliability, resiliency, performance, and security metrics are manual and subjective, varying across businesses and lacking objectivity, which hinders consistent evaluation and comparison across different applications.
Innovation Solution
A computer-implemented method and system using machine learning to identify patterns in developed code based on user-provided score tags, applying these patterns to other code blocks, and determining score tags to generate objective, aggregate scores for applications, thereby standardizing and improving the evaluation of reliability, resiliency, performance, and security metrics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual processes are used to determine application metrics, then flexibility in evaluation can be maintained, but objectivity and consistency across different applications deteriorate
Solution Approach 1:
The patent replaces manual mechanical evaluation processes with an automated machine learning system. The system uses trained models to automatically assess code blocks against multiple metrics (reliability, resiliency, performance, security), eliminating human subjectivity while maintaining comprehensive evaluation capabilities through computational analysis.
Solution Approach 2:
The system enables self-service evaluation where the machine learning models independently assess code blocks without requiring continuous human intervention. The automated system can evaluate multiple applications simultaneously, maintaining consistent standards across all evaluations while reducing the operational burden on human engineers.
2Productivity
If manual metric determination is used, then evaluation flexibility is maintained, but evaluation speed and accuracy deteriorate
Solution Approach 1:
The patent replaces manual assessment with automated machine learning inference systems that can rapidly evaluate code blocks against multiple metrics simultaneously. The trained models process code through computational analysis, achieving high-speed evaluation while maintaining accuracy through the learned patterns from training data.
Solution Approach 2:
The system performs preliminary training of machine learning models on labeled code datasets before actual evaluation. This preliminary action enables the system to quickly and accurately assess new code blocks without requiring real-time human expertise, significantly improving evaluation speed while maintaining high accuracy through the pre-learned patterns.
3Reliability
If subjective metrics are used by engineers, then adaptability to different business needs is maintained, but consistency across different businesses deteriorates
Solution Approach 1:
The patent implements a universal machine learning system that can evaluate multiple different metrics (reliability, resiliency, performance, security) using a single integrated platform. The system maintains consistency through standardized evaluation protocols while adapting to different business needs by loading different trained models or adjusting evaluation parameters without requiring manual recalibration.
Solution Approach 2:
The system achieves adaptability through parameter changes rather than structural reconfiguration. Different business metrics can be evaluated by adjusting model parameters, weightings, and evaluation criteria within the same framework. This allows consistent evaluation methodology while maintaining flexibility to accommodate different business-specific requirements through configurable parameters.
Data Source
AI summary
Disclosed are systems and methods for determining developed code scores of an application. The method may include: receiving, by a processor, one or more first score tags of one or more first blocks of developed code from a user through a computing device; identifying, by the processor, patterns in the developed code based on the received one or more first score tags; applying, by the processor, the identified patterns to one or more second blocks of the developed code; and determining, by the processor, one or more second score tags for the one or more second blocks of the developed code based on the applied patterns.