Call Stack Scoring for Application Vulnerability Prioritization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing software applications face challenges in identifying and mitigating security vulnerabilities, particularly through stack traces, which can expose them to exploitation and performance issues.

Innovation Solution

A system and method for evaluating call stacks using machine learning algorithms to determine the likelihood of vulnerabilities, providing a score for vulnerability severity, and generating notifications for remediation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If machine learning algorithms are used to evaluate call stacks and generate vulnerability scores, then application security and vulnerability identification capability are improved, but system complexity and computational resources required are increased

Engineering Contradiction:

Improveapplication securityVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces machine learning models as intermediary components between call stack analysis and vulnerability identification. These models serve as mediators that process call stack data and generate vulnerability scores, enabling automated security assessment without requiring complex manual analysis systems.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements self-service through automated vulnerability detection and scoring using machine learning algorithms. The call stack analysis system automatically evaluates security vulnerabilities without human intervention, generating scores and prioritizing issues autonomously based on trained model predictions.

Inventive Principle:

Principle #25Self-service

2Measurement precision

If machine learning models are trained using known vulnerabilities and call stacks, then vulnerability detection accuracy is improved, but training time and data processing requirements are increased

Engineering Contradiction:

Improvevulnerability detection accuracyVSAvoidtraining time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies preliminary action by training machine learning models in advance using historical vulnerability data and call stacks before deployment. This pre-training phase prepares the models for rapid vulnerability detection during actual operation, reducing real-time analysis delays while maintaining high accuracy through prior learning from labeled datasets.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12579276B2Application vulnerability score based on stack traces

Publication Date: 2026.03.17 CISCO TECHNOLOGY INC

AI summary

A computing system for identifying and scoring problems associated with call stacks. The computing system identifies call stacks associated with an application and determines a problem occurs in the application. The computer system compares a call stack of a first set of applications with a call stack of a second set of applications, wherein the call stack of the first set of applications includes the problem and the call stack of the second set of applications does not include the problem. The computer system generates a score indicating a likelihood that a particular call stack caused the problem based on whether the particular call stack is included in the call stack of the first set of applications, the call stack of the second set of applications, or both. The computing system generates a notification comprising the score indicating the likelihood that the particular call stack caused the problem.