Machine Learning Model for Automated Exploratory Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual exploratory testing is time-consuming and expensive due to its improvisational nature, limiting its effectiveness in identifying bugs that pre-defined test cases may miss.
Innovation Solution
A machine learning model is generated using test scenarios and user interface controls to automate exploratory testing, enabling the system to probe an application's functionality and identify bugs without human intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual exploratory testing is performed by human testers, then bugs caused by unanticipated behavior can be identified, but the process becomes time-consuming and expensive
Solution Approach 1:
The patent replaces the mechanical system of manual human testing with an automated machine learning-based testing system. The ML model learns from test scenarios and UI controls to autonomously perform exploratory testing actions, substituting human testers while maintaining the ability to identify unanticipated bugs through adaptive, data-driven exploration of application behavior.
Solution Approach 2:
The testing system performs self-service by having the machine learning model automatically generate test cases, execute tests, and identify bugs without requiring continuous human intervention. The model learns from available test scenarios and UI controls, then autonomously explores application functionality, reducing dependency on manual tester involvement while maintaining effective bug detection.
2Reliability
If manual exploratory testing is performed by human testers, then improvisational testing can identify unexpected bugs, but the cost increases
Solution Approach 1:
The patent replaces expensive manual human testing with an automated machine learning system. The ML model, trained on test scenarios and UI controls, performs exploratory testing autonomously, eliminating the need to pay human testers while maintaining the ability to identify unexpected bugs through intelligent, adaptive testing behaviors.
Solution Approach 2:
The system creates a virtual copy of the testing process through the machine learning model, which replicates and extends human testing capabilities. The ML model learns from existing test scenarios and UI controls to generate its own test explorations, effectively copying and improving upon manual testing approaches at lower cost.
3Productivity
If pre-defined test cases are used, then testing efficiency improves, but bugs caused by unanticipated behavior are missed
Solution Approach 1:
The patent transforms static pre-defined test cases into dynamic, adaptive test explorations. The machine learning model learns from test scenarios and UI controls to generate test cases that adapt to the application's actual behavior, enabling the system to explore unanticipated functionality while maintaining efficient automated execution. This dynamic approach combines the efficiency of automation with the flexibility of exploratory testing.
Solution Approach 2:
The system changes the parameters of test case generation by using machine learning to dynamically adjust test explorations based on learned patterns from test scenarios and UI controls. Rather than executing fixed test cases, the ML model modifies test parameters and exploration paths in real-time, enabling discovery of unanticipated bugs while maintaining automated efficiency.
Data Source
AI summary
Technologies are provided for automated exploratory testing using machine learning. In response to receiving an identifier for an application to be tested, a machine learning model can be generated that can be used to automate exploratory testing of the application. The machine learning model can be generated based on test scenarios associated with the application and user interface controls of the application. The machine learning model can comprise one or more data structures that model relationships between user interface control values and application functionality defined by the test scenarios. The machine learning model can be used to generate exploratory testing operations targeting the application. In at least some embodiments, the machine learning model comprises an artificial neural network comprising input layer nodes associated with user interface controls and/or hidden layer nodes associated with application test scenarios.