Mobile App Behavioral Model via Dynamic Crawling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Assessing the behavior of mobile applications is challenging due to their graphical user interface (GUI) and touchscreen interactions, making it difficult to determine user actions and transitions between states, which hinders development and testing processes.
Innovation Solution
A method that involves determining user actions from source code and dynamically crawling through the mobile application to generate a graphical user interface (GUI) behavioral model using a hierarchical finite state machine (HFSM), simulating user interactions to identify transitions and states, and modifying the application design for improved functionality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a GUI and touchscreen interface is used for mobile applications, then ease of operation is improved, but difficulty of detecting and measuring behavior increases
Solution Approach 1:
The patent creates a behavioral model that copies and represents the mobile application's behavior in an abstract, analyzable form. The model captures user interactions, state transitions, and application responses without requiring direct observation of the complex GUI and touchscreen interface, making behavior assessment feasible while preserving ease of operation benefits
Solution Approach 2:
The behavioral model acts as an intermediary between the complex GUI/touchscreen interface and the analysis process. Instead of directly analyzing the complex interface interactions, the model serves as a simplified representation that mediates the assessment of application behavior, maintaining ease of use while enabling effective measurement and detection
2Measurement precision
If dynamic crawling is performed to generate behavioral model, then measurement precision is improved, but productivity decreases
Solution Approach 1:
The system performs preliminary dynamic crawling during the model generation phase to capture complete behavioral information. By conducting the time-consuming exploration and state transition analysis upfront, the patent achieves high measurement precision in the behavioral model while improving productivity during subsequent development and testing activities, as the model can be reused without repeating the crawling process
Data Source
AI summary
A method of modeling behavior of a mobile application includes determining an action associated with the mobile application based on source code of the mobile application. The method further includes dynamically crawling through the mobile application based on the action and generating a graphical user interface (GUI) behavioral model based on dynamically crawling through the mobile application.