Web Page Element Identification via Probability Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing web page testing methods face challenges in identifying specific elements due to changes in attributes, leading to false positives and negatives, and incompatibility across different platforms, requiring significant developer intervention and resources.
Innovation Solution
A method that assigns probabilities to web page elements based on attribute weights, using thresholds and margins to determine the most likely specific element, with user prompting and weight updating to ensure accurate identification and action execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated programs use scripts to test web pages by simulating human operations, then testing efficiency is improved, but false positives and negatives occur due to attribute changes in web page elements
Solution Approach 1:
The patent changes the identification parameters from static attributes (ID, name, class) to dynamic visual characteristics (position, size, color, text content). This allows elements to be identified based on their visual presentation rather than mutable attributes, resolving the contradiction between automation efficiency and identification reliability when attributes change.
Solution Approach 2:
The patent replaces the mechanical attribute-matching system with a visual recognition system that uses image processing and coordinate-based identification. This substitution eliminates false positives and negatives caused by attribute changes while maintaining automated testing efficiency.
2Ease of operation
If testing methods rely on specific element attributes to identify elements, then element identification is straightforward, but compatibility across different platforms is lost
Solution Approach 1:
The patent creates a universal identification method that works across different platforms by using visual characteristics and coordinate systems that are platform-independent. The system can identify elements on various platforms (Windows, macOS, Linux, mobile devices) using the same visual recognition approach, achieving both simplicity and adaptability.
3Reliability
If developers manually intervene to identify elements when automation fails, then element identification accuracy is improved, but significant developer resources and time are required
Solution Approach 1:
The patent enables the testing system to self-correct and self-identify elements without developer intervention. When an element cannot be identified through scripts, the system automatically captures a screenshot, presents it to the developer once for manual selection, and then learns from this interaction to improve future automatic identification, significantly reducing ongoing developer time requirements.
4Reliability
If the system prompts users to identify elements when automatic identification fails, then element identification completeness is improved, but user interaction complexity increases
Solution Approach 1:
The patent introduces a screenshot as an intermediary between the automated system and the user. Instead of presenting complex technical attributes or code to the user, the system displays a visual screenshot with the uncertain element highlighted, making the interaction intuitive and simple while maintaining identification completeness.
Data Source
AI summary
A computer-implemented method, apparatus and computer program product, the method comprising: obtaining attribute weights associated with element attributes in a web page comprising elements, in regard of a specific element to be operated upon, a first margin, and a second margin; based on the attribute weights, determining a probability for each element in the web page to be the specific element; determining a first threshold indicating a difference between probabilities of two elements having the highest probabilities; determining a second threshold indicating a difference between a probability of an element having the highest probability and one; based on the first threshold, second threshold, first margin and second margin, determining whether the element having the highest probability is the specific element; and subject to the specific element being identified, performing an action upon the specific element.