Cross-OS Application Similarity Detection via Icon and Metadata Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods cannot detect similar applications across different operating systems (OSs), as they are designed to analyze specific types of code on a single OS, limiting their applicability to other OSs.
Innovation Solution
An electronic device and method that compares unique IDs, icon images, and string information across different OSs to determine similarity, using standardized icon images and language-independent string analysis, and applies weighted similarity values to classify applications as identical or similar.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If code analysis methods (package name, file difference, hash function) are used to detect similar applications, then detection accuracy on identical OS is improved, but applicability to different OSs deteriorates
Solution Approach 1:
The patent applies universality by creating a detection method that works across multiple operating systems (Android, iOS, Windows, etc.). Instead of OS-specific code analysis, the invention uses universal elements like application icons, package names, and metadata that exist across different platforms, enabling the same detection algorithm to function universally across diverse OS environments.
Solution Approach 2:
The patent introduces an intermediary approach by using application icons and metadata as intermediate representation layers between different OS code structures. These intermediaries serve as common ground for comparison across OS boundaries, allowing similarity detection without directly analyzing OS-specific code formats.
2Measurement precision
If OS-specific code analysis methods are used, then detection precision on that OS is improved, but device complexity increases due to multiple analysis methods
Solution Approach 1:
The patent reduces complexity by implementing a single universal detection method that handles multiple OS types. Rather than maintaining separate analysis pipelines for each OS, the invention uses a unified approach based on icons and metadata that works across all platforms, simplifying the overall system architecture.
Solution Approach 2:
The patent extracts the essential identifying features (icons, package names, metadata) from applications, separating these universal elements from OS-specific code implementations. This extraction allows the detection system to focus on platform-independent characteristics, reducing the need for multiple OS-specific analysis methods.
3Adaptability or versatility
If application format changes are made to meet different OS platforms, then compatibility with various OSs is improved, but detection of similar applications deteriorates
Solution Approach 1:
The patent uses application icons and metadata as intermediary elements that remain consistent across different OS platforms. While the underlying code formats change to accommodate different OSs, these intermediary elements serve as stable reference points for similarity detection, allowing the system to identify similar applications despite format variations.
Solution Approach 2:
The patent leverages visual characteristics of application icons (analogous to color changes in the principle) as platform-independent identifiers. By analyzing icon images, package names, and metadata rather than code formats, the detection system can identify similar applications across OSs where the visual and textual identifiers remain consistent even when code structures differ.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method, electronic device, and system for detecting similar applications are provided. The method includes receiving an input command for detecting an application which is implemented on a certain operating system (OS) and is at least similar to a certain application; collecting detailed information regarding the certain application; searching whether an application at least similar to the certain application exists, based on the collected detailed information; and displaying the search result in a preset mode.