Phishing Detection via Web Page Link Ratio Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for preventing phishing attacks, such as SSL secure connections and blacklists, are inadequate in detecting fake websites, leading to potential exposure of sensitive information.
Innovation Solution
A method and system that classify links in a web page into internal and external types, calculating their percentages and comparing them to a threshold to determine if a phishing attack is likely, warning users if external links exceed a certain percentage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If SSL secure connection and digital certificates are used to verify website authenticity, then the security level is improved, but it is still difficult to detect fake websites and the system complexity increases
Solution Approach 1:
The patent segments the security verification process into two layers: (1) traditional SSL/certificate verification for basic security, and (2) a new link ratio analysis mechanism that specifically targets phishing detection. This segmentation allows the system to maintain existing security infrastructure while adding a specialized anti-phishing capability without requiring complete system replacement.
Solution Approach 2:
The patent introduces an intermediary analysis mechanism that sits between the user and the website verification process. Instead of directly verifying website authenticity through complex certificate chains, the system uses link ratio analysis as an intermediary indicator to infer phishing risk, simplifying the verification process while maintaining effectiveness.
2Measurement precision
If comprehensive security verification methods are implemented to detect phishing websites, then the detection accuracy is improved, but the ease of operation for ordinary users deteriorates
Solution Approach 1:
The patent implements a self-service security mechanism where the browser automatically performs link ratio analysis without requiring user intervention. The system autonomously calculates the ratio of internal to external links and displays phishing warnings, eliminating the need for users to manually analyze website structures or understand complex security concepts.
Solution Approach 2:
The patent replaces manual security verification methods (where users would need to check certificates, analyze URLs, or verify website legitimacy) with an automated computational system that analyzes link ratios and provides instant phishing assessments, making security verification as simple as viewing a warning indicator.
3Ease of manufacture
If traditional security methods like blacklists are used to block phishing websites, then the implementation simplicity is improved, but the detection effectiveness deteriorates due to inability to detect new phishing sites
Solution Approach 1:
The patent applies preliminary action by analyzing the structural characteristics of phishing websites (link ratios) before they are added to any blacklist. By establishing a detection mechanism that evaluates link ratios, the system can identify and warn about phishing sites proactively, even before they are known to security databases, thus preventing phishing attacks on new and evolving threats.
Data Source
AI summary
A method, system and program product for preventing phishing attacks, wherein the method comprises: acquiring links in a Web page; classifying the acquired links according link types; and determining whether a phishing attack exists according to the classified links, wherein the links are classified into two types: internal links belonging to the same domain as the address of the Web page, and external links belonging to a different domain from the address of the Web page. By carrying out the method or system according to the above one or more embodiments of the present disclosure, since it is first detected whether a Web page is a fake website of a phishing attack before displaying the reproduced Web page to the user and the user is warned upon detecting a fake website, unnecessary losses due to phishing attacks can be prevented.