Research On Identification System Of Fishing Website Based On QR Code

QR(Quick Response)code is a kind of matrix type 2D barcode.With the rapid development of society,it has been widely used in people's daily activities.Scanning QR codes has become a daily "work" that people must do.However,it not only brings convenience to people,but also malicious attacks such as phishing websites and virus software.With the change of people's focus from the PC to more portable smart terminals,phishing attacks are also transformed from the traditional e-mail spam lure phishing attacks to the way of scanning QR codes to link directly to phishing websites.However,existing APPs that can perform QR code scanning,such as WeChat and Alipay,cannot identify phishing websites.Although 360 mobile phone guards,QQ mobile phone butlers and other APPs can identify phishing websites,they cannot be screened based on front-end QR codes.This paper analyzes and extracts the abnormal characteristics of phishing websites,and combines with the URL blacklist verification technology to design a recognition system that can be used on Android smart terminals to face QR code phishing websites.The main work of this article is as follows:1.Research on the status quo of QR code defense technology.Based on the characteristics of QR code,analyzed the mechanism and the reasons why QR code becomes the carrier of malicious attack.Combined with ZXing open source library to achieve QR code scanning and decoding capabilities.2.The blacklist of phishing websites published by well-known authoritative anti-phishing websites such as PhishTank and security alliances is used to establish a blacklist database of local phishing websites to implement an initial security assessment of URL links.At the same time,through the research of related papers and the research on phishing websites,select whether the URL link itself and the WEB page content have the features of the phishing website as a basis for follow-up URL security assessment.Separately extract abnormal features of these two parts,and build their own eigenvectors.For the URL link anomaly feature detection part,calculated the weight of each feature vector,and the Logit model is used as an identification algorithm to evaluate the URL security.For the WEB page anomaly feature detection part,security evaluation is performed according to the weight and size of each feature vector.Finally,the security of the URL that the user wants to visit is determined by the combination of the blacklist matching method and the abnormal feature vector discrimination method.3.Using Android integrated development tools combined with PC to set up the test environment,test the phishing website identification system proposed in the paper,and prove the effectiveness and feasibility of the phishing website identification system based on QR code that this article proposed.