Multi-source data fused phishing website identification method and system

Abstract

The invention provides a phishing website identification method and system fusing multi-source data. The method comprises the following steps: determining phishing websites and legal website samples; collecting multi-source feature data of each website based on the website URL; processing the URL multi-source feature data to obtain a high-dimensional feature vector of the website URL multi-source feature; performing word segmentation and word segmentation operation on the website URL, and splicing the word vector matrix and the character-level vector matrix to obtain a high-dimensional feature vector of the website URL; splicing the high-dimensional feature vector of the website URL multi-source feature and the high-dimensional feature vector of the website URL to obtain a feature vector of each website; combining the feature vectors and the labels of all the websites to form a sample data set, inputting the sample data set into a classification model for training, and serving the trained classification model as a phishing website identification model; and identifying a to-be-identified website based on the phishing website identification model, and judging whether the to-be-identified website is a phishing website. The invention provides a high-precision phishing website identification scheme.

Description

technical field [0001] The invention belongs to the technical field of phishing website identification, and more specifically relates to a phishing website identification method and system that integrates multi-source data. Background technique [0002] "Phishing website" is a kind of network fraud, which means that criminals use various means to counterfeit the uniform resource locator (URL) address and page content of the real website, or use the loopholes in the real website server program to post on the website Insert dangerous HTML codes into some webpages of the website to defraud users of private information such as bank or credit card account numbers and passwords. [0003] In the existing technology, the final discrimination result can be obtained through two model trainings to efficiently realize the identification of phishing websites. However, the feature selection of this technology is relatively simple, and richer features are not considered. In addition, the...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the defects of the prior art, the purpose of the present invention is to provide a phishing website identification method and system that integrates multi-source data, aiming at solving the problems of relatively simple features and low identification performance of the existing phishing website identification selection

Images