A malicious URL detection system and method based on automatic feature extraction

Abstract

The invention discloses a malicious URL detection system and a malicious URL detection method based on automatic feature extraction, which relates to the malicious URL detection field. The malicious URL detection system comprises a preprocessing module, a parallel learning module and a detection classification module. The preprocessing module takes the URL of the web page as an input, and convertsthe URL structural features, the web page text content and structural features, and the image features extracted by the preprocessing into three digital matrices containing feature vectors respectively. The parallel learning module uses three independent depth learning networks of different algorithms to process the three digital matrices to obtain three probability matrices. The detection and classification module inputs the three probability matrices to a fully connected network for further processing to give a final classification result. The invention combines the depth learning model ofthe text and the image with the malicious URL detection, comprehensively extracts various information of the web page, and improves the application scope and accuracy of the detection method.

Description

Technical field [0001] The invention relates to the field of malicious URL detection, in particular to a malicious URL detection system and method based on automatic feature extraction. Background technique [0002] With the rapid development of the Internet and the continuous expansion of network services, the scale of the Internet's web pages presents a development trend of "large base, fast growth, and frequent updates". Taking China as an example, as pointed out in the "Report on the Development and Security of Internet Sites in China (2017)", as of December 2016, the number of Chinese websites was 4.82 million, an annual increase of 14.1%. Abundant Internet services have greatly improved people's daily life on the one hand; on the other hand, they have also provided a broad development space for some cyber attacks (including phishing webpages, webpage Trojans, etc.). These network attacks often revolve around web pages, or design traps or exploit loopholes, and use various ...

AI Technical Summary

Problems solved by technology

Although the new technology of deep learning has been extensively researched, it is still not popularized in the actual URL detection software.

Images