Research And Design Of Webpage Trojan Detection System

In recent decades, people’s lives have great changes with the rapid development of Internet. People are also suffering from the Internet security threats, such as malicious code, network fishing, botnets, etc. when they fully enjoy the convenience of the network. But the web page Trojan which is one kind of malicious code evolved from Internet in recent years makes prospects of Internet security not clear. There are Trojan virus, botnets, poultry and other destructive behavior along with web page Trojans. Browsing the web page has become the main way which individuals are infected with Trojan. So it is necessary to strengthen the research on web page Trojan and exploration of detection technology, and restore the Internet a safe and orderly environment.In view of this situation, this paper makes a review of web page Trojan’s definition, characteristics, methods and principles of Trojan attacks firstly, analyzes the current popular technology of web page Trojan detection in detail, and finds out the shortcomings in those technologies. Then this article proposes a detection technology based on machine-learning, finds out several kinds of typical web page Trojan characteristics through the analysis of hierarchical structure of malicious web pages and excutes experiment together with algorithm to validate those features. Finally this paper designs a web page Trojan detection system. This system detects and collects web page Trojan through crawling the web page source code of malicious URL publiced by Internet, and these codes provide support for research of network security.The main research work of this paper is as follows:1. Researching the current status of Internet security, expounding the serious harmfulness of web page Trojan, and making a review of web page Trojan, including the web page Trojan’s definition, properties and characteristics, the method and principle of web page Trojan attack, web page Trojan defense, etc.2. Introducing the current web page Trojan detection technology, analyzing the popular detection method in detail and finding out the defects of these methods. And then this article proposes a detection method based on machine-learning. This method finds out several typical characteristics through analyzing the web page Trojan’s hierarchy. Then it proves the effectiveness of these characteristics through experiment.3. Designing a web page Trojan detection system. This department defines the design purpose of the system firstly and makes a summary of the system’s requirement analysis, general design and detailed design. Finally it completes the implementation of the system with Python language.4. Testing and verificating the designed system. This paper completes the test of system function and capability through executing some test cases which is designed by itself, and then tests the detection effectiveness of this system in comparison with other antivirus softwares.The system can detect the web page Trojans effectively and collect high quality of malicious web pages fast through validation. It effectively provides the support for the research of network security.