Stephy Maria A K (41718013) <br/><br/>
Automated malicious web page scanner using machine learning classifiers 

 - MPhil CS 2018-2019 
<br/><br/> The number and the importance of Web applications have increased rapidly<br/>over the last years. Also, the number of web attacks are increasing day by<br/>day. According to security experts, cross- site scripting is among the most<br/>serious and common threats in Web applications today. Cross- site scripting<br/>commonly occurs in the application-layer. It is classified into three categories<br/>such as Stored XSS, Reflected XSS and DOM Based XSS. XSS vulnerabilities<br/>target scripts embedded in a page that are executed on the client-side (web<br/>browser) rather than on the server- side. Since manually preventing the XSS<br/>attacks are time-consuming, error-prone and costly, the need for automated solutions<br/>has become evident. In this work, we propose a web page vulnerability<br/>scanner using machine-learning classifier to detect and prevent XSS attacks.<br/>Also, we propose an intelligent agent based page blocking tool to block malicious<br/>scripts/page. Here we are introducing multi-agents. The intelligent agents<br/>communicate each other regarding the XSS attacks.In this work, we are using 3<br/>agents. Level1 agent act as a sensor agent.It is responsible for handling incoming<br/>request from the browser for web pages and it alerts level2 agent about a<br/>URL request.Level2 agent is responsible for decision making and evaluation .It<br/>act as a web watcher to detect attack and alarms a system about the attack.Also<br/>alert level3 agent about the attack.Level3 agent act as an interface agent If<br/>attacks happened, the intelligent agent perform appropriate actions (blocking<br/>page/script). 
<br/><br/>
<br/><br/><label>Subjects--Topical Terms: </label><br/>WEB APPLICATIONS   <br/>WEB ATTACKS<br/>XSS ATTACKS<br/>MACHINE LEARNING