Deployement of Data Analytics in Defending Web Attacks

Abstract

The approach consists of analyzing the web application source code for input validation vulnerabilities for XSS attacks. We explore the use of a combination of methods to detect this type of vulnerabilities: static analysis and data mining. Static analysis in source code, but tends to report much non-vulnerability due to its undecidability. This problem is particularly difficult with languages such as PHP that are weakly typed and not formally specified. Therefore, we complement a form of static analysis, with the use of data mining to predict the existence of false positives. To predict the existence of false positives we introduce the novel idea of assessing if the vulnerabilities detected are false positives using data mining and use a combination of data mining algorithm to find the top-ranking classifiers to flag every vulnerability as false positive or not.