Up

An Efficient Model for Web Vulnerabilities Detection based on Probabilistic Classification

 Hot
File Size:
1.41 MB
Volume:
Volume 2, Issue 8 (August, 2016)
Publication No:
IJTC201608009
Author:
Mohit Kumar, Sachin Majithia, Dr. Shashi Bhushan
Downloads:
15 x

Abstract-
The vulnerabilities in the source code can invite the hacking attempts over the software source code. The hacking attempts over the source codes are usually made to steal the data or the session in order to extract the required data from the software storage or the user profile. To build the trusted software, which are considered non hack able or with minimum vulnerabilities, the vulnerability detection before publishing the final software distribution becomes the essential affair. We will work upon four vulnerabilities i.e. Cross site scripting, SQL injection, Remote code execution and File inclusion which are four top most vulnerabilities according to OWASP’s top 10 report. In this paper, the proposed work is described to improve the performance parameters i.e. recall, false alarm rate and accuracy of existing system for detection of these four vulnerabilities. In our proposed work we will use probabilistic classification and feature descriptor for enhancing the system performance.

Keywords:
OWASP,Web Vulnerabilities, Recall rate, Feature descriptor.