Cross site scripting attacks detection algorithm based on the appearance position of characters

Recently, attacks against web application, such as SQL injection and cross site scripting, tend to increase. The detection methods against web application attacks such as pattern recognition, parsing and listing methods (web application firewall) have been developed. However, the evasion techniques of web application attacks are developing still now. In this paper, we proposed a new detection algorithm against cross site scripting attacks by extracting an attack feature of cross site scripting attacks considering the appearance position and frequency of symbols. Our proposed algorithm learns the attack features from given attack samples. We prepared samples for learning and testing, to show the effectiveness of our proposed algorithm. As the result, our proposed detection method was successfully detected 99.5% attack test samples and 97.5% normal test samples.