Cross Site Scripting (XSS) is one of the most frequently occurring vulnerability. The impact of XSS can vary from cosmetic to catastrophic damages. However, detection of XSS efficiently is still an open issue. Cross site scripting has been dealt with static and dynamic analysis previously. Both techniques have shortcomings and fail due to frequent variations in XSS payloads. Therefore, in this paper, we have proposed the use of Genetic Algorithm (GA) along with Reinforcement Learning (RL) and threat intelligence to overcome XSS attacks. For validation, the proposed approach is applied on a real dataset of XSS attacks. Results show better performance of our proposed approach when compared to the approaches reported in the literature. In addition to better performance, our method is not only flexible to changes in XSS payloads, but the results are also more understandable to end users. Moreover, our approach shows improvement when the number of attacks is increased.

跨站点脚本（XSS）是最常见的漏洞之一。XSS的影响可能会有所不同，从外观损害到灾难性损害。但是，有效地检测XSS仍然是一个未解决的问题。以前，跨站点脚本已处理过静态和动态分析。两种技术都有缺点，并且由于XSS有效负载的频繁变化而失败。因此，在本文中，我们提出了使用遗传算法（GA）以及强化学习（RL）和威胁情报来克服XSS攻击的方法。为了进行验证，将建议的方法应用于XSS攻击的真实数据集。结果表明，与文献报道的方法相比，我们提出的方法具有更好的性能。除了提供更好的性能外，我们的方法不仅可以灵活地更改XSS有效负载，但是结果对于最终用户也更容易理解。此外，当攻击次数增加时，我们的方法显示出改进。