Recovering information from RIAs users access logs

Web应用程序，特别是使用诸如JavaScript、AJAX、HTML5等先进技术的“富Internet应用程序”（ria），在现代系统中所占的比例越来越大。如今，我们与计算机的大多数交互都涉及到Web和Web应用程序。随着ria变得越来越适合移动设备，这种趋势还会继续发展。尽管Web应用程序和ria有很多优点，但它们也受到安全漏洞的困扰，这是一个需要尽快解决的问题。像IBM这样的公司已经开发了一系列以使Web应用程序更安全为中心的产品和服务，并且现在正在将他们的重点转向ria。然而，这些工具和服务目前在有效处理ria方面的能力有限。在本研究中，我们的目标是通过两种方式为ria安全性做出贡献：首先，我们将挖掘应用程序的现有使用（HTTP）日志，以推断有关应用程序的一系列技术信息，然后使用该信息自动配置安全扫描器。这将允许自动检测ria中的安全漏洞。第二，仍然使用HTTP日志，但是来自单个用户，我们将自动重建该用户会话，在安全事件报告后，取证调查人员可以使用该重建；从而大大提高了调查的准确性，同时大大减少了执行调查所需的时间。

Web applications, and in particular "Rich Internet Applications" (RIAs), which use advanced technologies such as JavaScript, AJAX, HTML5 etc., represent an increasingly large proportion of modern systems. Nowadays, most of the interactions we have with computers involve Web and Web applications. This tendency will increase moving forward as RIAs become increasingly mobile-friendly.For all their advantages, Web applications and RIAs are also notoriously plagued with security vulnerabilities, a problem that needs to be resolved as soon as possible. Companies such as IBM have developed a series of products and services centered on making Web applications more secure, and are now shifting their focus toward RIAs. However, these tools and services are currently limited in their ability to handle RIAs efficiently.In this research, we aim to contribute to RIAs security in two ways: first, we will mine existing usage (HTTP) logs of the application to infer a series of technical information about the application, and then use that information to automatically configure security scanners. This will allow the automatic detection of security vulnerabilities in RIAs. Second, still using HTTP logs but from a single user, we will automatically reconstruct that user session, a reconstruction that can be used by forensics investigators after a security incident has been reported; thus dramatically improving the accuracy of the investigation while significantly cutting down the time required to perform the investigation.