当前位置: X-MOL 学术Comput. Secur. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
New Biostatistics Features for Detecting Web Bot Activity on Web Applications
Computers & Security ( IF 4.8 ) Pub Date : 2020-10-01 , DOI: 10.1016/j.cose.2020.102001
Rizwan Ur Rahman , Deepak Singh Tomar

Abstract Web bots are malicious scripts that automatically traverse the websites, fill the web form and illegally scrap the data from web sites. The never-ending threat of web bot is causing serious problems on the web applications. According to various web bot traffic reports, more than fifty percent of the total web traffic is coming from web bots. An effective safeguard against automated web bots is to detect the human user presence on the web applications. Most part of the existing research is focused on specific web bot detection such as form spamming bot, data scrapping bots, chat bots, and game bots. In this paper, the web bot detection model is proposed using combined supervised and unsupervised machine learning algorithms. In this paper, new Biostatistics features are proposed which is used to identify the human user presence on web applications. The Biostatistics features have proven very effective in discriminating human users from general web bots. Various attack scenarios are created for web bot attacks such as automated account registration, automatic form filling, and data scrapping to mimic the zero-day web bot attacks. The proposed model is evaluated by numerous experiments using standard evaluation parameters. The result analysis reveals that the proposed model is efficient in discriminating human users from web bots.

中文翻译:

用于检测 Web 应用程序上的 Web Bot 活动的新生物统计功能

摘要 Web bots 是一种恶意脚本,可以自动遍历网站、填写网络表单并从网站上非法抓取数据。Web bot 永无止境的威胁正在导致 Web 应用程序出现严重问题。根据各种网络机器人流量报告,总网络流量的 50% 以上来自网络机器人。针对自动化 Web 机器人的有效防护措施是检测 Web 应用程序上是否存在人类用户。大部分现有研究都集中在特定的网络机器人检测上,例如表单垃圾邮件机器人、数据抓取机器人、聊天机器人和游戏机器人。在本文中,使用监督和无监督机器学习算法相结合,提出了网络机器人检测模型。在本文中,提出了新的生物统计功能,用于识别 Web 应用程序上的人类用户存在。事实证明,生物统计功能非常有效地将人类用户与一般网络机器人区分开来。为网络机器人攻击创建了各种攻击场景,例如自动帐户注册、自动表单填写和数据抓取,以模拟零日网络机器人攻击。所提出的模型是通过使用标准评估参数的大量实验来评估的。结果分析表明,所提出的模型在区分人类用户和网络机器人方面是有效的。和数据抓取来模拟零日网络机器人攻击。所提出的模型是通过使用标准评估参数的大量实验来评估的。结果分析表明,所提出的模型在区分人类用户和网络机器人方面是有效的。和数据抓取来模拟零日网络机器人攻击。所提出的模型是通过使用标准评估参数的大量实验来评估的。结果分析表明,所提出的模型在区分人类用户和网络机器人方面是有效的。
更新日期:2020-10-01
down
wechat
bug