Scholarlink

With the rapid expansion of technology and increasing reliance on web-based service, it has become essential to protect our servers from attacks like Denial of Service (DOS) and Distributed Denial of Service (DDOS). These attacks, If not fixed with time, can lead to server downtime, degraded performance of the website, and compromised services. A common method of such attacks is overwhelming the server with excessive requests from bad bots, which mimic genuine human behavior. In this context, it is important to differentiate between humans and bots. This paper proposes an effective approach for identifying humans and bots by employing honeypots, behavioral analysis, and request analysis. The keystroke and mouse movement-based model achieves accuracies of 99.98% and 99.72%, respectively, in classifying requests as human or bot.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Behavioral Biometrics: A Comparison of Keystroke Dynamics and Mouse Trajectories for Bot Detection

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Disha Modi,
  • Brijesh Bhatt,
  • Bhavika Gambhava,
  • Jatayu Baxi

摘要

With the rapid expansion of technology and increasing reliance on web-based service, it has become essential to protect our servers from attacks like Denial of Service (DOS) and Distributed Denial of Service (DDOS). These attacks, If not fixed with time, can lead to server downtime, degraded performance of the website, and compromised services. A common method of such attacks is overwhelming the server with excessive requests from bad bots, which mimic genuine human behavior. In this context, it is important to differentiate between humans and bots. This paper proposes an effective approach for identifying humans and bots by employing honeypots, behavioral analysis, and request analysis. The keystroke and mouse movement-based model achieves accuracies of 99.98% and 99.72%, respectively, in classifying requests as human or bot.