<![CDATA[Denial of service (DoS) is an attack on a computer or server on an internet network that consumes computer resources until it can no longer perform its duties properly. The research objective is to develop a DoS attack detection and mitigation system based on the decision tree algorithm on server log analysis. The security method uses the decision tree algorithm because it has classification capabilities and produces simple classification tree decision rules. The system will monitor the spike of an IP in the server log to detect attacks and provide handling with IP Blocking techniques that are able to block the attacker's IP request for a certain duration. Python is used to study the data by generating a rule-based classifier then applied to the system using the PHP programming language and a separate PowerShell implementation so that it can run the system automatically. The database used is MySQL which consists of 2 tables, namely the request log table to store logs of requests that enter the server and ips throttle to store IPs that indicate attacks. The simulation results are the TPR accuracy value of 99.49% while the FPR error value is 0.14%, besides that the system successfully blocked 657 attacks but there were 135 incoming attacks and 17 normal requests were blocked. As a result, the system can predict attacks accurately and block the majority of incoming attacks although it still needs to be further optimised.]]>
Copyrights © 2025
