Flow Based Approximation Technique to Restrict Client Side Scripts from Phishing Attacks for Enhanced Internet Security using Fuzzy Rule Sets Malathi P.*, Dr. Vivekanandan P.** *Assistant Professor, Dhanalakshmi Srinivasan College of Engineering and Technology, Chennai, India **Professor, Department of Chemical Engineering, A. C. Tech, Anna University, Chennai, Tamilnadu, India Online published on 15 September, 2016. Abstract The most commercial activities are performed online through internet and highly depend on the internet services. The internet threats grow because of internet technology usage in many fields. There exists huge number of internet attacks and these attacks phish the information and perform various malicious activities to degrade the service provided. Most financial organizations accept user sensitive information like security codes and passwords through online and send them through a secure channel. But still these informations are stealed by malicious user to perform various attacks. The user does not know about the background process and they simply submit their details and wait for the result. The organizations have the responsibility to protect the user valuable information from malicious nodes or programs. There are many security measures present right now but lags with the classification accuracy and struggles with false positive results. To overcome all this, we propose a new approach which works based on the flow of packets and identify the malicious nodes and their packets to restrict the client side scripts. The proposed approach monitors the foreground and background processes which access the network interface. While a new process which accesses the gateway will be monitored and will be checked to find out the malicious packets. Every time a packet received, a log generated and it is called web log or traffic using which later processes will be done. The packet feature is extracted and generate log accordingly. The generated log is used to compute the flow approximation. The flow approximation which identifies distinct processes and their parent id, computes set of rules which specify the process id and process group id, process name, packet flow rate, memory, bandwidth, hop count and ttl values. Based on the log generated, the fuzzy rules are generated to perform flow approximation. The proposed approach has produced efficient results compared to existing works and has increased ratio of attack detection and reduces the time complexity. Top Keywords Internet security, Phishing Threat, Malwares, Flow approximation, Fuzzy Rule Sets, QoS. Top |