Optimized Outlier Based Web Bot Detection

Abstract

By the turn of century, the use of computers and accessing internet were rapidly increases. As the increasing the network access it increases the network attacks also. The nature of attacks may vary in each day. Todays trends of attacks are web bots. Web bots can be used for both useful and destructive purposes. Now a days attackers use bot nets for malicious intents. Bots are basically a computer program that surf multiple websites without the intention of the user to perform variety of tasks. If any web bots were present in network it may distort the analysis process which leads to incorrect pattern and cause wrong decision making. The web bots requests were different from genuine request. So it can consider web bots are example of outliers and detect them using outlier detection methods. In this project use Swarm Intelligent (SI) based technique called Particle Swarm Optimization technique (PSO) for detect outliers or web bots. The efficiency of PSO algorithm depends on its parameters. For improving the efficiency of PSO algorithm it need some changes in its parameters. So for improving the efficiency of outlier detection optimization based HPSO (Hierarchical Particle Swarm Optimization) algorithm were used.

Keywords: Clustering, Optimization, Outlier, PSO.

View PDF