Web Robot Detection Based On Fuzzy System and PSO Algorithm

Abstract

Web robots are applications which recursively and automatically overview the content of website documents. Some robots are considered to be malicious. Hence, identifying web robots is a classification challenge. In this research we mainly present a hybrid particle swarm optimization method and fuzzy system aiming at increasing efficiency over Web robot detection and simulation by MATLAB software. Evaluation criteria are considered: Specificity, Accuracy, F- Measure, Recall and Precision. Results of the study for the proposed method are respectively: 99.81, 96.92, 96.10, 91.39 and 99.58. The yields for proposed basic fuzzy system and fuzzy network algorithm and ANFIS neural fuzzy network algorithm indicate the priority of the proposed method other than algorithms being compared.