Designing And Application Of A Large-scale And Fast Malicious Web Page Recognition Method Based On Combination Of Kafka And Spark-streaming

People have become more and more dependent on various Internet applications,which may lead to customers much more risk exposuing to Internet attacks.The malicious web pages are the most typical attack modes,which has been widley accepted as great threat to Internet security.This thesis presents a comprehensive research on large-scale fast malicious web page recognition and its applications via theoretical analysis and experiments depending on the current threat of malicious web pages to Internet security.The thesis firstly gives a brief introduction of the malicious web page definitions,typical precaution methods,and some relative technologies such as Kafka and Spark-streaming,etc.Then,the methods of web page sample set acquisition and feature extraction are analyzed in detail.This thesis provides malicious web page identification and detection method designing,which introduces Weka tool to classify the training sample data and construct system model,respectively.Three different algorithms,i.g.the support vector machine algorithm,naive bayesian algorithm,and linear neural network algorithms are implemented and compared.Finally,this thesis proposes system application design.Test results show that the proposed system schene has better malicious web page recognition rate compared to other static detection methods,which can meet the needs of large-scale malicious web page recognitions.