Research On Online Analysis Method Of Online Shopping User Behavior Security Based On Petri Net And Complex Event Processing

With the rapid development of Internet and the rapid popularization of online shopping,more and more consumers choose online shopping to buy goods.As a new way of shopping,online shopping not only improves people's life style,but also promotes the rapid development of economy.However,due to the virtuality and insecurity of the online shopping environment,the inherent defects and network risks of the software system pose a great threat to the security of consumers' accounts and funds.In order to ensure the safety of consumers' funds and accounts in the online shopping process,it is necessary to carry out real-time monitoring on consumers'online shopping behavior,and timely identify users' abnormal purchasing behavior in the process of online shopping.In this paper,an online analysis method based on Petri net and complex event processing(CEP)is proposed.Petri net is a formal model that is generally applicable to distributed systems and can accurately describe the concurrent relationship between multiple events.Meanwhile,Petri nets have rich analytical techniques and other derived models(Colored Petri nets,CPN),which facilitate the modeling,verification,analysis and identification of risk behavior patterns of online shopping users.Therefore,Petri net is an ideal model for constructing the risk behavior of online shopping users.CEP is a real-time processing technology,mainly through the method of event pattern processing and associated data flow,which can quickly identify risky or exceptional cases requiring special processing.In order to effectively identify users' risk behaviors,Petri net is combined with CEP in this paper.Firstly,the definition of behavior identification network model is proposed and the CPN based users' risk behavior identification models of online shopping are described.Secondly,the model is transformed into the event pattern language according to the transformation algorithm to ensure the semantic and grammatical correctness.Finally,the risk behavior is identified through the CEP technology,and the feasibility of the method is verified through the system simulation.The major innovation of this dissertation is as follows:(1)Aiming at the abnormal behaviors of users in online shopping,this dissertation describes a risk behavior identification model of online shopping users based on colored Petri nets according to two different perspectives of single user and multiple users,and verifies the correctness of the identification model through quantitative analysis.(2)On the basis of(1),a recognition mechanism of online shopping user behavior was proposed,and the hierarchical color Petri net modeling method was used to identify online shopping user risk behaviors.(3)In order to effectively ensure the semantic and grammatical correctness of the event pattern language,this paper proposes an algorithm for transforming the event pattern language into the risk behavior recognition model of online shopping users,which lays a foundation for the real-time processing of the CEP engine.(4)A risk behavior identification system for online shopping users based on CEP is constructed.On the basis of(3),the event pattern language is embedded in Esper,a complex event processing engine,to analyze and reason the real-time data stream of online shopping users,and to identify and deal with abnormal or malicious behaviors in the transaction process in real time.