The Research And Application Of Toll Audit In Networked Expressway Based On Data Mining

The expanding scale of expressway toll collection network,the increasing of trafficmileage,and the high profit cause tax evasion more and more violent.The means of taxevasion are constantly changing.One hand these caused the running away of the nationaltoll, and on the other hand these caused the unfairness of the toll payment,and disturbedthe management of the expressway,and also took some difficulty for the expresswaymanagement.Normal checking should take the comprehensive analysis of the backgrounddata from the expressway toll payment network as the basic means.It was very poorefficiency for the manpower method of checking and analysis through the huge data of theexpressway toll payment network.Missed checking and wrong judging for the gravedefrauding was usually occurred.By using the Data mining for the handling of the currentdata of the toll payment network,we can fully digging the information value of data,andcan enhance work efficiency. Using tech means to prevent and combate evasion tollsbehavior effectively is the sally port and origin of force of the inspection of the highwaytoll payment network. At the mean time,It is significant to build up and improveLong-term mechanism for the expressway toll payment network inspection,and to realizethe checking means from manpower to science and technology assistant, and to improvethe information management of the expressway toll payment network.This thesis is to clean the part worthy of study and suspected data of the expresswaytoll payment network, and to build corresponding data warehouse, and to extract valuableor unfound information before by analysing and digging the information of the datawarehouse,and to satisfy the necessary of data analysis of inspection fees of the highwayOperation management unit.Through the Guangdong expressway toll payment network,westated the structure, composing and related principles and data model for the expresswaytoll payment network.According to the research Summary of anti-defrauding for everyroad of the expressway toll payment network, and through a deep study and analysis ofdefrauding,we extracted the common feature of the unusual characteristic。Using theadvanced data mining function of SQL Server2008R2and taking the history data of oneroad of the east Guangdong as digging target, we selected decision tree arithmetic toforecast OD directional time,so we carried out dynamic overtime setting.We usedrelated rule arithmetic to dig out the unusual flow data of Gang escapesuspicion.According to the digged result and combining the checking system of tollroad,we extracted original transaction flow and corresponding proof information (such as access images,high definition images, the passing vehicle video)to provide referencebasis for the road operation department to judge defrauding vehicle and anti-defrauding.