| With rapid development of Internet, the number of the people who use internet has reached 500 millions. And e-commerce has become important market behavior and life-style. Virtual property including net-money turns to an important medium in people’s behavior in internet. At present, crime and damage to virtual property is rampant. So protection and management of virtual property is important to maintaining stability in society.Because of huge quantity and various behavior styles of the users of virtual property, the data stream of users’ operation is huge, multidimensional and real-time. At the same time, it must be real-time and accurate for management system of virtual property to audit and trace abnormal operation. That is a challenge to traditional data management technology. To resolve the problem, we propose a model of data stream cube for discovering abnormal operation on virtual property. According to application, we optimize the data stream cube model, and study the technology of compression and storage, and design algorithms. The main work is as follows:1)On the basis of studying data stream technology, OLAP technology, data stream cube technology, we propose a data stream cube model of users’ operation on virtual property, and pre-process data stream, and design an algorithm.2)Study related improved technology of data stream cube, and propose partial materialization method on tracing dimensions, and optimize data stream cube model of users’ operation, and optimize materialization algorithm, and research partial update.3) Study the Quotient Cube technology, and optimize compression method according to query of data stream cube of users’ operation on virtual property, and adopt mixed storage model to storing data stream cube of users’ operation.4)Design and realize the system of storage of data stream cube of users’ operation on virtual property, and test the validity of the optimized model of data stream cube of users’ operation and compressed storage. |
PDF Full Text Request