| E-Government affair is such an electronic system by which efficiency is enhanced cost is lowered and service quality is improved using the modern technology such as information system and communication network to regenerate the organization and work flow. Nowadays it has become very important to the construction of national information system in the world. In order to ensure its better future, evaluation is indispensable, which can not only verify the effect of E-Government's deployments but also provide decision support.The development situation of E-Government is evaluated from three aspects: information maturity, government affair transparence and user satisfaction. Evaluation data is got from heterogeneous data sources and accessed in uniform format, whose security and reliability is guaranteed by several methods of cryptography when transferred on internet. Analytic hierarchy process (AHP), fuzzy neural network (FNN) and self organizing feature map(SOM) are used. AHP algorithm can check the situation of E-Government's development. Then FNN which combines the neural network and fuzzy theory can get rules from the evaluation data. SOM is used for cluster analysis and feature extraction. FNN is able to get rules for AHP to build expert matrix, and AHP can check a single city's situation while SOM can be used to evaluate the development as a whole.The reuse of the implementation of models is also considered, in which data and algorithms are designed independently. Through integrating three open source projects: sturts, Hibernate and spring, an evaluation system based on J2EE is built up, in which Struts is used as presentation tier and uses Hibernate as persistence tier and Spring as the middle tier to connect presentation tier and persistence tier and to make system with lower decoupling and facility maintenance.The models of AHP, FNN and SOM are all proved, in which performance and precision is guaranteed. Thus, an integrated solution is made for E-Government affair evaluation. |
PDF Full Text Request