Synthetic Evaluation Model About Performance Of E-government Based On PSO-BP Neural Network

The rapid development of economy pulls the unceasing progress of informationtechnology and computer network technology. It also greatly promotes the development ofinformation industry of our country. E-government has became the the product of trend ande-government performance evaluation has became a big problem in the process of theinformation. How to implement an effective e-government project? How to minimize the risk?Which standards can measure whether the electronic government affairs has reached theexpected effect? These problems have became a challenge for academia and the government.This article built a comprehensive and effective e-government performance evaluation modelbased on PSO and BP neural network. In view of the slow convergence speed of BP neuralnetwork, the module needs a long training time. The PSO optimization algorithm was appliedto the model to improve BP neural network model. It raised principle and algorithm of BPneural network based on PSO. We use the actual example to establish the model of trainingand validation.In this paper, the main research work is as follows:(1) theoretical research. This paper collected a large amount of relevant literature andnetwork information. On the basis of a large number of scholars at home and abroad we putforward this topic research background. On the relevant research achievements weresummarized. This article made the detailed overview with the theory of knowledge.(2) built e-government performance evaluation index system model based on thebalanced scorecard. The balanced scorecard is a new performance management system. It ismainly used for the implementation of the enterprise financial strategy. We builte-government performance evaluation index system model using the core content of balancedscorecard: the service object index, cost benefit index, internal operational indicators, learningand development indicators.(3) the construction of e-government performance evaluation model based on PSO andBP neural network. It put forward the e-government performance evaluation model based onBP neural network. It constantly adjust the parameters of the model and select the appropriatefunctions. It optimizes neural network training and learning process by using PSO algorithm.(4) to verify. We selected eleven city of Hunan province to carry out electronicgovernment affairs. It chose some data to train the model and used the remaining data toverify this model. By comparing the results of before and after optimization, the conclusion for the optimized neural network model of error is reduced and improved the convergencespeed.This study is intelligent algorithm and combinatorial optimization theory and methods ofcross and penetration. In order to establish a perfect and efficient e-government performanceevaluation model, I have carried on the beneficial exploration. Research results show that theBP neural network model has the defects of slow convergence speed and long training time.The model overcomes this deficiency by being optimized by PSO. In this paper, the researchprovides decision basis and the reference model for the government to carry out theperformance evaluation of e-government. It has important practical significance to theelectronic government affairs. It can also provide a certain reference significance for theenterprise performance evaluation.