Research On The Predictive Model Of The Usage Status Of Open Government Data

With the quantity increasing of open government data platform, there is a hot research issue about how to effectively enhance the level of government open data service. Public timely access to and use the data that government opened is an important way that the data resources affect social and economic development. Therefore,describing and forecasting the using status of open government data has a realistic significance. Using the scientific method to predict the usage status of government data after it opened reasonably,can be beneficial to stay current with the public's data needs as well as the change of demand, and help to enhance ability of data resources integration,and is helpful to improve the comprehensive service level of open government data.The purpose of this paper is to build open government data usage status prediction model by using literature method, combination of qualitative and quantitative, and empirical research, and then verify the feasibility and effectiveness on the base of analysis and research the usage of open government data resource at present. The main research content can be divided into the following three parts:First of all, this paper puts forward two dimensions to represent data usage status.By studying the related literature at home and abroad, and synthesizing characteristics and deficiencies of the related studies, puts forward two dimensions to represent data usage status. The dimension composed by the public attention and the public utilization which are work from the perspective of data user's behavior.Second, this paper builds a prediction model of the usage status of data. Author analyze the characteristics of open government data from two aspects: data attribute and data management, and then consider that using the data field, data format, data source,subject classification,open condition and update frequency as variables to make predictions. Through the comprehensive analysis of the advantages and disadvantages of the commonly used prediction method, author build a prediction model of the usage status of data by using optimized BP neural network model. The model is optimized by using genetic algorithm.Finally, this paper empirically analyses the prediction model. By using the data of Shanghai municipal government data service as a sample, this paper validates the data usage status representation and prediction model presented in this paper. The empirical result shows that the prediction model built in this article has great feasibility. Moreover,it reflected the validity of the method from the perspective of quantitative objectively.The research of this paper provides a feasible prediction analysis method so that the government can stay current with the usage status of their data resources better. This not only be helpful to improve the comprehensive service level of open government data, but also provide basic theories and practical basis to the study in terms of open government data.