Research On User Demand Prediction Model Of Open Government Data Platform

The opening of government data is a product that the information technology,political and social develop for certain period.Opening the data resources owned by the government to the society and sharing them with the public will promote the improvement of government governance capabilities and improve the level of development and utilization of public data resources.The establishment of Chinese government data open platform is concentrated in 2012.In 2013,the national data open platform “national data”(data.stats.gov.cn)established by the National Bureau of Statistics was officially launched.How to establish an open government data(OGD)platform and how to open data has become a research hotspot in recent years.The purpose of government data opening is to meet the needs of the public,improve the utilization of data resources,and then promote social and economic development.This paper focuses on the prediction model of user demand for government data open websites.Predicting the number of users' browsing and downloading of government data open platforms by optimizing BP neural network,reflecting the data needs of platform users and contribute to service-oriented government establishment.The paper uses the adaptive differential evolution algorithm to optimize the BP neural network,and establishes the user demand forecasting model of the government data open platform,and verifies the validity and effectiveness of the model.The major results are listed as follows:(1)Define the concept of “users” in the government data open platform to determine the measurement dimension of user needs.Analyze the user's objective use of the platform,and use the user access volume and user download volume of the government data open website as the quantitative indicators of user demand.User demand is predicted by using the gross of data sets,the total number of downloadable data sets,and the total number of machine-readable data sets as measurement dimensions.(2)Construct a combined forecasting model of user needs.Aiming at the random assignment of initial weights and thresholds to the standard BP neural network,it is easy to make the results into local minimum.After analyzing the validity of the improved method,the initial weight of the BP neural network is optimized by the selfadaptive differential evolution algorithm(SaDE).The threshold is searched for,and the obtained optimal results is taken as the initial input value,and further training is performed on this basis.(3)Verify the prediction model by numerical simulation.Firstly,the BP neural network and the proposed combination model are compared by a comparative example.The validity of the SaDE-BP neural network prediction model is verified,and the demand prediction has higher precision.Secondly,the forecasting model is applied to the prediction of the user demand of the Beijing government data service network.The results of this paper are for the government to accurately grasp and analyze the user needs on the data open platform,provide an operational forecasting method,providing a basis for the government to establish an open database by predicting the number of users of data open platform users and the amount of data downloads.It can more accurately predict user needs and target open data,thereby improving the utilization of government data to create higher social and economic value.