Research On The Visualization Of Agricultural Soil Data Based On The Hadoop Platform

The arrival of the age of the data,the deepened the fusion of traditional agriculture and modern technology,with the further development of agricultural informationization process,data scale agriculture also becomes more and more big,the traditional data analysis form have already can't meet the needs of all walks of life,so the data visualization analysis has become one of the important ways in data analysis at present stage.In addition,in recent years,data privacy and other issues are constantly intensified,so the data visualization analysis method in the big data environment is facing great challenges.Based on the modern big data computing framework,this research focuses on the visualization of agricultural soil big data.The main work of this research includes four aspects:?.According to the characteristics of agricultural soil data in practical research and application,it is divided into natural attribute characteristics and geospatial attribute characteristics,and the visualization characteristics of the them are analyzed and the visualization scheme is designed.?.Consideration to the characteristics of agricultural soil data applications in python lib related data analysis method and mutual information(MI)correlation index selection algorithms,such as integration of a series of agricultural soil data pretreatment method(including cleaning,concise,neat and reduction,etc.),and soil data sets on Hadoop for association rules mining based on decision tree classification and prediction.?.Based on a variety of high-dimensional data visualization methods,this paper designed a visualization example of a relatively standardized soil data set after pretreatment,and based on this visualization example,improved the visualization research scheme of soil data to achieve an ideal visualization effect.Aiming at the visualization of geospatial data,based on the existing WebGIS agricultural data visualization system,this paper analyzes the problems such as data display,rendering speed and visualization space in the visualization of agricultural soil geographic data,and proposes corresponding solutions.Aiming at the technical bottleneck of agricultural soil data analysis research,in order to fit the future development of related data visual analysis platform,the idea of constructing a federal visual model of agricultural soil data is proposed in the end of this topic,to solve the increasingly prominent data privacy and data barrier problems in the society.This idea has certain reference value for further visual analysis of agricultural data and data visualization in other fields.