Research And Application Of Data Visualization Based On WEB Front-End

In the era of big data,data is growing exponentially.How to comprehend the data quickly and discover the inherent information of the data is an urgent problem to be solved.As a new technology,data visualization provides a convenient and efficient way for people to know data intuitively,comprehend data quickly and discover the rules of data by mapping the data to visual graphics.At present,the realization of data visualization application software is based mostly on C/S structure.This application software,usually with single usage and bad cross-platform characteristics,is often highly professional.As a mainstream Internet technology,Web front-end technology provides not only extensive applications,good cross-platform characteristics and better visual interaction results for the realization of data visualization based on B/S structure,but also Internet platforms for data visualization representation,all of which make data more available and understandable.It has become a trend to realize data visualization based on Web front-end technology.In this thesis,data visualization technology of low dimension data and multi-dimension data are researched.Based on these researches,Web front-end technology is exploited to realize the data visualization system and consequently some practical application cases are used to verify the visualization system.The contributions of this thesis are summarized as follows:1.Researches on low-dimensional and multidimensional data visualization technology,Web front-end technology and visualization technology framework are done.Besides,the interoperability of parallel coordinate visualization technology based on Echarts3.0 is also extended in this thesis.2.A PCA-based parallel coordinate visualization method(PCAP)and a PCA and cluster based parallel coordinate visualization method(PCAKP)are proposed.Because of the high dimension,the data visualization of multidimensional data usually causes high density of lines and narrow distances between coordinate axes of parallel coordinate plots.In order to solve the problem,a PCAP method is proposed.In this method,PCA is used to reduce the dimension of multidimensional data and then parallelization processing is taken on the dimension-reduced data,all of which effectively improve the visualization results of parallel coordinates.Because of the high dimension and immense data amount,the visualization of multidimensional data usually causes the problems that visualization lines are intensive and overlapping,which will greatly affect the acquisition of the features and rules of the data.In order to solve this problem,a PCAKP method is proposed in this thesis.Based on PCAP,the K-means clustering algorithm is used to cluster the PCA-processed data,and then the parallel coordinate visualization method is taken to visualize the resulting data,which not only improves the parallel coordinate visualization results,but also helps to obtain the features and rules of the data.3.A Data visualization system is designed and implemented in this thesis.According to the requirement analysis of the data visualization system,workflows of the visualization system,the Web front-end module,the data visualization processing and Web interface prototype are designed in this thesis.With Web front-end technology and visualization framework,this thesis also realizes a visualization system.Finally,the actual case is taken to verify the usefulness and effectiveness of the visualization system.In this thesis,the data visualization technology is researched and with Web front-end technology and visualization framework,the data visualization system is realized which not only satisfies the requirements of low-dimensional and multidimensional data visualization and achieves the effect of ideal visual graphics,but also provides a more effective way to acquire features and rules of the data.