Analysis And Design Of High Dimensional Data Visualization Structure Model

Due to the rapid development of data collection and acquisition technology,the number and dimension of data is growing and high-dimensional data has become an important part of big data.A lot of valuable information is hidden in the complex high-dimensional data,and high-dimensional data visualization can help people get deeper rules.However,because large amount of information is carried by the high-dimensional data visualization interface,the hierarchical structure of the data information is not obvious,resulting in the inefficiency of information perception.Therefore,this paper is carried out through summarizing the characteristics and visualization process of high-dimensional data and analyzing the main influential factors and logical relations of high-dimensional data visualization.The goal of this article is to propose structural model of high-dimensional data visualization to guide the design of high-dimensional data.Firstly,the high-dimensional data attributes are deconstructed into entity attributes,associated attributes and temporal and spatial attribute based on the theory of information attribute theory and statistic theory,helping people to understand the characteristics of high-dimensional data information.Secondly,combined with three typical high-dimensional data types,the main features of the interface are summarized.Based on the user's cognitive characteristics and the main typical tasks,the causes of the problems in the process of high-dimensional information acquisition are analyzed.The perceptions,attention,memory and thinking needs of the user in the process of information acquisition from the high-dimensional data visual interface are put forward,and the high-dimensional data visualization user cognitive model with three typical types is established.Then,we put forward the specific design method based on the information characteristics of high-dimensional data and the user's cognitive needs.And a global view of high-dimensional data "information structure-visualization" mapping model is established,which reveals the core human-computer relationship of high-dimensional data visualization.Then,high-dimensional data visual design methods and strategies are combined and mapped to the visual structure,visual and interactive three aspects.Finally,according to the high-dimensional data "information structure-visualization method" mapping model,"bed and breakfast room high-dimensional data "is selected as an example to be designed.The design scheme breaks the complex interface form of the traditional high-dimensional data and adopts the user-defined data filtering layering,ensuring that the interface is simple ang the special demand of the high-dimensional data user's behavior and cognition are fully considered.Besides,feedback design,fault-tolerant design,associated aided tool design,system design and so on are applied in the scenario,which effectively reduces the user's cognitive load.The research content of this subject will fill the theoretical blank of the high-dimensional data user's cognitive model and the structure model of the "information structure-visualization" method,providing some reference for the future research on the visualization of high-dimensional data.