拉美一体化发展特征可视分析方法

Visual Analysis Method for the Development Characteristics of Latin America Integration

对拉美一体化发展特征分析的关键在于探索各国间的属性信息,但评估国家发展的因素众多且存在大量对结果无关因素的干扰。针对拉美国家数据的高维多元特征,提出一种利用层次数据划分和子空间分析相结合的可视分析模型,可同时分析属性相关性和局部子空间下国家的降维结果,再结合视觉隐喻等多种可视化手段和交互联动视图,设计了一个基于拉美国家数据的高维多元可视分析系统,支持对多个国家的关联对比和多属性的时变演化规律分析。对拉美国家数据从整体到个体、从宏观到特征事件进行筛选和探索,有助于认识拉美一体化发展的特征从而制定精准的扶贫计划。结合领域专家的需求进行了系统建模和案例分析,验证了本系统的有效性和适用性。

The key to the analysis of the development characteristics of Latin American integration is to explore the attribute information between countries.However,there are many factors to evaluate the development of the country and a large number of factors that interfere with the results.A visual analysis that combines classification and subspace analysis is proposed aiming at the high-dimensional multivariate characteristics of Latin American data.It can simultaneously analyze the dimensionality reduction results of countries under attribute correlation and local subspace.Combined with visual metaphor and other visual means and interactive views,a high-dimensional multi-visual analysis system based on data of Latin American countries is designed.It supports the correlation comparison of multiple countries and the analysis of time-varying evolution of multiple attributes,screening and exploring Latin American data from the whole to the individual,from macro to feature events,which helps to understand the characteristics of Latin American integration and to develop accurate poverty alleviation plans.Combined with the needs of domain experts,the system modeling and case analysis are carried out to verify the effectiveness and applicability of the system.