测绘学报 ›› 2017, Vol. 46 ›› Issue (10): 1672-1677.doi: 10.11947/j.AGCS.2017.20170286

• 地图学与地理信息 • 上一篇    下一篇

多模态时空大数据可视分析方法综述

朱庆, 付萧   

  1. 西南交通大学地球科学与环境工程学院, 四川 成都 611756
  • 收稿日期:2017-06-02 修回日期:2017-07-24 出版日期:2017-10-20 发布日期:2017-10-26
  • 作者简介:朱庆(1966-),男,博士,长江学者特聘教授,博士生导师,研究方向为多维动态GIS与虚拟地理环境。E-mail:zhuq66@263.net
  • 基金资助:
    国家自然科学基金(41471320);国家重点研发计划(2016YFB0502303)

The Review of Visual Analysis Methods of Multi-modal Spatio-temporal Big Data

ZHU Qing, FU Xiao   

  1. Faculty of Geoscience and Environmental Engineering, Southwest Jiaotong University, Chengdu 611756, China
  • Received:2017-06-02 Revised:2017-07-24 Online:2017-10-20 Published:2017-10-26
  • Supported by:
    The National Natural Science Foundation of China(No. 41471320);National Key R&D Program of China(No. 2016YFB0502303)

摘要: 时空大数据可视分析是近年国际大数据分析与数据可视化领域研究的热点前沿,也是全空间信息系统的核心研究内容之一。本文针对时空大数据多源、多粒度、多模态和时空复杂关联的特点,按照描述性、解释性和探索性3个层次分类归纳了现有时空大数据可视分析方法,论述了时空大数据可视分析中多模态特征筛选、新型人机交互分析以及探索性可视推理等技术难点和主要发展动态。研究表明,以数据可视化为主的描述性可视分析研究相对成熟,以可视环境下交互式挖掘分析实现问题诊断为主的解释性可视分析已成为当前大数据分析的焦点,而面向复杂问题协同决策的探索性可视分析方法则是大数据分析有待突破的重要发展方向。

关键词: 时空大数据, 可视分析, 全空间信息系统, 多模态数据

Abstract: The visual analysis of spatio-temporal big data is not only the state-of-art research direction of both big data analysis and data visualization, but also the core module of pan-spatial information system. This paper reviews existing visual analysis methods at three levels:descriptive visual analysis, explanatory visual analysis and exploratory visual analysis, focusing on spatio-temporal big data's characteristics of multi-source, multi-granularity, multi-modal and complex association.The technical difficulties and development tendencies of multi-modal feature selection, innovative human-computer interaction analysis and exploratory visual reasoning in the visual analysis of spatio-temporal big data were discussed. Research shows that the study of descriptive visual analysis for data visualizationis is relatively mature.The explanatory visual analysis has become the focus of the big data analysis, which is mainly based on interactive data mining in a visual environment to diagnose implicit reason of problem. And the exploratory visual analysis method needs a major break-through.

Key words: spatio-temporal big data, visual analysis, pan-spatial information system, multi-modal data

中图分类号: