热带地理 ›› 2020, Vol. 40 ›› Issue (2): 184-193.doi: 10.13284/j.cnki.rddl.003239

• “地理空间智能技术及应用”专题 • 上一篇    下一篇

自发地理信息在灾后恢复监测中的应用研究框架

严滢伟1,2, 马大伟3(), 范红超4   

  1. 1.广东省遥感与地理信息系统应用实验室//广东省地理空间信息技术与应用公共实验室//广州地理研究所,广州 510070
    2.南方海洋科学与工程广东省实验室(广州),广州 511458
    3.武汉大学 遥感信息工程学院,武汉 430079
    4.挪威科技大学 土木与环境工程学院,特隆赫姆 7491
  • 收稿日期:2020-03-12 修回日期:2020-04-07 出版日期:2020-03-10 发布日期:2020-05-15
  • 通讯作者: 马大伟 E-mail:dawei.ma@whu.edu.cn
  • 作者简介:严滢伟(1987—),男,江苏人,助理研究员,博士,主要从事自发地理信息(众源大数据)研究,(E-mail) yanyingwei@u.nus.edu。
  • 基金资助:
    国家自然科学基金(41901330);国家自然科学基金(41771484);广东省科学院发展专项资金项目(2020GDASYL-20200103005)

A Research Framework for the Application of Volunteered Geographic Information in Post-Disaster Recovery Monitoring

Yan Yingwei1,2, Ma Dawei3(), Fan Hongchao4   

  1. 1.Key Lab of Guangdong for Utilization of Remote Sensing and Geographical Information System// Guangdong Open Laboratory of Geospatial Information Technology and Application// Guangzhou Institute of Geography, Guangzhou 510070, China
    2.Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou 511458, China
    3.School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China
    4.Department of Civil and Environmental Engineering, Norwegian University of Science and Technology, Trondheim 7491, Norway
  • Received:2020-03-12 Revised:2020-04-07 Online:2020-03-10 Published:2020-05-15
  • Contact: Ma Dawei E-mail:dawei.ma@whu.edu.cn

摘要:

文章探讨了如何有效利用自发地理信息(Volunteered Geographic Information, VGI)大数据促进灾后恢复监测工作。首先概述了国内外VGI相关研究的发展现状,明确了VGI用于灾后恢复监测研究的不足,然后提出了一个基于VGI大数据的灾后恢复监测应用的研究框架,助力于灾后恢复监测各类具体恢复目标(如旅游业恢复、工商业恢复、生活常态恢复)的实现。该研究框架包含数据获取、数据质量控制和数据挖掘3个核心组成部分。其中,数据获取对象以VGI为主,以传统官方权威数据为辅;数据质量控制主要是通过模糊逻辑专家系统和人工神经网络(深度学习)确保VGI适用性;数据挖掘则是以变革式范例为理论基础,利用定量和定性结合的方法调查灾区基建、经济和安全3个灾后恢复主要方面的状态。最后,文章还讨论了当前利用VGI大数据促进灾后恢复监测所存在的一些局限性,包括VGI来源的可持续性问题、各VGI平台应用程序接口的数据获取限制问题和VGI应用所涉及的用户隐私问题。

关键词: 自发地理信息, 大数据, 灾害管理, 灾后恢复监测, 应用研究框架

Abstract:

In recent years, many studies on the application of Volunteered Geographic Information (VGI) to natural disaster management have been reported. To date, it has been shown that VGI can provide big geospatial data, incorporating rich content and extensive spatiotemporal coverage, in a real-time and cost-effective manner. Thus, it can facilitate disaster management by filling information voids known to occur in traditional geospatial datasets. However, existing studies have mostly focused on disaster prevention, preparedness, and response phases, and few studies have focused on post-disaster recovery. The aim of the proposed work is to help bridge this research gap by investigating how VGI data can facilitate post-disaster recovery monitoring, both in general, and for specific aspects such as tourism, business, industry, and the daily routines of residents. In the work reported here, we first review the development of VGI research in the decade since 2007, and then propose a research framework for post-disaster recovery monitoring, based on VGI data. The research framework involves three key components—data acquisition, data quality control, and data mining. Data acquisition is generally referred to as VGI collection (for example, OpenStreetMap, Twitter, and Flickr data), while authoritative data (such as remote sensing data, official statistics, and field survey data) can be collected as ancillary information. Data quality control is based on a fuzzy expert system, which considers Linus’ law, metadata, data lineage and provenance, geographic contexts (Tobler’s first law of geography), user credibility, spatiotemporal data density, and user activeness, and leverages artificial neural networks (deep learning) for optimizing the fuzzy rule sets of the expert system. Data mining is based on the transformative paradigm, using a combination of quantitative and qualitative approaches, involving text mining, spatial statistics, and machine learning. Using this combination, the status of infrastructure (roads, electricity, water, transport, housing, buildings, and telecoms), the economy (external sources of economy, internal sources of economy, and services), and safety (reputation, secondary disasters, health, security, and stability), which are the three main measures of post-disaster recovery, are investigated. This work thus paves a way for future studies related to this topic, identifying both research and practical implications. We also identify limitations in using VGI for post-disaster recovery management. The first of these is that a VGI source may rapidly become obsolete due to the fast development of cyberspace, making it imperative for researchers to keep the technique up-to-date, adapting it to the diverse and emerging VGI sources. The second limitation is that certain VGI platforms do not provide all the data available in their databases, and have imposed constraints on their Application Programing Interfaces (APIs). The third limitation is related to privacy issues pertinent to VGI data use. We suggest that the future work should also focus on seamlessly integrating VGI with traditional post-disaster recovery monitoring approaches, such as remote sensing, field surveys, and resident interviews.

Key words: Volunteered Geographic Information, big data, disaster management, post-disaster recovery monitoring, research framework

中图分类号: 

  • P208