在救灾智联网中推测信息孤岛救助需求强度的空间信息扩散模型

Geospatial information diffusion model for inferring the demand intensity of assistance in information islands based on the Internet of intelligences for disaster mitigation

重大自然灾害中的信息孤岛,是精准救助的盲区。本文认为,只有发挥互联网的优势,从“隔空判灾”走向“采点外推”,才能较高精度地快速评估信息孤岛中的灾情和救助需求。本文依托救灾智联网,由已观测地理单元上的数据,用空间信息扩散模型构造背景数据与救助需求的因果关系,据此推测信息孤岛中的救助需求。以2019年长宁6.0级地震为例子,以“相对暴露度”、“人口”和“GDP”等为背景数据,用启动应急响应3小时内网络爬虫数据中受灾乡镇名字出现频率的强弱表征应急救助需求强度,我们由10个镇得到的数据构造因果关系,推测了3个信息孤岛镇的救助需求强度。由于空间信息扩散模型的普适性,该模型对识别非线性的救助需求因果关系,具有一定的优势。

The information island in major natural disasters is the blind spot of precise rescue.This paper argues that only by giving full play to the advantages of the internet,from“judging disasters from a distance”to“extrapolating from mining points”,can the disaster situation and demands of assistance in information islands be evaluated quickly and accurately.In this paper,relying on the data of Internet of intelligences for disaster mitigation and the observed geographical units,the causal relationship between the background data and the demands of assistance is constructed by using the geospatial information diffusion model,and the demand of assistance in the information island is inferred accordingly.Taking Changning M s 6.0 earthquake in 2019 as an example,with“relative exposure”,“population”and“GDP”as background data,the demand intensity is characterized by the frequency of the names of affected towns in the data of web crawlers within 3 hours of starting emergency response.We construct causal relationship from the data obtained from 10 towns,and speculate the demand intensity of 3 information isolated towns.Because of the universality of geospatial information diffusion model,this model has certain advantages in identifying the nonlinear causal relationship of demands of assistance.