基于微博数据挖掘的九寨沟7.0级地震灾情时空特征分析

2017年8月8日21时19分九寨沟发生7.0级地震,震后数小时里,大量与地震相关的信息广泛传播,互联网社交媒体高度关注,九寨沟地震成为最热议话题。本文以新浪微博为例,获取了距震中200km范围内震前、震后24h的微博数据,通过对数据清洗、分类和挖掘,分析了此次地震微博的数量、灾情分类、词频统计、时间序列和空间分布等特征,同时与实际灾评结果进行了对比分析。研究结果表明,对震后社交媒体数据进行充分挖掘,分析提取地震灾情关键信息,有助于对灾情的宏观把握,对救灾决策部署有一定的参考意义,是解决震后灾情获取难度大、覆盖小、时效性差等问题的一种有效的辅助手段。

Analysis of the spatial and temporal characteristics of disaster-information about the Jiuzhaigou,Sichuan MS7.0 earthquake based on data mining of Sina Weibo

At 21:19 p.m., August 8, 2017, a 7.0-magnitude earthquake struck Jiuzhaigou, Sichuan Province. In a couple of hours, a great deal of the earthquake-related information spread on internet. The earthquake drew much attention from the social media and soon became a hot topic. In this paper we searched those Sina Weibo users who are within the range of 200km from the epicenter, and copied their Weibo data released 24 hours before and after the earthquake event. After cleaning, mining, and classifying these data, we analyzed their characteristics such as quantity, word-frequency, and classification, spatial and temporal distribution. We found that extracting data from the social media would help governments learn overall the post-earthquake information, and on this basis make decisions and arrangements for earthquake relief.