微博意图分类在地震事件应急中的应用研究

Research on the Intention Classification of Weibo in the Emergency of Earthquake Events

探究微博意图分类在地震事件应急中的研究,通过爬虫程序获得和地震相关的微博文本数据,对微博文本数据进行预处理后得到所需要的分类数据,然后通过WEKA平台应用决策树、支持向量机和朴素贝叶斯多项式模型进行文本分类实验探究。实验结果表明采用支持向量机模型分类准确度最高达到90%以上。通过对微博地震数据的意图分类探究,可以及时得到地震中寻求帮助和提供救助的相关信息,合理调度应急资源。

Mainly explores the research of micro-blog＇s intention in the earthquake event emergency, obtains the micro-blog text data related to theearthquake through the crawler program, and preprocesses the micro-blog text data to get the classified data we need, and then applies thedecision tree, the support vector machine, and the naive Bayes polynomial model through the Weka platform, carries out the type of textclassification experiment. The experimental results show that the accuracy of the support vector machine model is over 90%. By studyingthe intention classification of micro-blog seismic data, gets information about help and rescue in time, and rationally dispatch emergencyresources.