根据公共安全网络舆情研究的需求,将文本分类技术应用于突发事件应急管理中,提出了基于TFIDF分类方法的突发事件引发的网络舆情信息分类方法,设计类别样本并读入新闻文本,对文本进行中文分词,通过计算新闻文本和类别样本的相似度将新闻文本分到相似度最大的类别之中。通过编程实现了按照事件类型和地理位置两种分类方式对新闻文本进行分类,程序分类结果验证了该方法的实用性。
According to research needs on network public opinion in public safety,applying text classification technology in the field of emergency management for incidents,this paper presents a TFIDF-based classification method for network public opinion triggered by the incidents,which first designed the samples of categories and inputted the text of the news corpus,then proceeded Chinese word segmentation on the text,and at last assigned the news to the most similar category by calculating the similarity of the news text to the category sample.By programming the classification of news text in two types of event and of location is realised in this paper,and the results verified the practicality of the method.