基于文本挖掘的公共事件分析(2012-2014) ——类别、涉事者、地理分布及演化
What, Who, Where, and How: Text Mining of Public Events in China (2012-2014)
本研究以2012-2014年间我国3359个公共事件为对象,利用描述统计、文本挖掘方法以及空间地理可视化呈现等方法,定量地描述公共事件的基本特征。研究发现,公共事件参与者身份、事件类型以及事件发生的空间地理特征皆呈现出不平等分布趋势。具体而言,涉事群体/个体身份多集中于政府部门、官员和公检法机关;事件类型中“制度危机与社会公平类”问题占有压倒性优势;从地理分布来看,公共事件多发生于经济较发达地区;从事件演化模式来看,不同类别的事件呈现出竞争关系,即大部分事件发生频度呈负相关关系。
This study describes the dynamic pattern and the features (i.e., what, who, and where) of public events happened in China during the last two years. The data were extracted from two independent databases of Chinese public event. By analyzing the 3,359 cases using descriptive statistics and text mining techniques, this study found that the event type, the identity of involved social groups/individuals, and the geographical location of public events are severely unequally distributed. In specific, the event type that most frequently happened is the events regarding “institutionalization crisis and social equality”; the location where public events often occurred concentrate on developed regions, such as the Southeast Coast; and the identities of deeply involved social groups/individuals include government departments and officers, public security organs, procuratorial organs, people's courts, and ordinary people. In addition, it is found that the frequency of different types of public events competed with each other during the past two years as evidenced by the negative correlation coefficients of event frequency across event types.
/
| 〈 |
|
〉 |