为解决在互联网文本信息爆炸性增长的前提下,在大规模文本数据中如何发现隐含的、 有价值的潜在知识的问题,提出基于多层次文本聚类的文本知识挖掘方法,针对不同规模的文本数据进行不同粒度的聚类,实现不同层次知识的挖掘.针对最广义层次的文本知识挖掘可实现各主题事务划分,针对子级分类数据的文本知识挖掘可发现下一层次主题分类,针对自定义层次的文本知识挖掘可发现该事件中存在的具体细节.对诉求实际数据的分析结果表明,该方法可在所有诉求数据中挖掘出各种诉求主题,精确挖掘出其中的细节问题,为管理者提供数据和决策支持,提高服务效率.

The premise of Internet social governance is grasping the current situation of Internet social security risk correctly. Therefore, the Internet social security risk assessment has important practical significance. [Method/Process]Based on the current situation and security problems of Internet society in China, this paper constructed the risk assessment index system of Internet so﹣cial security in five dimensions, namely, political security, economic security, social security, cultural security and ecological security. The security cloud data of Tencent ranked the status of Internet social security in 31 provincial administrative regions and 32 central cities in mainland China. [Result/Conclusion]The overall situation of Internet social security risk in China is not optimistic and there are great differences among different regions. Local governments need to improve the ability to deal with the Internet social security risk from the i﹣deological guidance, Internet financial supervision,Internet content construction, Internet crime, infrastructure security and other aspects.