Analyze and Recommend News Comments in E-Government

With the development of Internet, more and more public users prefer to present their viewpoints of government policies. They often comment on some emergencies through news, blogs and so on. Their opinions influence decision makers of government to make right decisions. However, large numbers of news and related comments are produced when an emergency occurs and officers are very difficult to read and analyze all of them in seconds. Specially, comments usually are short texts and common clustering technologies are not suited to analyze them. In this paper, we firstly propose a framework based on semantic web technologies to recommend news and related comments in order to aid different officers to get their interesting news rapidly. Then, a new short text clustering method is discussed to analyze related comments. Finally, a news recommender system based on above approaches is introduced.