| The online community is an important platform for people to express anddisseminate opinions and topics. The Valuable topic of online community usuallyinvolved in the issues of politics and people’s livelihood, and it is the basis andsource of the internet public opinion. The research of the method of topic featureextraction and valuable topic identification in the online community provide atechnology management tools and methods for monitoring and managing the topicof online community evolved into internet public opinion, the research hasimportant theoretical significance and practical value.This study research on the method of topic feature extraction and valuabletopic identification in the online community by used the theory and method of textdata mining, social network analysis, and bioinformatics. By constructing themodel of topic feature extraction of online community, this paper describes thesubject text data and use the method of multi-feature fusion classification andAgglomerative Hierarchical Clustering to extract topic feature. Based on the theoryof combination of bioinformatics and the method of social network analysis, thispaper proposes a detection method of hot topics that describes the members ofonline community and the relationships between topics in online community as thenetwork of protein-protein interactions; and identifies the key recognition elementsof hot topics in internet community by constructing the monitoring analysis modelof hot topics parameters; the method of combined parameters asynchronousdetection was established to identify valuable topic in the online community;finally, experiment verifies the rationality and effectiveness of this method. Thismethod provides theoretical support and effective method to monitoring andguiding the network public opinion for the government.This subject originates from problems of the Humanities and Social SciencesFoundation of the Ministry of Education (10YJA630055). |
PDF Full Text Request