Study On The Evolution Law And The Security Evolution Index System Of The College Network Public Opinion

As of June2011, the total number of the netizens in china reached485million, and the internet has become the fourth media of modern information exchange and dissemination. According to the relevant investigation, almost all colleges and universities in our country have established campus network. College or university students are the most active, positive and sensitive social group, easily using the internet to comment and express their views about focus or hot issues of social. They have become the main force generating the public opinion of network and the important effecting object of the public opinion of network. In recent years, the emergence and development of the various college and university unexpected events is almost inseparable from the spread of the network. In order to ensure the harmonious and stable development of the campus, the research of the college public opinion of network is urgently needed.The paper has researched the background of the college public opinion of network, pointing out the present situation of domestic and international research and the problems of the study, putting forward the main content of this thesis and the research ideas; then put an overview of the college network public opinion, defining its meaning, characteristics, and the effect; and then design questionnaire to research the effects of the network public opinion on college students. Based on the background, significance and the present research, objective analysis of the connotation, characteristics and the role, pointing out the impact of the college network public opinion on the college students, and then make sure of the necessity and the value of the network public opinion supervision, leading to two core thesis of the paper: the evolution law and the security evolution index system of the college network public opinion.About the problem of the evolution law of the evolution law of the network public opinion, it is divided into three sides to research:the time evolution law, the theme evolution law, and the disseminator evolution law. Based on the statistical analysis methods, I got the time evolution law: normal distribution type, exponential distribution, poisson distribution, and wave type distribution, putting out the response management measures based on the evolution law. Based on the texting mining to analyze the examples, I got the condition of the theme evolution in the different stages of the public opinion development, between different development stages of public opinion, and the main content of guide and the monitoring priorities are different. Using the social network analysis methods, researching the disseminators of the examples, which lead to get the disseminator evolution law of the different stages of the public opinion development, and dig out the more central and active disseminators, guiding and managing the disseminators, in order to grasp the development direction of network public opinion.About the problem of the building of the safety evaluation index system of network public opinion of the colleges and universities, firstly, through the literature review and the date collection to determine the safety evaluation index, then broke each indicators into small indicators, which is easy to quantify. Secondly, filtered the indicators and determined the weights of each indicators through the expert survey and the hierarchy analysis. Finally, built the safety evaluation index system of network public opinion of the colleges and universities.The Innovation of the paper is that based date mining sofeware, social network analysis software, and the empirical investigation analysis, analyzed the evolution of the network public opinion, and built the safety evaluation index system of the network public opinion in line with the colleges and universities, which is the three innovations of the paper. If combine the research with the warning system of the network public opinion, the practical significance will be more prominent, so the study also has a long way to go.