关键词: 大学生, 网络行为追踪, 融媒体, 网络信息推送

Abstract: With the rapid development of artificial intelligence, fusion media, mainly represented by Weibo, Wechat, B station and Tiktok, have become the main means of communication in the era of information explosion. The matrix spread of fusion media is extremely rapid, and the network space constructed by them has the characteristics of circle segmentation, social negativity and extreme behavior. In the important period of the formation of their outlook on life, world outlook and values, college students are extremely vulnerable to the impact of multi-culture and social trends of thought. They have always been the outbreak point of contradictions of social transformation and ethos trends. As the youngest and most active group of Internet users, the network behavior of “post-00s” college students and its characteristics are worthy of our study in depth. The research on fusion media network information push based on the perspective of network behavior tracking of “post-00s”college students is of great significance for universities and government departments to understand, supervise and govern the ideological dynamics of college students scientifically and precisely.
   Aiming at the shortcomings of existing research methods to deal with the preference relationship, in this paper, the intelligent hybrid push model of fusion media information is constructed based on the internal relationship between the information browsing habits and preferences of “post-00s” college students. An integrated model combining DEMATEL and TOPSIS methods with intuitive fuzzy numbers (IFN) is presented. In different time periods, the information set that college students are most concerned about is effectively sifted from a large number of network information. The internal correlation degree of information is extracted through analyzing the interaction between indicators and considering the level of criteria. The types of network information that college students pay most attention to in different time periods are identified. Then, a personalized preference model of college students’ financial media network information push is designed. Finally, the validity and applicability of the proposed method are verified by empirical data simulation. The research results show that the fusion media information push model based on the perspective of online behavior tracking of “post-00s” college students can carry out detailed differentiation calculation and classification screening for big data. It can provide the reference for network supervision departments to obtain, judge and filter information. It is of great theoretical and practical significance for universities and government departments to grasp the ideological dynamics of college students and fully and deeply participate in ideological and political education of college students under the environment of integrating media. It provides scientific and effective technical support for promoting the right to speak in cyberspace and establishing a community of shared future in cyberspace.
   In the actual process of network information push, the amount of network information is very large, and the classification of network information is also very important. Especially, in the age of artificial intelligence, it is more and more urgent for intelligent algorithms to drive the modernization of network governance. How to build an intelligent push algorithm for cloud computing is the next direction of research. In addition, the weight refinement of secondary indicators and the differential treatment of indicators are also worthy of further study. Therefore, in the future, we hope to put forward a decision-making method that is as fair as possible to the differentiation index system of college students’ Internet users’ preferences.

Key words: college students, network behavior tracking, fusion media, network information push