Research On Aggregated Influence-based Data Collection And Spreading Mechanism In Online Social Networks

The relationship between online users is expression of users' relationship in the real world to a certain extent.Compared with face-to-face communication,online social networks have stronger capabilities of information releasing,spreading and sharing.Hot topic can spread among millions of online users in just a few minutes,and then becomes top trending searches with high probability.To trace and quickly cool down rumors,hot topics,group emotions and even the trend of the international situation,we must collect data related to specific topics in online social networks for public opinion analysis.To collect and analyze the important data under a specific topic,it is necessary to find the influential user set corresponding to a specific topic,in other words,to find the target user set with the greatest value.At the same time,hot topics spread regularly among online users.Therefore,to efficiently select valuable content from online social networks,data collection and dissemination mechanism research is imperative.In this thesis,focus on the online social networks,combined with the needs of the project "Research on targeted convert collection and anti-traceable Access for Internet Data[BH2018-CF03-1]" with China Academy of Electronics and Information Technology of CETC,we conduct a comprehensive study on the data collection and information spreading.The main contributions are described as follows:(1)Due to big data generated by online social networks,large-scale network analysis usually confronts with unbearable computational complexity,resulting in failing to achieve expected results within the specified response time.Taking aggregate influence into account,we propose an algorithm to find the target users and then collect data.To retrieve and analyze the most important data within predetermined time,we study the problem of targeted covert collection with multiple granularity levels for Internet data in this paper.To find target data(or web users interchangeably),we use a tree indexes with the largest spanning tree to rank Internet data.Finally,we propose an effective collect algorithm for Internet data acquisition under different granularity levels.We validate our proposed scheme with the real data,which are gathered from an online social platform.The experiment results verify that our algorithm can find the target data and collect data covertly with different granularity levels.(2)Since heterogeneous users have different influences on the effect of information spreading,we classify users into different levels,and propose a topic spreading mechanism based on status theory.For an online social network with multiple influence levels of users,the update rules of users' degree of acceptability are designed firstly by taking the status of users,time window and parallel prorogation into consideration.Secondly,The STB model(Status Theory Based Model)is proposed to simulate the information spreading in online social networks.Thirdly,the seed nodes are chosen based on aggregate influence of users.Finally,we use crawled data as dataset to verify effectiveness of status theory for information spreading.Compared with classical information propagation models,the proposed algorithm can achieve better prediction accuracy.