基于双曲映射算法的社会网络演化建模及传播源点定位方法研究
发布时间：2018-11-28 15:01
【摘要】：在社会网络中,各种谣言不断传播,对国家和社会的稳定造成极大的威胁,有效地定位信息传播源点对于预测传播范围、控制传播过程等具有重要的意义。社会网络的最主要的特性是动态特性,即随着时间的推进,社会网络中的节点可能增加或减少,其中的边也会增加或减少。因此,能够通过建立社会网络的演化模型很好地模拟社会网络的演化规律对于定位信息源点至关重要。本文进行的研究主要以两大前提条件为基础：一是假设社会网络的演化只是从边的增加角度进行；二是假设已知当前的传播拓扑和定位时间与真正传播拓扑形成时间的差值。本文与之前源点定位算法的最大不同是,考虑到社会网络的动态演化,从而在定位性能上较扩展的基于观察点的单源点定位算法有相对的提高。本文从社会网络的动态特性出发,考虑到在信息源点定位的情形中,源点定位时的网络拓扑与真正的传播拓扑不同,应用EPSO模型对社会网络进行建模,同时采用双曲映射算法预测社会网络中的链接。应用双曲映射算法,根据当前传播拓扑预测新生成的链接,将预测出的链接从当前传播拓扑中删除得到估计出的真正的传播拓扑,以这个拓扑为基础使用单源点定位算法预测信息传播源点。本文采用双曲映射算法在合成网络和实际网络上预测将来的链接,其传播模型是随机传播模型,这个模型是SI模型。在定位时进行了各种对比实验,比如在同一个传播拓扑下,不同的观察点部署策略；在同一个传播拓扑下,不同观察点部署比例。从实验结果来看,我们的算法的定位性能总体上优于扩展的基于观察点的单源点定位算法。由此,可以推断出本文提出的定位算法在社会网络上的信息传播源点定位中效果明显,它对于社会网络上的谣言定位和控制有重大作用。
[Abstract]:In the social network, the spread of various rumors poses a great threat to the stability of the country and society. It is of great significance to locate the source of information effectively for predicting the spread range and controlling the communication process. The most important characteristic of social network is dynamic characteristic, that is, the nodes in social network may increase or decrease, and the side of social network will increase or decrease. Therefore, it is very important to simulate the evolution law of social network by establishing the evolution model of social network for locating information source points. The research in this paper is mainly based on two prerequisites: one is to assume that the evolution of social networks is only from the angle of the increase of edges; the other is to assume the difference between the known current propagation topology and the time of localization and the time when the real propagation topology is formed. The biggest difference between this paper and the previous source point localization algorithm is that considering the dynamic evolution of social network, the single source point location algorithm based on observation point is relatively improved in the performance of localization. Based on the dynamic characteristics of social network and considering that the network topology of source point location is different from the real transmission topology, the EPSO model is used to model the social network. At the same time, hyperbolic mapping algorithm is used to predict the links in social networks. The hyperbolic mapping algorithm is applied to predict the newly generated links according to the current propagating topology, and the estimated true propagation topology is obtained by removing the predicted links from the current propagating topology. Based on this topology, a single source location algorithm is used to predict the source points of information propagation. In this paper, hyperbolic mapping algorithm is used to predict the future links on the synthetic network and the real network. The propagation model of hyperbolic mapping algorithm is a random propagation model, and this model is a SI model. A variety of comparative experiments are carried out, such as the deployment strategies of different observation points under the same propagation topology, and the different deployment ratios of the observation points under the same propagation topology. The experimental results show that the performance of our algorithm is better than that of the extended single source location algorithm based on observation point. Therefore, it can be inferred that the localization algorithm proposed in this paper is effective in locating the sources of information spread on social networks, and it plays an important role in the localization and control of rumors on social networks.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09;G206
[Abstract]:In the social network, the spread of various rumors poses a great threat to the stability of the country and society. It is of great significance to locate the source of information effectively for predicting the spread range and controlling the communication process. The most important characteristic of social network is dynamic characteristic, that is, the nodes in social network may increase or decrease, and the side of social network will increase or decrease. Therefore, it is very important to simulate the evolution law of social network by establishing the evolution model of social network for locating information source points. The research in this paper is mainly based on two prerequisites: one is to assume that the evolution of social networks is only from the angle of the increase of edges; the other is to assume the difference between the known current propagation topology and the time of localization and the time when the real propagation topology is formed. The biggest difference between this paper and the previous source point localization algorithm is that considering the dynamic evolution of social network, the single source point location algorithm based on observation point is relatively improved in the performance of localization. Based on the dynamic characteristics of social network and considering that the network topology of source point location is different from the real transmission topology, the EPSO model is used to model the social network. At the same time, hyperbolic mapping algorithm is used to predict the links in social networks. The hyperbolic mapping algorithm is applied to predict the newly generated links according to the current propagating topology, and the estimated true propagation topology is obtained by removing the predicted links from the current propagating topology. Based on this topology, a single source location algorithm is used to predict the source points of information propagation. In this paper, hyperbolic mapping algorithm is used to predict the future links on the synthetic network and the real network. The propagation model of hyperbolic mapping algorithm is a random propagation model, and this model is a SI model. A variety of comparative experiments are carried out, such as the deployment strategies of different observation points under the same propagation topology, and the different deployment ratios of the observation points under the same propagation topology. The experimental results show that the performance of our algorithm is better than that of the extended single source location algorithm based on observation point. Therefore, it can be inferred that the localization algorithm proposed in this paper is effective in locating the sources of information spread on social networks, and it plays an important role in the localization and control of rumors on social networks.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09;G206
【相似文献】
相关期刊论文 前10条
1 徐红波;;空间填充曲线映射算法研究[J];科技信息(科学教研);2007年35期
2 孙培展;袁国良;;改进的隐式空间映射算法的研究[J];电子设计工程;2012年09期
3 赵文庆;基于性能驱动的工艺映射算法[J];计算机辅助设计与图形学学报;1992年03期
4 黎洪松;;一种改进的自组织特征映射算法[J];中国民航学院学报;2006年01期
5 徐德智;黄利辉;陈建二;;一种新的基于树分割的本体映射算法[J];小型微型计算机系统;2009年11期
6 吴国福;窦强;窦文华;;基于查表的空间填充曲线映射算法[J];国防科技大学学报;2010年05期
7 陈];;心动阵列的自动映射算法[J];计算机研究与发展;1992年05期
8 黄胜;吴川川;杨晓非;王辉;张卫;;一种基于临近原则的虚拟网络映射算法[J];电信科学;2013年12期
9 柳玉起;李明林;冯少宏;易国锋;;基于有限元映射算法的试验网格显示及其应用[J];华中科技大学学报(自然科学版);2007年03期
10 王琳珠;单_，
本文编号：2363216
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2363216.html
