数据挖掘的光纤网络信息扩散方向挖掘模型

发布时间：2018-12-12 12:57

【摘要】：经典的扩散方向挖掘模型及其扩展模型多数从光纤网络结构出发研究信息扩散方向,较大程度上忽略了节点属性和信息内容的影响,存在信息扩散延迟问题。为此,提出数据挖掘的光纤网络信息扩散方向挖掘模型。深入挖掘光纤网络中作为扩散主体和客体的节点属性特征以及信息内容特征,对光纤网络节点间扩散概率和扩散延迟进行建模,建立光纤网络端至端信息扩散方向挖掘模型,利用随机梯度下降算法学习信息扩散方向挖掘模型中的各个特征的权重,结合光纤网络的网络拓扑特点,考虑了光纤网络中的不同感染状态,引入感染用户的衰减函数,从而达到对信息扩散过程进行预测的目的。实验结果表明,所提模型反映出的扩散方向规律更符合信息在光纤网络中的扩散过程,具有较高的预测准确率。
[Abstract]:Most of the classical diffusion direction mining models and their extended models study the information diffusion direction from the perspective of optical fiber network structure, ignoring to a large extent the effects of node properties and information content, and there exists the problem of information diffusion delay. Therefore, a data mining model of optical fiber network information diffusion direction mining is proposed. In this paper, the attributes and information content characteristics of nodes as the subject and object of optical fiber network are deeply excavated, and the diffusion probability and diffusion delay between nodes are modeled, and the model of information diffusion direction mining for fiber optic network is established. The random gradient descent algorithm is used to learn the weight of each feature in the model of information diffusion direction mining. Considering the network topology characteristics of optical fiber network, the different infection states in the fiber network are considered, and the attenuation function of the infected user is introduced. In order to achieve the purpose of information diffusion process prediction. The experimental results show that the diffusion direction law reflected by the proposed model is more consistent with the diffusion process of information in optical fiber networks and has a higher prediction accuracy.
【作者单位】：商丘工学院信息与电子工程学院;
【基金】：基金项目:河南省科技厅(62102210361)
【分类号】：TN929.11

【相似文献】