基于节点和信息特征的社会网络信息传播模型

发布时间：2018-02-11 01:06

本文关键词： 社会网络信息传播传播模型节点属性传播预测　出处：《计算机研究与发展》2015年01期 　论文类型：期刊论文

【摘要】：随着在线社会网络的快速发展,越来越多的人开始利用微博或Twitter来传播信息或分享观点.研究社会网络中的信息传播规律对于意见领袖挖掘、舆情监控、品牌营销等有着重要意义.虽然有关社会网络中的信息传播模型已经得到广泛研究,但是影响网络中节点之间信息传播的因素有哪些,以及如何刻画信息传播过程,仍然是一个有待深入研究的重要内容.传统的传播模型及其扩展模型更多地从网络结构出发研究信息传播,很大程度上忽视了节点属性和信息内容的影响.从多个维度提取信息传播的特征,包括节点属性特征和信息内容特征,对节点间传播概率和传播延迟进行建模,提出一个细粒度的在线社会网络信息传播模型.利用随机梯度下降算法学习模型中的各个特征的权重.另外,针对模型的传播预测功能,在新浪微博真实数据集上进行了实验,结果表明,在预测准确率方面,所提出的模型要优于其他同类模型,如异步独立级联模型、NetRate模型.
[Abstract]:With the rapid development of online social network, more and more people begin to use Weibo or Twitter to spread information or share views. Brand marketing is of great significance. Although the information communication models in social networks have been widely studied, what factors affect the information dissemination between nodes in the network, and how to describe the process of information dissemination, It is still an important content to be studied deeply. Traditional communication models and their extension models study information dissemination more from the perspective of network structure. The effects of node properties and information content are largely ignored. The characteristics of information transmission are extracted from multiple dimensions, including node attribute features and information content features, and the propagation probability and propagation delay between nodes are modeled. This paper presents a fine-grained online social network information transmission model. The stochastic gradient descent algorithm is used to learn the weight of each feature in the model. In addition, aiming at the propagation prediction function of the model, an experiment is carried out on the real data set of Sina Weibo. The results show that the proposed model is superior to other similar models in prediction accuracy, such as asynchronous independent cascade model and NetRate model.
【作者单位】：国防科学技术大学计算机学院;数学工程与先进计算国家重点实验室;解放军信息工程大学;
【基金】：国家“八六三”高技术研究发展计划基金项目(2009AA012201)
【分类号】：TP393.09

【参考文献】