基于用户与节点规模的微博突发话题传播预测算法

发布时间：2018-03-06 09:33

本文选题：微博网络　切入点：话题传播　出处：《通信学报》2013年S1期 　论文类型：期刊论文

【摘要】：突发话题传播建模与预测的主要目的是对网络中可能产生不良影响的、紧急性突发事件的后续传播进行控制。目前微博网络中的话题传播与预测研究尚处于起步阶段。通过对病毒传染模型、消息传播模型以及话题传播模型的深入研究,提出一种基于微博粉丝关系、用户活跃度和影响力的话题传播模型,将微博用户集合划分为感染用户、易染用户和免疫用户,分析感染用户和易染用户的粉丝关系,预测下个时间窗口内被感染的用户规模。沿用话题传播模型研究中的"内外场强"概念,通过研究发现"内场强"和"外场强"有特定的比例关系,基于用户群的规模大小,分别提出基于用户和节点规模的话题传播预测算法。相关实验表明,基于用户的算法预测更为准确但是时间复杂度较高,基于节点规模的算法则更适合大规模数据集的处理。
[Abstract]:The main purpose of the modeling and prediction of burst topic propagation is to have a negative impact on the network. At present, the research on topic propagation and prediction in Weibo's network is still in its infancy. Through the in-depth study of virus transmission model, message transmission model and topic transmission model, This paper presents a topic communication model based on Weibo fan relationship, user activity and influence. The user set is divided into infected users, susceptible users and immune users, and the fan relationship between infected users and vulnerable users is analyzed. To predict the scale of infected users in the next time window. Using the concept of "internal and external field strength" in topic propagation model research, it is found that there is a specific proportional relationship between "internal field strength" and "external field strength", based on the size of user group. The experiments show that the algorithm based on the user size is more accurate but the time complexity is higher and the algorithm based on the size of the node is more suitable for the processing of large-scale data sets.
【作者单位】：哈尔滨工程大学信息安全研究中心;国家计算机网络应急技术处理协调中心;
【基金】：国家自然科学基金资助项目(61170242,61272536) 国家高技术研究发展计(“863”计划)基金资助项目(2012AA012802) 中央高校基本科研业务费专项基金资助项目(HEUCF100601)~~
【分类号】：TP393.092

【相似文献】