论文部分内容阅读
超链路预测首次被提出,是指在超网络结构上,对未知的超链接及未来超链接的预测。超链路预测是对现有超网络研究的一项重要推进,实现了对复杂超网络的结构演化、调节干预和预测分析等。本文采用网络结构相似性方法,引入超三角形的概念作为相似性的度量指标,提出基于超边相似性的超链路预测算法,并将其应用于实际的舆情超网络模型———钱云会事件中,最后用AUC方法评价超链路预测与链路预测在舆情超网络模型中的优劣。
The first proposed hyperlink prediction is the prediction of unknown hyperlinks and future hyperlinks in the super network structure. Hyperlink prediction is an important advance for the existing super-network research and realizes the structural evolution, regulatory intervention and prediction analysis of complex super-networks. In this paper, we use the similarity of network structure to introduce the concept of hyper-triangle as a measure of similarity, and propose a hyperlink prediction algorithm based on hyper-edge similarity and apply it to the actual super-network model of public opinion --- Qian Yunhui In the event, the AUC method was used to evaluate the pros and cons of hyperlink prediction and link prediction in the public opinion hyper-network model.