属性异构信息网络节点重要度评估方法
Evaluation method of node importance of attribute heterogeneous information network
吴亚洁 1邓喆1
作者信息
- 1. 贵阳职业技术学院信息科学系,贵阳 550081
- 折叠
摘要
为了提供预测网络行为、优化网络结构等的决策依据,提出属性异构信息网络节点重要度评估方法研究.根据网络拓扑结构与评估需求,进行节点重要度评估指标的选择;在实际评估过程中,根据网络类型、属性异构信息数量与评价中的实际需求等进行指标的调整;为确保在评估过程中所有指标具有可量化的特点,进行指标的预处理与信息熵权重的计算;根据指标的权重向量,将其与标准化矩阵进行加权相乘,得到指标对应的加权矩阵,以此为依据,实现对节点重要度的计算与综合评估.实验结果表明:所提方法可以实现对节点重要度评估结果的量化,能够进行精准有效的节点重要度排序,排序评估效果较好,且能够得到更多节点的重要度分布结构信息.
Abstract
In order to provide the decision basis for predicting network behavior and optimizing network structure,the impor-tance evaluation method of attribute heterogeneous information network nodes is proposed.According to the network topology struc-ture and the evaluation requirements,adjust the network type,the quantity of the index and the quantification of the indexes in the evaluation process,multiply the index and the standardized matrix according to the weighted matrix,and realize the calculation and comprehensive evaluation of the node importance.The experimental results show that the proposed method can quantify the impor-tance of node evaluation results,can conduct accurate and effective node importance ranking,the ranking evaluation effect is good,and can get more importance distribution structure information of nodes.
关键词
属性异构信息/信息熵权重/评估方法/重要度/网络节点Key words
attribute heterogeneous information/information entropy weight/evaluation method/importance/network node引用本文复制引用
出版年
2024
NETL
NSTL
万方数据