基于TOPSIS的时序网络节点重要性研究

时序网络考虑事件发生的顺序，可以更准确地刻画复杂系统的演化特征。本文采用多属性排序方法（TOPSIS）对时序网络不同时间片段节点的影响力进行综合评价。具体的思想是通过计算不同层间相似性指标值与正理想解和负理想解的欧式距离，根据其接近正理想解和远离负理想解的程度对层间耦合关系的度量方法进行排名。基于Workspace数据集的实验结果表明，以优先链接指标（PA）度量时序网络时间层耦合关系，所挖掘出的重要节点准确率最高，在各时间层上平均达到50.82%。该文的工作为从多属性角度分析时序网络提供了借鉴。

Node Importance Identification for Temporal Networks via the TOPSIS Method

Temporal networks could describe the evolution characteristics of complex systems more accurately by considering the sequence of events. In this paper, the multi-attribute sorting method (TOPSIS) is introduced to comprehensively evaluate the influence of different time slices of temporal networks. By calculating the Euclidean distance of different inter-layer coupling indexes to the positive-ideal solution and the negative-ideal solution, this method ranks the indexes according to the measurement that the results are close to the positive-ideal solution and far away from the negative-ideal solution. The relevant experiments on Workspace datasets show that the Preferential Attachment Index (PA) to measure the temporal coupling relationship can dig out the highest accuracy, the average of 50.82% on each layer. Our work may shed some lights for analyzing temporal networks from multi-attribute perspective.