CAJ | 학술논문

近些年来，随着对各种网络图数据分析的不断深入，越来越多的研究开始关注复杂网络的数据分析。面对现实生活中不断的动态变化，且更加复杂、多维度的社交网络或实体关系，亟需寻找新的方法来支持这样的应用需求。针对维度不断动态变化（增边或删边）的全动态多维网络上的节点重要性度量问题，提出了多维网络局部介数中心度度量算法以及动态环境下的局部介数中心度更新算法。首先将多维网络映射为一维映射图，并在此基础上利用 K 最长路径约束的局部介数中心度来衡量节点重要性，进而提出了两步过滤剪枝算法来加快动态网络环境下的局部介数中心度更新过程。接着对算法进行了理论分析，并分析了算法的应用场合。最后通过在模拟和真实数据集上的实验，探讨了最长路径约束 K 的大小选择，并展示出文中所提的全动态多维网络局部介数中心度更新算法具有良好的加速性能。
근사년래，수착대각충망락도수거분석적불단심입，월래월다적연구개시관주복잡망락적수거분석。면대현실생활중불단적동태변화，차경가복잡、다유도적사교망락혹실체관계，극수심조신적방법래지지저양적응용수구。침대유도불단동태변화（증변혹산변）적전동태다유망락상적절점중요성도량문제，제출료다유망락국부개수중심도도량산법이급동태배경하적국부개수중심도경신산법。수선장다유망락영사위일유영사도，병재차기출상이용 K 최장로경약속적국부개수중심도래형량절점중요성，진이제출료량보과려전지산법래가쾌동태망락배경하적국부개수중심도경신과정。접착대산법진행료이론분석，병분석료산법적응용장합。최후통과재모의화진실수거집상적실험，탐토료최장로경약속 K 적대소선택，병전시출문중소제적전동태다유망락국부개수중심도경신산법구유량호적가속성능。
With the growing demands of the complex network analysis,more and more people begin to concentrate on analyzing complex networks.New analysis methods are needed when facing with more complex multi-dimensional social networks or entity relations which are changing frequently.This paper introduces a method to measure the importance of nodes on fully dynamic multi-dimensional networks.We propose a metric which is called local betweenness centrality and its updating algorithm to solve this problem.Firstly,we transform the multi-dimensional networks into single-dimensional networks by corresponding mapping function. Secondly,we introduce a local betweenness centrality metric with K longest distance constraint. Thirdly,the two-step pruning updating algorithm is introduced.We provide the theoretical analysis of this algorithm.Moreover,we make experiments on both synthetic and real multi-dimensional networks to illustrate its impressive efficiency.