CAJ | 학술논문

提出了基于节点吸引力的分裂社区发现算法，算法认为网络中节点之间是存在一定吸引力的，不同相连节点之间的吸引力大小差异很大。如果相连两个节点的度数越相近，那么这两个节点相互吸引力越小；否则认为度数较大的节点对度数较小的节点的吸引力大。在本算法中，通过计算网络中节点之间的吸引力，删除节点间吸引力最小的连边，重复计算吸引力，并删除相应的边，实现网络的分裂，通过计算网络分裂后社区的中心性，实现社区的发掘。实验证明该算法可以有效划分网络社区。
제출료기우절점흡인력적분렬사구발현산법，산법인위망락중절점지간시존재일정흡인력적，불동상련절점지간적흡인력대소차이흔대。여과상련량개절점적도수월상근，나요저량개절점상호흡인력월소；부칙인위도수교대적절점대도수교소적절점적흡인력대。재본산법중，통과계산망락중절점지간적흡인력，산제절점간흡인력최소적련변，중복계산흡인력，병산제상응적변，실현망락적분렬，통과계산망락분렬후사구적중심성，실현사구적발굴。실험증명해산법가이유효화분망락사구。
This paper presents a community discovery algorithm based on nodes attraction. This algorithm consid-ers that there is some certain attraction between nodes in the network, there exist big differences between the at-tractive sizes when connected to different nodes:The closer the degrees of connected to nodes is, the smaller of the attraction, otherwise the bigger the degree of the node is, the stronger the attraction it becomes as opposed to the smaller degree node. By calculating the attraction between network nodes, the algorithm adopted in this paper deletes the side which is the least attraction between two nodes, and then repeats calculating the attraction of con-nected to nodes, and delete the side, to realize the division of the network, through the calculation in the network community centricity, at least to realize discovery of communities. Experiments prove that the algorithm can effec-tively divide different network communities.