摘要
针对基于种子扩展的重叠社区检测算法存在因种子选取质量不高而导致重叠社区检测结果准确度较低的问题,提出一种利用图嵌入、聚类和K-shell相结合的新的种子选取策略来进行种子扩展的重叠社区检测算法。算法利用提出的新的种子选取策略得到种子集,根据社区度量函数即电导性最优的原则不断进行种子扩展完成社区划分。研究结果表明,改进的种子扩展的重叠社区检测算法提高了检测结果的准确度。
For the overlapping community detection algorithm based on seed extension,the accuracy of overlapping community detection results is not high due to the low seed selection quality,a new overlapping algorithm was proposed,namely GCSE,which uses graph embedding,clustering and K-shell for seed expansion.In this Algorithm,the new seed selection strategy was used to get the seed set,and then the community division was completed by continuously expanding the seed according to the principle of the best conductivity.The results show that the improved seed-expanding overlapping community detection algorithm is proposed to improve the accuracy of detection results.
作者
段瑞玮
张公敬
DUAN Rui-wei;ZHANG Gong-jing(College of Computer Science and Technology, Qingdao University, Qingdao 266071, China)
出处
《青岛大学学报(自然科学版)》
CAS
2021年第1期64-69,共6页
Journal of Qingdao University(Natural Science Edition)
基金
国家自然科学基金(批准号:61872205)资助。
关键词
重叠社区
图嵌入
聚类
K-SHELL
种子扩展
社区检测
overlapping community
graph embedding
clustering
K-shell
seed expansion
community detection