摘要
近年来,高质量社区的挖掘和发现已经成为社会网络研究一个热点.其中,基于标签传播的社区挖掘算法(Label Propagation Algorithm,简称LPA)由于具有近似线性时间复杂度且无须预先定义目标函数和社区数量等优点而得到广泛关注.但是,LPA算法的标签传播过程存在不确定性和随机性,影响了社区发现的准确性和稳定性.提出一种新的基于标签传播的社区发现算法LPA_SI(Label Propagation Algorithm based on Significance and Influence).首先,采用新的节点重要性度量方法对节点进行排序;其次,提出一种新的标签影响力计算方法更新每个节点的标签;最后,在真实数据集和人工数据集上的实验表明,LPA_SI在复杂度相近的情况下能够显著提高社区发现的质量,并具有较好的稳定性.
In recent years, mining and detecting of high quality communities has become a hot orientation in social network research. Among them,the community detection algorithms based on label propagation （ LPA ） receive broad attention for the advantages of near-linear complexity and no prerequisite for any object function or cluster number. However, the propagation of labels contains un- certainty and randomness, which affects the accuracy and stability of the LPA algorithms. In this paper, a novel community detection algorithm based on LPA called LPA_SI { LPA based on Significance and Influence ） is proposed. Firstly,the vertices are sorted accord- ing to a new vertex importance measure. Then, the label of each vertex is updated according to a new label influence measure. The ex- periments on both the artificial datasets and the real-world datasets demonstrate that the LPA_SI algorithm significantly improves the community quality while the detection stability is preserved.
出处
《小型微型计算机系统》
CSCD
北大核心
2015年第6期1171-1175,共5页
Journal of Chinese Computer Systems
基金
国家自然科学基金项目(61103175
61300104)资助
教育部科学技术研究重点项目(212086)资助
福建省科技创新平台建设项目(2009J1007)资助
福建省自然科学基金项目(2013J01230)资助
福建省高校杰出青年科学基金项目(JA12016)资助
福建省高等学校新世纪优秀人才支持计划项目(JA13021)资助
关键词
社会网络
社区发现
标签传播
标签影响力
节点重要性
social network
community detection
label propagation
label influence
node significance
