一种优化的标签传播方法

社区发现是复杂网络分析领域的一项重要研究内容,而标签传播算法因在分析复杂网络时具有时间复杂度低等优点,获得广泛关注.但标签传播算法中的随机策略降低社区划分结果的稳定性和效率.为解决随机策略引起的问题,提出了一种优化的标签传播方法.该方法引入标签权重,并与标签一起组成二元组,根据标签二元组、节点间的联系度等因素为节点分配初始化标签;同时,在标签传播过程中,根据节点间的联系度等因素进行标签更新.实验结果证明了该方法的有效性和有用性.

Optimized Label Propagation Solution

Community discovery is one of the important research topics of complex network analysis.And label propagation algorithm has attracted much attentions due to the low time complexity exhibited when discovering communities in complex networks.However,the random strategies adopted in the algorithm degrade the stability and efficiency of community division results.In order to handle the problem caused by the random strategies,an optimized label propagation solution is proposed.The solution introduces label weights which forms a tuple with the corresponding labels,assigns initial labels to nodes based on label tuples and the connection degree between nodes.The connection degree between two nodes is also exploited during label propagation.The experimental results show the effectiveness and usefulness of the proposed solution.