This paper applied the hypergraph based model and cluster algorithm in community structure discovery. 摘要将超图模型以及基于此的聚类算法应用到社区结构发现的领域。
Based on the clustering method, the mixing matrix was first estimated by a new potential function using the sparseness of sources. 首先，利用变换后信源的稀疏特征，采用一新的势函数通过聚类算法估计混叠矩阵。
