实际复杂网络数据的合理预处理是社团结构分析的基础与前提,对划分结果有着显著的影响.本文通过定义节点间的关系强度,将社会调查得到的社交网络数据处理成社团结构算法中通用的邻接矩阵,提出了处理实际复杂网络数据使之适用于经典社团算法分析的一种方法.同时,本文进一步给出了关系强度强联系定义和关系强度弱联系定义,并比较了算法在这两种定义下展示出的不同性能.
The data pre-processing of actual complex network is an important basis of researching community structure detectingalgorithm, which has a significant influence on the accuracy of algorithm. A new method of process is proposed to transmit theoriginal network data into standard adjacent matrix through the relationship strength definition. Several classical community detectingalgorithms under the definition are analyzed.