欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对异质社交网络社区识别问题,提出一种基于随机游走层次社区识别算法。提出异质网络层级吸引力度量函数,构建异质网络随机游走模型;设计了一种基于2-hop互随机游走的异质网络节点相似性度量函数;通过将该相似性函数推广到层次聚类并设计相应的相似矩阵校准方案,异质社区识别任务可以在较短的时间内迭代完成。人工合成网络和真实网络上的仿真实验验证了算法的可行性和有效性。
英文摘要：
In order to solve the problem of identifying community structures in heterogeneous social networks, a hierarchical community detection algorithm was proposed based on random walks. A heterogeneous random walk model was built by measuring the attraction between network layers and the transition probability between nodes in homogeneous networks. Then, a heterogeneous network node similarity function was proposed based on 2-hop mutual random walks. Finally, the similarity function was extended to hierarchical clustering so the multi-relational community structure could be obtained iteratively in a relatively short period of time. Competitive experiments on both synthesized and real-world social networks demonstrate the effectiveness and feasibility of the proposed algorithm.
同期刊论文项目
同项目期刊论文
期刊信息
位置：