摘要： 现有的社交网络快速划分社区算法存在质量低、不能充分利用节点链接信息的问题,而效果较好的划分算法也存在时间复杂度高、无法应用于大规模社交网络的问题。为此,提出一种基于MapReduce 的社区发现算法。利用PGP 算法内信任者推荐模型迭代计算用户之间的信任强度,通过社区传播的方式聚合节点。在经典数据集上和大规模新浪微博数据集上进行实验,结果表明,该算法能有效度量用户间的信任度,得到准确的社区发现结果。
关键词:
社交网络,
社区发现,
信任度,
并行化,
信任推荐,
微博
Abstract: In the research area of community detection in social network,there exists problem that some fast algorithms
for large scale network are resulting in low quality community results,and lacking of model and mechanism to express user link attributes,some algorithms for with comparatively satisfactory detection result having high time complexity. This paper proposes a community detection algorithm for massive-scale social networks using MapReduce. This paper uses a new recommend trust model,which is evolved from PGP(Pretty Good Privacy),to compute the trust degree between users iteratively. More importantly, it proposes a community propagation model to assign nodes into communities. Finally,it conducts experiments with some typical network datasets and Sina microblog datasets,which shows the model this paper proposed can availably compute the trust degree between users,and a better result of community detection is gained.
Key words:
social network,
community detection,
trust degree,
parallel,
trust recommendation,
microblog
中图分类号: