图 书 馆
自分裂框架:从非重叠到重叠集群Ego-splitting Framework: from Non-Overlapping to Overlapping Clusters |
|
| 课程网址: | http://videolectures.net/kdd2017_epasto_ego_splitting_framework/ |
| 主讲教师: | Alessandro Epasto |
| 开课单位: | 视频讲座网 |
| 开课时间: | 2017-10-09 |
| 课程语种: | 英语 |
| 中文简介: |  |
| 课程简介: | We propose a new framework called Ego-splitting for detecting clusters in complex networks which leverage the local structures known as ego-nets (i.e. the subgraph induced by the neighborhood of each node) to de-couple overlapping clusters. Ego-splitting is highly scalable and flexible framework, with provable theoretical guarantees, that reduce the complex overlapping clustering problem to a simpler and more amenable non-overlapping (partitioning) problem. We cann solve community detection in graphs with tents of billions of edges and outperform previous solutions based on ego-nets analysis. More precisely, our framework works in two steps: a local ego- net analysis, and a global graph partitioning. In the local step, we first partition the nodes’ ego-nets using non-overlapping clustering. We then use these clusters to split each node of the graph into its persona nodes that represents the instantiation of the node in its communities. Then, in the global step, we partition these new persona nodes to obtain an overlapping clustering of the original graph. |
| 关 键 词: | 自我分裂; 复杂网络; 局部结构; 重叠聚类 |
| 课程来源: | 视频讲座网 |
| 数据采集: | 2022-11-20：chenxin01 |
| 最后编审: | 2022-11-20：chenxin01 |
| 阅读次数: | 24 |