基于P-Rank的RDF有向图的分布式存储  

Distributed Storage of RDF Directed Graph Based on P-Rank Algorithm

作　　者：冷泳林[1] 申华[2] 鲁富宇[3] 

机构地区：[1]渤海大学信息科学与技术学院,辽宁锦州121000 [2]鞍山师范学院数学与信息科学学院,辽宁鞍山114005 [3]渤海大学教务处,辽宁锦州121000

出　　处：《重庆理工大学学报（自然科学）》2015年第1期91-95,共5页Journal of Chongqing University of Technology：Natural Science

基　　金：辽宁省社科基金资助项目(L14AGL002)

摘　　要：随着语义网数据的迅猛增长,RDF大数据存储成为数据存储领域研究的热点问题,而分布式存储是解决RDF大数据可扩展性的一种有效途径,数据分割则是实现分布式存储的关键。利用图聚类思想实现RDF数据的有效分割,基于RDF数据模型的图特性,首先利用PRank节点相似度算法计算RDF图结点间的相似度,然后使用AP聚类算法对相似度矩阵进行聚类,进而实现RDF数据的分割和分布式存储。实验结果表明:该方法能有效完成RDF数据的聚类分割,使得聚类结果中类间相似度较小,而类内相似度较大。With the rapid growth of the semantic web data, RDF （Resource Description Framework） storage has become a hot research issue in data storage field currently. Distributed storage is an effective way to solve the scalability of RDF big data, and data partition is a key issue for distributed storage. We used graph clustering idea to realize the effective partition of RDF data. RDF can be de- scribed as a directed graph, so in this paper we used P-Rank （Penetrating Rank） algorithm to calculate the similarity of RDF graph nodes, and then the AP clustering algorithm was implemented to cluster the similarity results, so as to realize the distributed storage of RDF data. The experimental results show that, this method can complete the RDF data partition effectively, and can make the intra-cluster similarity smaller, and the similarity within the inter-cluster is larger.

关 键 词：资源描述框架 有向图 P-RANK AP聚类 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]