纯Peer to Peer环境下有效的Top-k查询
发布时间：2019-03-05 11:24
【摘要】：目前大多数的Peer-to-Peer(P2P)系统只支持基于文件标识的搜索,用户不能根据文件的内容进行搜索.Top-k查询被广泛地应用于搜索引擎中,获得了巨大的成功.可是,由于P2P系统是一个动态的、分散的系统,在纯的P2P环境下进行top-k查询是具有挑战性的.提出了一种基于直方图的分层top-k查询算法.首先,采用层次化的方法实现分布式的top-k查询,将结果的合并和排序分散到P2P网络中的各个节点上,充分利用了网络中的资源.其次,根据节点返回的结果为节点构建直方图,利用直方图估计节点可能的分数上限,对节点进行选择,提高了查询效率.实验证明,top-k查询提高了查询效果,而直方图则提高了查询效率.
[Abstract]:At present, most Peer-to-Peer (P2P) systems only support search based on file identification, and users can not search according to the contents of files. Top-k query has been widely used in search engines and has achieved great success. However, because P2P system is a dynamic and decentralized system, it is challenging to carry out top-k query in pure P2P environment. A hierarchical top-k query algorithm based on histogram is proposed. Firstly, a hierarchical approach is used to implement distributed top-k query, and the merging and sorting of the results are dispersed to each node in P2P network, which makes full use of the resources in the network. Secondly, the histogram is constructed for the node according to the result returned by the node, and the upper limit of the possible score of the node is estimated by using the histogram to select the node, which improves the efficiency of the query. Experimental results show that top-k query improves the query efficiency, while histogram improves the query efficiency.
【作者单位】： 中国人民大学信息学院 中国人民大学信息学院 中国人民大学信息学院
【基金】：国家自然科学基金 国家高技术研究发展计划(863) 国家重点基础研究发展规划(973) 北京市科技计划重大项目~~
【分类号】：TP393.09
[Abstract]:At present, most Peer-to-Peer (P2P) systems only support search based on file identification, and users can not search according to the contents of files. Top-k query has been widely used in search engines and has achieved great success. However, because P2P system is a dynamic and decentralized system, it is challenging to carry out top-k query in pure P2P environment. A hierarchical top-k query algorithm based on histogram is proposed. Firstly, a hierarchical approach is used to implement distributed top-k query, and the merging and sorting of the results are dispersed to each node in P2P network, which makes full use of the resources in the network. Secondly, the histogram is constructed for the node according to the result returned by the node, and the upper limit of the possible score of the node is estimated by using the histogram to select the node, which improves the efficiency of the query. Experimental results show that top-k query improves the query efficiency, while histogram improves the query efficiency.
【作者单位】： 中国人民大学信息学院 中国人民大学信息学院 中国人民大学信息学院
【基金】：国家自然科学基金 国家高技术研究发展计划(863) 国家重点基础研究发展规划(973) 北京市科技计划重大项目~~
【分类号】：TP393.09
【参考文献】
相关期刊论文 前1条
1 黄维雄,黄铭钧,陈建利,王晓宇,凌波,周傲英;一种基于自配置策略的新型Peer to Peer平台系统[J];软件学报;2003年02期
【共引文献】
相关期刊论文 前10条
1 蔡铭;孙乐昌;潘祖烈;姜欣;李永祥;;利用智能疫苗防治多网络蠕虫的方法[J];北京邮电大学学报;2007年02期
2 茅洪波,于万波,魏小鹏,姜丽丽;BDI agent导引修改机制用于Peer to Peer系统文件操作[J];大连理工大学学报;2003年S1期
3 聂荣;张洪欣;吕英华;雷振明;;P2P网络的研究与进展(下)[J];电信科学;2008年04期
4 姜灵;李绍滋;;一种混合式P2P文件共享与检索模型[J];福建电脑;2007年03期
5 唐九阳;葛斌;张，
本文编号：2434852
本文链接：https://www.wllwen.com/kejilunwen/sousuoyinqinglunwen/2434852.html
