天津科技大学学报
天津科技大學學報
천진과기대학학보
JOURNAL OF TIANJIN UNIVERSITY OF SCIENCE & TECHNOLOGY
2015年
4期
70-72,77
,共4页
熊聪聪%庞朝辉%王兰婷%耿世洁
熊聰聰%龐朝輝%王蘭婷%耿世潔
웅총총%방조휘%왕란정%경세길
云平台%案例检索%MapReduce%最近邻法%改进
雲平檯%案例檢索%MapReduce%最近鄰法%改進
운평태%안례검색%MapReduce%최근린법%개진
cloud platform%case retrieval%MapReduce%KNN%improvement
针对大数据处理需求提出基于云平台的案例检索算法.利用MapReduce技术改进了案例检索算法中常用的最近邻法,使其能够在多个服务器节点上并行执行,从而提高在海量数据情形下的案例检索速度.实验表明:基于云平台的案例检索速度高于单节点检索,集群节点的数量对案例检索有一定的影响.
針對大數據處理需求提齣基于雲平檯的案例檢索算法.利用MapReduce技術改進瞭案例檢索算法中常用的最近鄰法,使其能夠在多箇服務器節點上併行執行,從而提高在海量數據情形下的案例檢索速度.實驗錶明:基于雲平檯的案例檢索速度高于單節點檢索,集群節點的數量對案例檢索有一定的影響.
침대대수거처리수구제출기우운평태적안례검색산법.이용MapReduce기술개진료안례검색산법중상용적최근린법,사기능구재다개복무기절점상병행집행,종이제고재해량수거정형하적안례검색속도.실험표명:기우운평태적안례검색속도고우단절점검색,집군절점적수량대안례검색유일정적영향.
A new case retrieval technique based on cloud platform was presented in order to process large amount of data. The KNN used in case retrieval was improved by using MapReduce technology,so that it can be executed in parallel across multiple server nodes and improve the speed of case retrieval among large amount of data. Experiments show that the case-based reasoning based on cloud platform is faster than that of the normal mode and the node number of the cluster has a cer-tain effect on case-based reasoning.