计算机应用研究2018，Vol.35Issue(2)：333-338,6.DOI:10.3969/j.issn.1001-3695.2018.02.003
基于改进Eclat算法的资源池节点异常模式挖掘
Mining anomaly pattern of nodes in resource pool based on improved Eclat algorithm
摘要
Abstract
In cloud computing,mining anomaly pattern of nodes plays an important role in promptly diagnosing node states of resource pool.According to the data characters of computing resource,storage resource and network resource,it preprocessed the information of node status.And it used association rule algorithm to mine the rules among the nodes' parameter status of resource pool,e.g.," high level-high level" and "low level-high level".This paper proposed an improved Eclat algorithm,called i-Eclat,which based on vertical data format of restrictive attribute connection.i-Eclat reduced the number of connections by transforming the format of state data and constructing a non-frequent 2-itemsets,as well as it constructed the storage structure to limit the redundant attribute connections.Extensive experiments have been conducted to demonstrate that this method can find the hiding rules among resource pool's nodes.And it also shows that i-Eclat outperforms traditional methods on computing efficiency,especially on processing big data sets.关键词
模式异常挖掘/关联规则/资源池/i-Eclat算法/云计算Key words
anomaly pattern mining/association rule/resource pool/i-Eclat algorithm/cloud computing分类
信息技术与安全科学引用本文复制引用
高强,张凤荔,陈学勤,王馨云,耿贞伟,周帆..基于改进Eclat算法的资源池节点异常模式挖掘[J].计算机应用研究,2018,35(2):333-338,6.基金项目
国家自然科学基金资助项目(61602097,61272527) （61602097,61272527）
四川省科技支撑计划资助项目(2016GZ0065,2016GZ0063) （2016GZ0065,2016GZ0063）
中央高校基本科研业务费资助项目(ZYGX2015J072) （ZYGX2015J072）
