FOLAPE：基于位向量索引的快速OLAP引擎研究  

FOLAPE: fast OLAP engine research based on bit vector index

作　　者：牟昕[1] 唐常杰[1] 左劼[1] 段磊[1] 巩杰[1] 姜页希[1] 朱军 

机构地区：[1]四川大学计算机学院,成都610065 [2]中国出生缺陷检测中心,成都610041

出　　处：《四川大学学报（自然科学版）》2010年第5期1025-1032,共8页Journal of Sichuan University(Natural Science Edition)

基　　金：国家自然科学基金(60773169);"十一五"国家科技支撑计划项目(2006BA105A01)

摘　　要：提高联机分析处理OLAP效率是学术界和工业界共同关注的课题.传统OLAP在处理复杂应用时,运算代价随维度增大而增大,为解决这一问题,做了如下工作:(1)分析现有OLAP模型的不足,设计了一种采用位向量索引技术的快速OLAP模型,FOLAPE;(2)FOLAPE支持索引的增量式更新,提高了数据更新效率;(3)针对FOLAPE特点,设计了新的聚集函数算法;(4)在真实数据上进行了实验验证.实验表明,FOLAPE相对于最新OLAP模型,其平均查询时间最低减少90%,最高减少95%,同时存储空间需求也显著降低.On-Line Analytical Processing （OLAP） is an important topic focused by research community and industry. In complex applications, the costs of traditional OLAP fast increase with the increment of dimensions. To resolve this problem, the authors made the following contributions： （1） Analyze the inadequacies of traditional OLAP models, and propose a new model named as Fast OLAP Engine （FOLAPE）. （2） FOLAPE support incremental updates without reconstruction of index, hence it is efficient and scalable. （3） Design new aggregation functions based on traditional aggregation strategies and the characteristics of FOLAPE. （4） Conduct extensive experiments on real datasets and comparison with the latest OLAP model shows that the average query time is decreased by 90% at least and 95% at most, while the storage space is greatly decreased in the mean time.

关 键 词：联机分析处理 位向量索引 聚集 LRU 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]