Cc-bitmaps: An Effective Index Technology Of The Closed Cube

Data warehouse is generated accompanied by the development of information and decision support system. With the development of data warehouse growing scale, the effective mass data storage, fast query and analysis capabilities are more demanding.Online analysis processing by pre-aggregated the mass data to generate and save the data cube (also called multi-dimensional data), provides the ability to fast query. However, the complete pre-computation has led to the explosion of data volume.Therefore, the effective OLAP data cube index technology for reducing the disk storage space and improving the query performance is requested to study to solve these problems.Quotient Cube is a classical lossless data cube storage structure, saving the data cube semantic information while not solving the problem of incremental update.QC-Trees constructs a semantic query tree based on Quotient Cube, not only preserves the original semantic information, but also further reduces the storage. Meanwhile, it provides effective query and incremental update. Closed Cube studied on Quotient Cube further streamlines the cube storage structure while preserving the original drill up and down semantics.However, the experiment displayed the efficiency of query and incremental update algorithms are not high.This paper proposes an effective OLAP index structure CC-Bitmaps based on in-depth study of the characteristics of Closed Cube and QC-Trees. CC-Bitmaps is based on Closed Cube combining the ideas of prefix and suffix sharing and bit coding ideas to further compress the needs of disk storage space, and to provide more effective query performance than QC-Trees, and also to provide effective incremental update.This paper also proposes a new lists intersection algorithm CCListsIntersection, which can gradually reduce range in the process to find the data when intersection by each dimension value, providing a faster query response capacity in comparison with the tradictional intersection algorithm.Finally, the implementation of CC-Bitmaps, QC-Trees and other algorithms and a detailed experimental comparison from the storage space, query performance is presented to confirm the effectiveness of CC-Bitmaps data cube index structure and navigate the next step work.