| On-Line Analysis Processing (OLAP) is a data analysis technology which appearing with the Data Warehouse (DW) technology, it can make use of systemic data in DW and help people build data snapshots quickly from many different point of views, these snapshots reflect the action characteristic of the enterprise and can be understood by users. With the development of computer application technology, new data format appears continually, and this has put forward a new requirement for the efficiency and applicability of the OLAP tool. The existing OLAP tool can not deal with many kinds of data source effectively, and can not establish many data storage modes according to user's requirement. This paper provides SEU_OLAP system which can effectively deal with several data source, such as Data Warehouse, OLTP database and multidimensional database, and can build many data storage modes according to user's requirement, it can also hide the details of data modes from user's point of view. The system provides a friendly interface and can display multidimensional data in a graphical way. This paper mainly includes the following content:1) Describing the architecture of OLAP system, and expatiating the architecture of the OLAP modeling tools and OLAP front-End Tools.2) The difference of physical schema is hidden from user's point of view through semantic objects , and users can manipulate their familiar semantic objects to obtain analytical results.3) Several key technologies which affect OLAP query efficiency are discussed, and a materialized view selection algorithm is presented. Materialized View Selection Problem (VSP) is the main problem in implementing the OLAP system. After studying those existing VSP algorithms and analyzing their characteristics, a simple and easy implemented VSP algorithm that can run efficiently is introduced.4) Design and implement of multidimensional data visualization in SEU_OLAP system with DirectX technology. |
PDF Full Text Request