| 摘要: |
| 数据仓库的一个重要用途是利用时态视图向用户提供历史信息.因为在传统关系数据模型中增加了对时间的支持,而且时态视图的更新不仅来自于基表更新,还包括时间前进,所以,目前对非时态视图维护的研究成果不适用于时态视图,并且已有的一些时态视图维护算法也不适用于数据仓库.以历史关系模式为对象,根据增量式维护方法的原理,采用纯删除、纯插入的计算方法,用代数语言给出了5种基本历史关系代数运算的更新传播算法,由这5种历史关系代数组合定义的时态视图都可用迭代方法得到其增量维护计算式.所采用的纯删除、纯插入思想也可移用于其他历史 |
| 关键词: 数据仓库 时态操作 视图维护 历史关系模式 增量式计算 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| Maintaining Temporal Views in Data Warehouses |
|
LI Qi,BAI Ying-cai
|
| Abstract: |
| An important use of data warehouses is to provide historical information by temporal views. Because it makes temporal extension to traditional relation data model, the materialized temporal views may be needed updated not only when base relations change, but also as time advances. So presently the research results for the non-temporal views are not applicable for the temporal views, and also the existed temporal view maintaining techniques are not appropriate for the data warehouses. Based on a historical relational model,according to the pure deletion and pure insertion concept,the change propagation are presented in this paper computation in relational algebra format for the five prve primitive historical relational operators.The incremental computaion of temporal views defined by the five operators can be obtained bu iterative method.At the same time,the pure insertion and the pure deletion con caaepted by this paper can be used in the temporal view maintence under other historical relational model. |
| Key words: data warehouse temporal operation view maintenance historical relational schema incremental computation |
