SAS Banking Intelligence 方案讲述说明.ppt

分段存储区（Staging Area） 各个业务系统的数据首先直接快速传输到分段存储区 再从分段存储区经过清洗、转换、映射等复杂的数据移动处理转移到目标数据仓库中 ODS（Operational Data Store） 所有数据进入系统之前，首先进入ODS，减少对业务系统的影响 在ODS当中主要完成的工作是错误数据的修改和缺少数据的补录 基础数据仓库（Baseline） 数据结构完全按照决策支持的需求而设计 在基础数据仓库中，存储着详细的信用卡业务数据 数据集市（Data Mart） 数据集市是数据仓库的子集，通常已经将详细数据聚合为汇总数据 数据结构设计为星型结构，可以直接支持通过ROLAP服务器的多维分析 Explanation for this slide: There are a lot of pre-built elements that have significant advantages to a customer. First, all these pre-built elements are all derived from specific business pains. Second, by having these pre-defined a customer will save a lot of implementation time. This is a section of the telecommunications intelligence model. The models comprise a logical model describing the way a telecommunications company manages their business and the flow of data around it- that‘s what you see here. That logical model is then mirrored in a physical model which charts where the data is physically located. Analytical models provide the ?formula“ for generating intelligence - for example, to determine which customers are likely to churn the elements might include outstanding bill balance, period as a customer and arrears history. These models can be changed and extended without the need to start again from scratch – this is how the customisability and extendability of the solutions is ensured. Taking the archtitectural analogy once more, you could view the applications as individual houses, for which the architectural drawings are pre-supplied. However, the models also provide the optimal town plan with roads and utility supply lines (electricity and water) already marked in, so that when you choose to build a new house, you have both the architectural drawings for it, and all of the specifications to ensure it will meet town planning requirements. The SAS Management Console, as a standardized, common management tool, enables the enterprise to support technologies across the information value chain from a single point of administration. By levera