Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

Share this Page URL
Help

Chapter 8. Data Warehouse Modeling Techn... > Base Properties of a Data Warehouse - Pg. 83

Figure 37. Base Properties of a Data Warehouse. A data warehouse is an integrated collection of databases rather than a single database. It should be conceived as the single source of information for all decision support processing and all informational applications throughout the organization. A data warehouse is an organic thing, and it tends to become big, if not big from the beginning. In addition to the obvious requirement that a data warehouse should satisfy the needs of end users, there is also a great need to achieve maximum consistency throughout the whole data warehouse environment, at the level of primitive data and derived data, and also within the information derivation processes themselves. A data warehouse contains data that belongs to different information subject areas, which can be the basis for logically partitioning the data warehouse in several different (conceptual or even physical) databases. A data warehouse also contains different categories of data. It contains primitive data (the System of Record) either represented and organized as an accumulation of captured source data changes, business events and transactions, or as an interpreted and well-structured historical database. In many cases both representations of primitive data are present in the data warehouse and are positioned and mapped to form an integrated collection of data that represents the corporate memory. Another major category of data in the data warehouse is that which is condensed and aggregated in information analysis databases having a format and layout that is directly suitable for end users to interpret and use. A data warehouse also usually contains support databases, which are not directly of interest to end users for their data analysis activities but are important components in the process of capturing source data and delivering consistent information to end users. Clearly, data warehouse modeling must consist of different kinds of modeling techniques. The System of Record is usually best if not modeled using the same modeling techniques as the end-user-oriented information analysis databases. If, in addition, one considers that end users may be dealing with decision support tools (query and reporting, OLAP, data mining, ...) and informational applications that have different usage and development characteristics, it becomes clear that data warehouse modeling is in fact a compilation of different modeling techniques, each with its own area of applicability. Chapter 8. Data Warehouse Modeling Techniques 83