Market Forces (Jul 2006)
Derivations of Initial Data Warehouse Structure by Mapping Operational Database on Transaction Patterns
Abstract
Data warehouses improve the quality of integrated information in the organization for decision-making.The data for the data warehouse comes from online transaction systems. Typically, an involved process ofanalysis precedes the actual design phase of a data warehouse (1999). The analysis process becomes moredifficult because of the costs involved in hiring experienced staff and the privacy issues arising from theuse of external consultants (2000). Peter Coad’s transaction pattern (2000) is a higher-level description ofa generic business process (or a template) that has helped in the analysis and design of a wide range ofbusiness domains. The thesis of this paper is that the mapping of the operational databases on thetransaction pattern facilitates the derivation of initial data warehouse structure. During the mappingprocess, the relationships, roles and attributes of the players defined by the transaction pattern help us inidentifying the instances of the pattern in the database. Through these instances, we can then derive theinitial data warehouse structure i.e. the attributes of the fact and dimension table(s). The data warehousestructure thus derived reduces the need for an extensive information analysis of the needs ofthe user and the dependency on the experienced personnel for data warehousedevelopment. This paper proposes a three-step derivation methodology that is illustratedusing a case study of an organization’s operational database.
