DATA WAREHOUSE METHODOLOGIES

Data Warehouse Methodologies

Introduction

With the explosive growth experienced in the past few years, rising need for data security and management, data warehousing, as a form of storing current as well as historic data, has played a key role in the integration process. A data warehouse is a database for analysis of data and data reporting. It involves the integration of data from one or more sources, creating a central repository of data.it stores both current and historical data for use in creating reports for senior management. In support of the growing market; a number of data warehousing methodologies have been developed, having been cited as a priority project by IT experts. This paper seeks to compare various methodologies on the basis of some common attributes.

Data warehouse methodologies have a set of general tasks which include; first, business requirement analysis ,such as interviews and brainstorming, used to elicit business queries, that are analytic questions that managers pose, after which they are prioritized by estimating the risk attached to the answers to the questions. Second, data design which includes data modeling techniques such as dimensional modeling and entity relational modeling. Architecture design, that involves; creating conceptual models, which serves as a blueprint for the data requirements of the firm. It allows planning, maintenance, learning and reuse. The iterative approach is recommended as it is data driven. In this approach, Data is first gathered, then integrated and finally tested, with programs written against it followed by an analysis of the programs.

Methodologies are bases on some common attributes. These include: first, the core competency attribute of the companies, which depends on the segment they are in .for instance, the core-technology vendors that sell database engines use data warehousing schemes that take advantage of the gradations of their database engines. The infrastructure vendors, on the other hand, deal in the warehouse infrastructure tools which manage metadata by help of repositories that aid in extracting, transferring, loading data or creating end-user solutions.

Second, are the requirements modeling an attribute that concentrates on methods of capturing business requirements and developing information models on the basis of these requirements? Third, is the data modeling attribute that focuses on data modeling techniques used to develop physical and logical models. Fourth, is the architecture design attribute, whose design is affected by the designing strategies available such as the enterprise wide design the data mart design, all left for the firm to decide? Fifth, is the implementation strategy attribute, whose methodology varies between system development lifecycle approach and a RAD approach, with the iterative prototyping approach being more preferred?

In addition, the metadata management attribute, is an important aspect in data warehousing, as most vendors focus on metadata management. Among other attributes, is the scalability attribute, that is highly dependent on the type of database management system used, which can also be achieved by increasing the disk space that would in turn increase the firm’s overheads.

Conclusion

In conclusion, the development management attribute is of vital importance. In today’s dynamic economy, where mergers and acquisitions are in the rise, there is, therefore, need for constant re-scoping, restructuring, re-planning of priorities and redefinition of business objectives to accommodate changes in the data warehouse. Advances in technology and divestiture are other sources of change. Therefore, change management is crucial in the choice of data warehousing methodology.

References;

Johnson, A., & McGinnis, L. (2011). Performance Measurement in the Warehousing Industry. IIE Transactions, 43(3), 220-230.

Sen, A., & Sinha, A. P. (2005). A Comparison of Data Warehousing Methodologies. Communications of the ACM, 48(3), 79-84.