Separating Data into Data Marts - Research Paper Example

Separating Data into Data Marts Introduction The research begins by giving a brief of a data mart. It mainly concentrates on various recommendations suggested in order to come up with an efficient data mart.Data mart A data mart is a decision support system. It is a subset of the company’s data and activities. Data marts often have business related purposes, such as, measuring the impact of marketing promotions and forecasting sales performance. Recommendations The driving business tool for the data mart is the need for useful information.

It is necessary to understand the independent business needs of a business environment. This can only be achieved by involving those who have a share in the data mart. This includes both business sponsors and end users. Together, they should agree on the information needs that the data mart should fulfill; the data sources, the technical requirements and the success criteria. This happens in the early stages of the design.Industry and Corporate Factors It is necessary to develop a plan for delivery before implementing a data mart.

This entails information requirements and prioritizing the user’s needs. It is also necessary to develop a corporate philosophy that embraces true value on an optimized environment. Future Growth Strategies and Scalability For future growth, data needs to be integrated and managed consistently from across an organization. For instance, a financial data mart needs to provide equal access to common financial data that would be the foundation for corporate performance management. Other areas would include data marts to assist in decision making while screening the complexities associated with access to top corporate data structure.

Corporate experts should assist in organizing the data marts. End users should have direct access to the information they need to deliver relevant business decisions.Future design and the structure of data marts Future data mart designs should incorporate the process of logical design. This will involve arranging data into a set of logical relationships called entities and attributes. This is particularly crucial in keeping the business requirements on hand throughout the design process. It is necessary to make a wish list of data elements from the business’ user needs.

This involves setting up a list of data elements, data attributes, grouping data reasonably and setting up relationships between data. It is essential to understand the source of data; this can be done by categorizing them by size (Kimball & Ross, 2002). Classification of data involves traditional presentation of facts, dimensions, and relationships between them in data mart applications. The figure below is a complete description of energy consumption per person type for floors in a building.

The benchmarks can be created over floors. It draws a clear picture of energy consumption and identifies some of the efficiency opportunities.Data mart basic structure illustrationRecommendation on Data mart implementation It is paramount to consider the following steps in the implementation of a data mart:Design the schemaCreate a solid storagePopulate the data mart with records from source systemsAccess it to make informed decisionsManagement over time.Supporting Evidence on Recommendation Designing the data schema would help in gathering the various business requirements, identifying the data sources and designing the logical and physical configuration of the data mart.

The creation of a physical storage step would allow for the establishment of the best way of setting up tables and the access structure. The accessing step would allow queries submitted, through the front-end tool, to execute quickly and in an efficient manner (Scheps, 2008). The management step would help in managing the growth of data and optimization of the system for better performance. It would also help in secure access of data. ReferencesKimball, R. & Ross, M. (2002). The data warehouse toolkit: The complete guide to dimensional modeling. 2nd ed. Hoboken, NJ: Wiley.Scheps, S. (2008).

Business intelligence for dummies. Hoboken, NJ: For Dummies.