Data Warehousing Systems Report

Introduction

In today’s world, data mining is a significant business process that is used to get valuable and important information. This paper will focus on technical part of a business plan that is aimed at ensuring that there is data warehousing component in the data warehousing process.

Technology overview introduction

The datawarehouse can be taien and commercialized to act as a business. Business information systems are currently used widely in every industry to store information and data for future use. Data mining and data warehouse are the common processes found in information technology. The term data warehousing has dominated much energy and focus of the systems world. Data warehousing use is now a concentration area in companies because data warehousing is required in any organization in order to compete in today’s market place, they also need easy access to information that has a great impact on business decision. Data warehousing is a subject-oriented, time-variant, non-volatile, integrated collection of data used in support of management decisions. Subject-oriented is the data pulled from the various sources that are relevant to support various needs. It therefore allows better organization of data and also available in a summarized form. The organized data is integrated into the system so as to provide a consistent programme and format. Another feature of data warehousing is time-variant, data should be organized by time and stored using several time slices such as monthly, weekly, yearly or quarterly. The time-variant approach enhances reported summaries consistently across the departments and also allows historical data to be efficiently integrated. In data warehousing, non-volatility does not allow alteration of the database once the data has been written but it should be added as an addition on the database (Bernstein, 2000).

Technology review

The use of data warehousing provides powerful, cohesive access to unstructured and structured transactional and operational data in real time. This is because of the robust way employed in the manner the database has been created. Data warehousing fulfils its purpose when implemented in the business. Data warehousing systems allows easy access to information for both current requirements and also future requirements. So long as the user is granted permission to the system, some information about the company can be extracted easily once an individual has a requirement; information fed into the system brings out data as required. This therefore saves time and also makes information consistent.
One of the benefits of data warehousing systems is the better management of customer opportunity or customer relationship. Information about customers is useful in increasing ways of marketing. This information can be successfully used to acquire and retain customers and also increase or gain their products and services usage. Data warehousing is also making data available across corporate organizations for better usage and understanding between non-integrated functions like sales and marketing. Data warehousing systems also provides clean, value added, standardized data to create information from disparate sources enhancing, standardizing, and cleaning it to create information this is the benefit that cannot be quantitatively measured. Other benefits of data warehousing to an organization include the following. Chronological information for competitive and competent analysis; Enhancement of disaster recovery plans because of data backup; Can produce data extracts, reports, can also be done from the outside sources; Data warehousing can act as a repository for transaction processing systems; and data warehousing systems set up does not require technical skilled workers (Vassiliadis, 2000).

The diagram below shows the components of a data warehouse.

Data warehousing sets stage for data mining that is effective. Data warehousing improves greatly the chances of success in data mining. Over the years, companies have integrated data warehouse in order to enhance efficiency of data mining and also develop better business process. Data mining is the progression of data analysis from dissimilar standpoint and summarizing the data into practical information that can be used to cut costs, increase profits, or both. Data mining software is a methodological and systematic tool for analyzing and evaluating data. It assists its users in evaluating and analyzing data from many different angles or scope, proportions, dimensions, categorizing it, and reviewing and summarizing the identified relationships. In technical view, this is the procedure of finding patterns or relationships among all fields in the relational database.
Data warehouse architecture can be viewed in many perspectives and provide a meaningful way to view and analyze it. A successful data warehouse system depends on the database staging process that derives data from integrated Online Transactional Processing system. There are five data warehouse architectures commonly used namely Independent Data Marts, Data Mart Bus Architecture, Hub-and-spoke Architecture, Centralized Data Warehouse Architecture, and Federated Architecture (Lujan-Mora et al., 2003).

Independent Data Marts

Independent data marts are also known as small scale or localized data warehouse. It is used mainly by divisions or departments of a company in order to provide individual operational databases. This data mart is simple and consists of different forms derived from many design structures from many inconsistent database designs.

Data Mart Bus Architecture

Bus architecture allows data marts located in many different servers, this allows data warehouse to function in virtual mode and combine all processes and data marts as one data warehouse.

Hub-and-spoke Architecture

The hub is the centralized server responsible for information exchange and spoke handles transformation of data for all the operation data stores in the region. Hub and spoke focuses mainly on building maintainable and scalable infrastructure for data warehouse.

Centralized Data Warehouse Architecture

This architecture is build based on the hub-and-spoke architecture without dependent data mart component. It copies and stores the heterogeneous external and operational data to a data warehouse that is single and consistent. This architecture has one data model which are complete and consistent from all data sources.

Federated Architecture

This is an integration of many heterogeneous data marts, operational data store or database staging, combination of reporting and analytical application systems. This architecture focuses on integrated framework so as to make data warehouse much more reliable.

There are two types of data warehouses

Enterprise Data Warehouse (EDW). This is a large database repository that crosses over all business functions in the entire organization and include data from every department, division, and unit in the organization. EDW is a large repository of current and historical transaction data of the whole organization. EDW can contain hundreds of terabytes or gigabytes, and even petabytes of data sometimes.
The second is Data Mart; this is a collection of subject areas that are organized for decision support based on a given office or department needs. Data Mart usually serve as an analytical and reporting solution within an organization for a particular department such as marketing, customer service, sales, and/ or accounting,

References

Bernstein, P., Rahm, E., 2000. Data warehouse scenarios for model management. In: Proceedings of the 19th International Conference on Conceptual Modeling (ER’00), LNCS, vol. 1920, Salt Lake City, USA, pp. 1–15.
Lujan-Mora, S., Trujillo, J., 2003. A comprehensive method for data warehouse design. In: Proceedings of the Fifth International Workshop on Design and Management of Data Warehouses (DMDW’03), Berlin, Germany.
Vassiliadis, P., 2000. Data Warehouse Modeling and Quality Issues. Ph.D. Thesis, Department of Electrical and Computer Engineering, National Technical University of Athens Greece.