Index - Major Sections
Home

**InHCc HMIS**

Site Map
Health Economic and Reform

Benefits

Discussion

Data and Data Analysis

Health Management

Product and Services
References
Team

_______________

Index - Same Level Subject

Index - Child Subjects

Introduction

The concept of data warehouses began to rise as organizations found it necessary to use the data they were collecting through their operational systems for future planning and decision-making. Assuming they use the operational systems, they had to build queries that summarized the data and fed management reports. Such queries, however, would be extremely slow because they usually summarize large amounts of data, sharing the database engine with every day operations, which in turn adversely affected the performance of operational systems. The solution was, therefore, to separate the data used for reporting and decision making from the operational systems. Hence, data warehouses were designed and built to house this kind of data so that it can be used later in the strategic planning of the enterprise.

Therefore data warehousing requirements differ between a data warehouse and a database transactional operating system. Data accumulated in a data warehouse is used to produce informational reports that answer questions like "Who?" and "What?" about the original data.

The Data Warehouse is a separate and a different type of database than a Day-to-Day transactional database.

Definitions:

"Is subject-oriented, integrated, time-variant, nonvolatile collection of data in support of management decisions (Inmon W.H., 1996)."

"the place where people can access their data."  (Ralph Kimball)

“A method for an organization to improve decision-making through improved access to information."

“A home for ‘redundant’ data that originates from distributed sources.”

“It is the architect associated with the coordinated and periodic copying of data from various sources, both inside and outside the organization, into an environment optimized for analytical and informational processing”

“The process of creating an architect information-management solution to enable analytical and informational processing despite platform, application, organizational, and other barriers” (Simon, A. R.,  1997, pg 12)     

Benefits         

Data Warehousing offers an effective and cost efficient means to integrate data from disparate sources systems.

è    Does not access operational data

è    Provides for individual formatting of data

è    Provides for cross system analysis. It integrates data from several diverse, heterogeneous sources.

è    Does not require consensus on format of data (can create what is needed)

How a Data warehouse differ from Traditional operational data stores

Operational Data	Decision support Data
Application-oriented: data serves a particular business process or functionality	Subject-oriented: data serves a certain subject of the business, such as client, health care indicators, etc.
Detailed data	Summarized or refined data
Structure is usually static	Structure is dynamic where new data cubes can be created as needed. Existing cubes can also be modified by adding new dimensions, or dimension levels.
Targets data-entry people	Targets managers and decision makers
Volatile (can be changed).	Non-volatile (is not changed after it is inserted).
Requirements are usually known before the design of the system.	Requirements are not totally understood prior to the design of the system
Follows the classical development life cycle	Completely different life cycle.
Performance is important because of the large number of concurrent users that may access the data	Performance issues are more relaxed, since a far smaller number of people are expected to access the data simultaneously; thus there are no serious concurrency issues to worry about
Transaction-drive	Analysis-driven
Must be highly available for the end users (back-up and recovery plans are very well planned)	Does not have the same degree of high availability requirement, and plans for back-up and recovery are more relaxed.
Usually retrieved in units or records at any given time. This leads to processing only a small amount of data in any given process (transaction).	Usually retrieved in sets of records. This leads to processing large amounts of data in one single process (such as finding out a particular trend based on data collected over several years).
Reflects current situation.	Reflects values over time (historical)
When managed (administered) it is usually managed as a whole	Usually managed in pieces, or smaller sets.

Links

Back to Top