Earlier we noted that file management approaches are often sufficient to support traditional transaction processing. Without question, database management systems can improve the efficiency of processing data by eliminating data redundancies, improving data integrity, and so forth. However, the big change that database management systems have enabled is the realization of event-driven data processing systems. As noted earlier, event-driven systems are oriented toward the concept that complete data describing business events should be kept in its original form, where multiple users from throughout the organization can view and aggregate event data according to their needs.
At the heart of this movement toward event-driven systems is a fundamental shift in the view of information processing in business organizations. Traditionally, organizational Information Systems have been focused first on capturing data for the purpose of generating reports, and using the reporting function to support decision making. Increasingly, management is shifting to viewing the primary purpose of organizational Information Systems as decision support while reporting is secondary. This perspective leads to a focus on aggregating and maintaining data in an original form from which reports can be derived, but users can also access and manipulate data using their own models and their own data aggregations. In Chapter 5, we will discuss Information Systems such as business intelligence and expert systems that are designed to improve decision making. If you look ahead to the figures in Technology Insights 5.2 and 5.3 you will notice that both types of support systems generally require access to detailed data stored in databases.
The strategic shift toward event-driven systems is further embodied in two contemporary concepts that are driving new database management systems implementations: data warehousing and data mining. Data warehousing is the use of Information Systems facilities to focus on the collection, organization, integration, and long-term storage of entity-wide data. Data warehousing’s purpose is to provide users with easy access to large quantities of varied data from across the organization for the sole purpose of improving decision-making capabilities. Data mining is the complementary action to data warehousing. Data mining refers to the exploration, aggregation, and analysis of large quantities of varied data from across the organization to better understand an organization’s business processes, trends within these processes, and potential opportunities to improve the effectiveness and/or efficiency of the organization. The “warehouses of data” analogy makes sense as the software to support data storage is akin to physical warehousing approaches used to store and retrieve inventory—when an item needs to be restocked on the store shelf, there must be some system whereby the item can be located in the warehouse and retrieved.
Data warehousing and data mining opportunities are enabled and enriched through the use of event-driven systems focused on capturing data that provide comprehensive views of business events. However, neither effective event-driven systems nor data warehouses are possible without effective implementation of database management systems. Both objectives are dependent on the massive data integration and data independence made possible through database technology. Both warehousing and data mining may also be limited if well-designed database models that provide for future information needs are not effectively implemented. This process starts with the information requirements analysis and successful attainment of an understanding of all users’ potential data and information needs.
What do the concepts of data warehousing and data mining mean?