Week 5 Project
/
DSS to Manage Decision Making
Let us now discuss how DSS works in conjunction with business intelligence to manage decision making in an organization. Business intelligence is the intersection of the needs of the business and the available information necessary to make the best business decisions. Business intelligence IS include the Internet, extranets, intranets, online transaction processing (OLTP), operational data stores (ODS), data warehouses (DWs), data marts (DMs), analytical applications (AAs), data mining applications, statistical analysis applications, predictive modeling applications, reporting systems, and data islands.
Let us explore the concepts of DSS.
Decision Support Systems
A DSS provides new sets of capabilities for user control and nonroutine decisions. With DSS, effort required to link users to structured information �ow is reduced and greater emphasis is placed on models, ad-hoc queries, display graphics, and assumptions. The two basic types of DSS are:
Data-Driven DSS: Is a system that supports decision making by allowing users to analyze and extract useful information previously buried in large databases. Often data from transaction processing systems (TPS) is collected in data warehouses for this purpose. Online analytical processing (OLAP) and data mining is then used to analyze the data.
Model-Driven DSS: Is primarily a stand-alone system that uses a model to perform what-if and other types of analyses. These systems were often developed by end-user divisions or groups not under the control of central IS. Their analysis capabilities were based on a model or strong theory combined with a user interface. This made the model easy to use.
The types of information that data mining can yield include:
Associations and occurrences linked to an event. For example, a person who purchases corn chips from a supermarket also purchases cola 65% of the time.
Classi�cation recognizes patterns that describe the group to which an item belongs by examining existing, classi�ed items and by inferring a set of rules.
Work is clustered in a manner similar to classi�cation when no groups have been de�ned.
Forecasts sales �gures with the help of existing values.
Sequences occur when events are linked over time. For example, if somebody purchases a house then the person will also purchase a new refrigerator within two weeks 65% of the time.
Components of DSS
A DSS database is a collection of historical and current data from several groups or applications. The source can be a massive data warehouse or a small, computerized database. A DSS software system is a collection of software tools used for data analysis such as data mining and OLAP tools or a collection of analytical and mathematical models. A model is an abstract representation that illustrates the relationships or components of a phenomenon. A model can be mathematical, physical, or verbal. Statistical modeling software helps establish relationships such as relating product sales to differences in income, age, or other factors among communities. Optimization models use linear programming to determine optimal resource allocation to either maximize pro�ts or revenue or minimize time or costs. A classic use of optimization models is to determine the proper mix of products within a market to maximize pro�ts. Model sensitivity analysis models ask what-if questions to determine the impact of changes in multiple factors on outcome.
After learning the concepts of DSS, let us focus on its applications and uses.