What Is A
Decision Support System?

What is it?

Simply put, a decision support system is a computer program that makes it easier to make a decision. A simple example would be a spreadsheet.

Typically, however, a DSS is considered to be an interactive computerised system that analyses data and presents it so that knowledge workers can make business decisions more easily in environments which are rapidly changing and not easily predictable. It may include expert systems or artificial intelligence and helps managers and planners to extract useful information from data gathered from a wide range of sources, including documents in a variety of electronic formats, personal knowledge, business models, industry data etc., so that they can identify and solve problems.

A decision support system can be used for such things as comparing sales figures from different periods, for analyzing the consequences of various decision alternatives, deciding whether to launch a new product and in medical decision making.

 

A (very!) brief history

Decision support started when people began applying computer technology to the work being done on group decision-making in the 1970s, when it was described as “a computer-based system to aid decision-making". Then we had what-if spreadsheets and rules based software which allowed for a major leap forward. In 1990 data warehousing and online analytical processing widened the scope of DSS, and later applications to analyze information available on the Web were introduced.

As the various technologies improved DSS found their way into the fields of management, education, agriculture and medicine.

 

Types of models

As is common in new fields or areas, there is no universally accepted classification, rather there are several different classifications proposed by different people.

For example, Haettenschwiler talks about passive, active and cooperatives DSS. He makes this distinction in relation to how the user utilizes the system. A passive system obviously is an aid in the process of decision-making but does not elicit suggestions or solutions. In contrast, an active DSS will bring out alternative decision suggestions.  A cooperative system allows the decision maker to revise and refine suggestions provided by the system until a final solution is settled upon.

Daniel Alter developed a classification very early on and his idea is that decision support operations extend along a single dimension, ranging from extremely data oriented to extremely model oriented. At one end are file drawer systems that simply provide access to data items. For example, real-time equipment monitoring or inventory monitoring and reorder systems.

In the middle are analysis information systems with decision oriented databases and small models. For example, product planning and analysis or sales forecasts based on a marketing database. At the other end are suggestion decision support systems using logic models that suggest specific decisions for a well understood task, for example insurance renewal rate calculation or credit scoring.

Daniel Power describes them differently, this time based on the mode of assistance. There are five classes.

  • Communication driven DSS allows more than one person (in an office or via the web) to work on a task, for example, a web-conferencing program.
  • Data driven DSS or data oriented DSS allows for processing of data that is stored in a database or data warehouse. The information is categorized chronologically, weekly sales, monthly costs, etc., and the DSS is used to answer specific queries, e.g., showing trend lines in sales, growth, costs etc.
  • Document driven DSS has to do with the organization of information in a multitude of electronic formats. An example is a search engine that deals with html pages, pdf, image files, video files etc.
  • Knowledge driven DSS provides problem-solving expertise stored as rules procedures and algorithms, e.g., in stock trading a stop loss limit is seen as a knowledge driven model.
  • Model driven DSS allows for processing of information using quantitative models. For example, organizing job rotas, predicting future costs, tax planning are typically done using model driven DSS.

Power also differentiates in terms of scope, with an enterprise-wide DSS serving large data warehouses and many managers, whereas a desktop, single user DSS is a small system running on one managers personal computer.

 

Components

A decision support system is composed of the following components

  • 1. data management
  • 2. user interface
  • 3. model management
  • 4. knowledge management.

Of course the users themselves are another very important aspect! While there are high tech computers and artificial intelligence at work, few organizations leave the decision making to computers and ultimately the humans make the decisions...

Read more about the applications and benefits of DSS

Like this page?






Tips for dealing with psychopaths and narcissists

Fortnightly newsletter with practical tips and ideas
Learn more...
Plus!
'7 Vital Do's and Don'ts of Decision Making' when you subscribe!