DECISION SUPPORT SYSTEM (DSS)

DECISION SUPPORT SYSTEM (DSS)

DSS combine data analysis and sophisticated models to support non-routine decision making. This is useful in helping the managers make decisions on ill-defined problems in rapidly changing environments. They provide the user with an interactive interface and bring together analyses and models to make sense of existing internal and external data.

There are several major capabilities of DSS (Turban and Aronson 2001);

• Provide support in semi-structured and unstructured situations
• Support several sequential and interdependent decisions
• Support intelligence, design, choice and implementation phases of decision making
• Support a variety of styles and processes
• Are adaptive and flexible over time
• Are user-friendly with strong graphical capabilities
• Improve accuracy, timeliness and quality of decision making
• Have substantial modelling capability to allow experimentation with different strategies under different scenarios.

There are multitude of DSS on offer in the marketplace. Besides that, there are also two distinction would be to separate them into MODEL-DRIVEN DSS & DATA DRIVEN DSS. 

Below is Model-Driven DSS provide a range of statistical, financial, forecasting and management science models that may be applied at strategic, tactical or operational levels.

The DSS may contain between a few and several hundred models encompassing:-

1. Strategic Models – tend to help high-level strategic planning processes within organization
2. Tactical Models - assist in allocating and controlling organizational resources such as capital budgeting and human resource planning
3. Operational Models – help to support day-to-day decision making, such as loans approval and quality control processes.
4. Analytical Models- cover methods of analysis such as statistical models or specific financial models.

Besides that, in contrast – Data Driven DSS : are more focused on examining patterns and relationships in large amounts of data. They used Knowledge evaluation tools such as online analytical processing (OLAP), to provide multidimensional analyses and data mining techniques looking at  à

1. Associations: are occurrences linked to a single event
2. Sequence: are events linked over time
3. Classification: recognizes pattern in certain groups such as loyal or fraudulent customers
4. Clustering techniques: can help to determine different groupings of certain customers where the classification don’t necessarily exist.
5. Forecasting techniques: can help to predict values for certain variables
6. Data mining techniques: vary considerably in the variety of approaches adopted, from fuzzy logic to neural networks. There are different industrial sectors illustrate typical analyses that data mining techniques may help to uncover:        - Banking industry

       - Retail and Marketing

       - Insurance

In order to purchase a DSS, an organization need to decide between custom-made solutions or off-the-shelf solutions.

1. Off-the-shelf solutions - are the lower costs, flexibility and application of many business problems in the same sectors
   
2.  Custom-made solutions – allow the organization to differentiate themselves from a competitors and use a more sophisticated approach based on different configuration. There are 7 classification;-

1. Text-oriented DSS – using technologies such as Web-Based documents, hyperlinks and intelligent agents

2. Database-oriented DSS – featuring strong report generation and query-searching capabilities

3. Spreadsheet-oriented DSS – such as Excel which uses statistical and financial models and techniques

4. Solver-oriented DSS – functions or procedures used for optimizing certain variables such as the optimal ordering quantities of certain resources based on historical data

5. Rule-oriented DSS – often expert systems linked to procedural and inferential (reasoning) rules such as evacuation of a tall building in case of fire in certain parts of it

6. Compound DSS – containing two or more aspects of the above five classification

7. Intelligent DSS – similar to rule-oriented DSS that can learn using agent technology and machine learning techniques