Question

An example of a problem for the DSS and clearly define the problem? and the 3 possible solutions.

Answer 1

DSS problem:

Decision Support System(DSS) is one of the Decision Support disciplines supported by computer, but there is no universal definition of it. Some are:

1. Sprague and Carlson:

"DSS is an interactive computer-based system that helps decision makers to utilize data and models to solve unstructured problems."

2. Finlay:

"DSS is a computer-based system that aids the process of decision making."

3. Turban :

"DSS is an interactive, flexible, and adaptable computer-based information system, especially developed for supporting the solution of a non-structured management problem for improved decision making."

Considering above definitions, DSS ranges from systems answering of simple queries to systems modeling of a complex human decision making process. Therefore, it is easy to put a variety of information systems into the DSS class.

Problems

As it has been already pointed out, DSSs deal with human decision making. Thus, they have to deal with human inconsistency and uncertainty as well. One of the aspects for improvement can be the user interface.Another improvement can be found in representing of decision problem and decision problem solving methods. Most of the problems are not new, they have been in the research focus of many scientists, but a need to find ever better and better solutions is surely still present. More about the user interface and some specific problems in Model Driven-DSS and Knowledge Driven -DSS are introduced in following sections. :

Model-driven and Knowledge-driven DSS A simplified preview of main objectives and differences between MD-DSS and KDDSS is presented in Figure 1

Figure 1

The Figure consists of five components:

1) User – a person who uses DSS

2) User Interface – part of the system which communicates with the user

3) Model Base – a set of analytical and optimization tools which perform the decision making process

4) Inference Engine – a part of the system which makes conclusions

5) Knowledge Base – information (knowledge) including data and rules stored in a database

Problems in model-driven DSS

Most of the problems in model-driven DSS concern decision making and decision theory problems, such as:

1. Uncertainty.

Very often, it is impossible to determine complete and unambiguous information. Choosing the right representation of such information can be crucial to make a right decision.

2. Inconsistency.

Avoiding human inconsistency is impossible, but some mechanisms to deal with inconsistent data or at least to notify if contradictory information exists are necessary.

3. Multi-criteria decisions.

Finding the optima in multi-criteria decision problem requires also attribute ranking. The problems are how to measure such a thing and how to combine measurements to produce a decision.

4. Decision instability.

Finding a solution is one part of the problem, another one is to find a stable solution. Decision should be carefully analyzed and presented with its advantages and disadvantages.

Problems in knowledge- driven DSS

Most research topics in knowledge-driven DSS relate to the following problems:

1. Representing the knowledge

Expert knowledge is represented by rules. The easiest form is if-then rule, but reducing a part of an expert knowledge to such a rule is sometimes impossible. System should be able to handle rules that are more complicated.

2. Getting knowledge from experts.

It is reasonable to think that experts are not going to write down if-then rules. The need for a suitable interface, yet likely to native language, is inevitable.

3. Resolving rules conflicts.

Setting the rules by humans naturally leads to rules conflicts. A good algorithm for resolving that problem is needed.

4. Producing a conclusion.

Both knowledge facts and rules arise from large amount of data. Producing a conclusion in acceptable period of time becomes very important.

5. Learning.

Recently, an actual topic is computer learning. The possibility to incorporate learning in the process of decision making sounds very attractive. Some new trends in solving some of above problems are incorporating fuzzy logic and a growing number of heuristic algorithms in modeling of the decision making process.