Dicodess is a framework for building model-based distributed cooperative decision support systems. At first we will present the underlying principles that need to be understood when dealing with Dicodess. We will then discuss collaboration when using the DSS. The interested reader may get more information about Dicodess at (Gachet, 2004, 2009).
Dicodess encapsulates the underlying mathematical model into a graphical user interface (GUI) which spares the user from the technical details of a modeling language. By doing manipulations in the GUI the user actually specifies and finally generates a complete decision support model. This process is called structuring semi-structured problems. Figure 1 shows the abstractions Dicodess uses to support the process.
Abstraction of the decision process
To structure a problem completely three things need to be specified: The situation (which is based on facts, but could also comprise hypotheses and assumptions), the task (which can be influenced by the problem statement), and exogenous decisions (which are dependent on external constraints like the decision maker's will or some law). Dicodess uses distributed decision support objects (DDSO) to represent among other things the elements described above. These pieces of knowledge are mostly independent, reusable, and exchangeable. They can be managed (created, edited, exchanged, deleted, reused, combined, etc.) separately by their respective object manager. Figure 2 shows a DSS user's object managers.
Dicodess’ object managers