Intelligent User Support in Multicriteria Decision Support

Abstract

In an attempt to enhance the relationship between user and system, Multicriteria Decision Support Systems (MCDSS) which incorporate some form of intelligent support have been developed. Intelligent MCDSS (IMCDSS) are potentially useful to the multicriteria practitioner in (at least) two respects. Firstly, they can act as an “intelligent assistant” to the multicriteria analyst both when working interactively with decision makers in a decision making environment and when carrying out “backroom” analysis. Secondly, by supporting the D.I.Y user they may enable models to be handed over to clients with less need for ongoing support. There is also the potential to make modelling tools more available for D.I.Y use by managers, whilst guarding against their misuse. Intelligence takes on many guises, both in how it is generated and how it is relayed to the user. This paper discusses our research in the area of intelligent user support over the last two years. We begin by taking a brief look at initial studies, which examined a number of decision support systems (namely, V·I·S·A, SIMUL8, Frontier Analyst, and Decision Explorer). This highlighted possible support areas as well as analysing different methods of defining, capturing and relaying intelligence. The paper then goes on to present and discuss a prototype intelligent MCDSS designed and implemented by taking an existing multicriteria package and building in different forms of “intelligence”. The next stage of the research is to evaluate the effectiveness of the IMCDSS and we will discuss ways in which this might be done.