A framework and prototype for intelligent multiple objectives group decision support systems.

Abstract

The objectives of this research are threefold: (i) to develop a conceptual framework and a prototype in order to extend the application capability of a category of multiple objective decision support systems (MODSS) techniques; (ii) to explore the combined functionalities of knowledge-based expert systems (ES) and MODSS through embedding an intelligent front-end, and (iii) to develop a new system and process of dealing with multiple objective decision making (MODM) models in a group decision support system (GDSS) framework. Ultimately, a system that integrates MODSS, ES and GDSS is generated, which is then evaluated in a laboratory experimental setup. This integrated system contains a sufficient number of MODM methods to solve MODM problems, provides an ES-based guide to select and use the most suitable MODM method, and has the capability to aggregate individual decision makers' preferences to produce a compromise solution of an MODM problem in different forms and styles of group meetings. The system is supported by a set of group decision making (GDM) methods which combine the preferences of the individual group members and thus increases the confidence of each group member in the compromise solution.The research is conducted using a multiple-methodologies approach using the system development methodology as the backbone. The conceptual framework of the integrated system is elaborated to integrate multiple system elements into one facility at the application system level based on functional and resource integration. A prototype implements this conceptual framework as an intelligence-based and graphical user interface (GUI)-based MODSS that works in an individual/group environment. Both the conceptual framework and the prototype are called Intelligent Multiple Objectives Group Decision Support Systems (IMOGDSS).Initial evaluation of the IMOGDSS is encouraging, which is conducted in the form of testing a number of hypotheses in an experimental setup. This research thus makes contributions in both theoretical and application domains. Five major contributions are listed below:It develops a unique conceptual framework of integrating MODSS, ES and GDSS effectively to deal with MODM problem in individual/group decision making under a knowledge-based intelligent architecture.It provides a new application of ES, that is, utilising knowledge-based ES to select the most efficient MODM method for each particular decision maker (or decision group) in a particular decision problem.The complete method management function of the MODM methodology base guides the decision makers to use the most suitable method to solve their decision making problems, allows them to use multiple methods to resolve complex problems, that could not otherwise be solved with a single MODM, and also allows the group members to get solutions from different methods.This study produces an opportunity to select and apply the 'best' aggregation model to aggregate the individual solutions of an MODM problem through integrating various GDM methods in a methodology base.This study implements a two-stage configuration of group decision support software that provides a GUI-based hierarchical procedure for solving MODM problems with intelligent guidance in a decision group. The two-stage group decision making procedure is able to help the decision makers to analyse, understand and interact cooperatively in the group decision making process to reach a compromise solution.