Multiple Criteria Decision Making Methods for Business-to-Business Electronic Commerce

The objective of this thesis is to develop appropriate methods and tools to deal with decision making problems arising in business-to-business electronic commerce. In this context, we propose MCDM methods for the problem of supplier selection and the problem of customer evaluation for order acceptance aiming to tackle effectively and efficiently issues of complexity and uncertainty which are inherent in the respective decision processes. Regarding supplier selection, we present a novel evaluation method with two stages. A satisficing technique is used in the first stage of this method for the prequalification of the suppliers. The final evaluation of suppliers is performed in the second stage through the use of a fuzzy AHP technique. As far as customer evaluation is concerned, we present a novel class of methods extending TOPSIS. On the one hand, the proposed class of methods tackles the issue of uncertainty of the decision maker preferences through the use of linguistic variables and their representation as fuzzy numbers. On the other hand, the proposed class of methods may be parameterized according to the behavioral pattern of the decision maker towards risk. Moreover, a simulation experiment is designed and executed in order to analyze and study the proposed class of methods. The results of the simulation experiment show that it is adequate to distinguish three instances of the proposed class of methods that correspond respectively to risk-averse decision makers, to risk-neutral decision makers, and to risk-seeking decision makers. Finally, we present a novel agent-based DSS architecture for flexible electronic marketplaces. This architecture is employed for the implementation of a Decision Support System incorporating an application for supplier selection which is based on the proposed decision method.