Partitioned fuzzy integral multinomial logit model for Taiwan's internet telephony market

Customer satisfaction and ease of use are core values within the quality movement. Thus, a key issue in today's design activities is economically achieving these values by finding the quality attributes most valuable to customers. The multinomial logit model is a popular model for this purpose, although it relies on the assumption that its attributes are mutually independent. In social problems, however, attributes are often interdependent. To overcome this problem, some researchers have suggested using fuzzy integral, applying the idea of fuzzy measure to solve the non-additive problem. Yet that method is unsuitable for stated preference data, which is based on hypothetical choice data. This study combines factor analysis, the Choquet integral, and the stated preference method to propose a different way to solve these problems: the partitioned fuzzy integral multinomial logit model. Taiwan's Internet telephony market is used to illustrate this model, revealing that the partitioned fuzzy integral multinomial logit model is a successful model.