LAUSR

A novel model is presented in this article to derive the expected value from a ranking of alternatives, with which to tackle some uncertain problems, such as different types of att0itudes, in the creation of personalized quantifier caused by personal preference. There is no doubt that we consider here a reverse process of traditional ranking methods in decision making. More specifically, a representative sample of multi-attribute alternatives is prepared, and an involved person is asked to do nothing but provide, following personal preference or judgment on the whole, a ranking of alternatives of this sample. The relationship between the alternatives’ importance weights and their ranking positions is investigated. A TOPSIS-based model can then be established to derive his/her personal expected value from this ranking. The idea of ordered weighted averaging (OWA) aggregations is fully taken into account in the modeling. We use this technique to create a personalized quantifier, thus making the creation much more efficient and safer compared with what we did before. Experimental results show that different rankings of alternatives may lead to different expected values that eventually determine different types of attitudes. Thus, the developed technique could be used as a tool for uncertainty modeling in more complex situations in which various personality traits have to be taken into account, especially for decision making under uncertainty.