A Perceptual Computer based method for supplier selection problem

The supplier selection is one of the most important components of production and logistics management for many companies. The supplier selection problem involves different kinds of words in which all of the criteria are weighted using one kind of words and the performance evaluations for all sub-criteria are another kind of words. How to aggregate these words without losing information is a very daunting task using a type-1 fuzzy set (Tl FS) approach. This paper applies a new methodology- Perceptual Computer (Per-C)-to help solve this hierarchical group multi-criteria decision making problem. The Per-C has three components: encoder, computing with words (CWW) engine and decoder. First, the interval approach (IA) is used to obtain interval type-2 fuzzy set (IT2 FS) word models for the words in a pre specified vocabulary. Second, a linguistic weighted average (LWA) is used to aggregate all the words modeled by IT2 FSs. Finally, a centroid-based ranking method is used to rank the supplier selections, and a similarity measure is used to obtain similarities of the supplier selections. A committee of judges organized with P decision makers and K experts decides the winning supplier selection as the one with the highest ranking and least similarity to other suppliers.