Entropy-Based Customer Value Assessment with Interval Numbers

This paper proposes an entropy-based approach to customer value assessment while customers´ information are interval numbers. The decision matrix in the form of interval is normalized firstly, In the process of determining the attribute weights, the normalized decision matrix is transformed into a definite one by maximizing entropy of the attributes. Then, the weights of attributes are determined based on minimizing the weighted distance between the alternatives (i.e., the customers) and ideal one. According to the distances of every customer to the ideal one, the customers can be ranked or grouped. An example is used to illustrate the proposed approach.