Data normalisation techniques in decision making: case study with TOPSIS method

Abstract

Data normalisation is essential for decision-making methods because data has to be numerical and comparable to be aggregated into a single score per alternative. In multi-criteria decision-making (MCDM), normalisation must convert criteria values into a common scale, thus, enabling rating and ranking of alternatives. Therefore, it is a challenge to select a suitable normalisation technique to represent an appropriate mapping from source data to a common scale. There are some attempts in the literature to address the subject of normalisation, but it is still an open question which technique is more appropriate for any MCDM method. Our research contribution is an assessment approach for evaluating normalisation techniques. Here, we focus on six well-known normalisation techniques and on TOPSIS method. The proposed assessment process provides a more robust evaluation and selection of the best normalisation technique for usage in TOPSIS.