Segmentation of food consumers according to their correlations with sensory attributes projected on preference spaces

Abstract Consumer segmentation strategies aim to identify sectors of the population of consumers with different criteria of preferences. The segmentation procedure used here is based on the correlation coefficients between consumers and sensory attributes, which helps to interpret preferences according to the intensity of sensory characteristics. The correlation coefficient between a consumer and an attribute is computed from the acceptability scores given by the consumer to the samples and the mean intensities of the attribute evaluated in the same samples by a group of trained panellists. Thus, a consumer is characterised by as many coefficients of correlation as quantified attributes and the similarity of coefficients between consumers is used to segment them by cluster analysis.