Use of Cluster and Factor Analysis for Grouping Bulgarians Honeys according to Botanical Origin

Abstract

Fourteen samples from acacia (Robinia pseudoacacia), lime(Tilia sp.), sunflower (Helianthus annuus), thistle (Onopordum acanthium), forest and honeydew honeys have been characterized according to their parameters of mineral and sugar content, color characteristics and refractive index. The samples have been grouped in two clusters–monofloral lime honey and monofloral acacia honey on the basis of the investigated characteristics by cluster analysis. The honey samples from sunflower and honey from Onopordum acanthium are incorrectly classified according to their botanical origin as lime honey while the honeydew and lime honey are determined correctly. By applying factor analysis the investigated indicators reduced to 6 factors explained 91.55 % from the entire variation. The parameters with the highest factor weight influence on the distribution of honey in clusters.