[The identification of the primal wine production with the protected designation of origin with the appliance of cluster metrics]

Abstract

The article considers the guidelines of the choice of identification criteria, allowing to verify and confirm the geographical name of the origin of domestic primal wine production, thereby confirming their legal status. A priori the production of wine with protected designation of origin includes the use of certain raw materials with predetermined organoleptic and physical-chemical characteristics, which can be confirmed by respective tests. The aim of the work was to develop a robust differentiating criterion that allows one to determine the authenticity and origin of wine materials relative to the standard. Material and methods. The authors presented a clustering technique, which allows on the basis of test results and developed digital identification criteria to verify the origin of wine materials from Krasnodar and Rostov-on-Don regions. As a criterion, the data from the analysis of mineral and trace element composition of primal wine production in these regions have been used. Results. The article postulates following: the main concern of clustering, methods of identification from the perspective of food production using food regression model, information on fundamental clustering metrics, fields of appliance according to the approach of the identification of the product with indication of geographic place of origin. Based on the results of the analysis of the content of 21 mineral substances (10 in μg/l and 11 in mg/l), a regressive model of the primal wine production was built. Based on this model, cluster centers were identified. The resultant model allows us to distinguish the two mentioned wine regions and form a spatial digital discrimination criterion based on the proximity to one of the established cluster centers. Conclusion. The proposed model can be adapted to identify the production of different branches of the food industry.

Keywords: Voronoi diagram; clustering; digital identification criterion; food product with geographical indication; identification; regression model.