Food pairing analysis of Turkish cusinie via culinary fingerprints

The concepts of culinary fingerprint and food ingredient pairing are considered emerging approaches within the discipline of gastronomy and culinary arts. The term culinary fingerprint refers to the study of patterns related to the use of ingredients that are closely identified with a particular culture, have been preserved through intergenerational transmission, and are frequently employed in the traditional dishes of national cuisines. This approach specifically focuses on the frequency of ingredient use in the preparation of dishes characteristic of Turkish cuisine. On the other hand, the concept of food ingredient pairing can be described as grounded in the hypothesis that the presence and quantity of shared aromatic compounds between food items predict positive outcomes when these items are combined. This approach employs data-driven methods to interpret cooking patterns, dish structures, and underlying culinary principles. The objective of this thesis is to provide a scientific evaluation of the culinary fingerprint and food ingredient pairing, both considered contemporary approaches within Turkish cuisine. A content analysis was conducted on Turkish dishes obtained from various websites. For data analysis, the software MAXQDA 24.7.0 was utilized. Frequency analysis was employed to express the proportional usage of ingredients. Through these methods, representative recipes for Turkish cuisine were identified. For the food ingredient pairing component of the study, data on flavor compounds were obtained from FlavorDB2, an online flavor compound database. The shared flavor compounds between ingredient pairs in the collected recipes were identified, and their averages were calculated using the formulas employed in ingredient pairing analysis. Subsequently, a synthetic cuisine was constructed using the copy-mutant method to achieve the most realistic configuration. The shared flavor compounds of ingredient pairs within this synthetic cuisine were determined and subjected to the same calculation procedures. The differences between the representative recipes of Turkish cuisine and those of the synthetic cuisine were then computed. The average shared flavor compound value for Turkish cuisine was found to be 47.18927199, while that for the synthetic cuisine was 54.00306349. Subtracting the synthetic cuisine's average from the Turkish cuisine's average yielded a result of –6.8137915. In this context, Turkish cuisine exhibits the characteristic of negative flavor pairing. It can be suggested that, in Turkish dishes, the greater the shared flavor compounds between ingredients, the less frequently these ingredients co-occur. The analysis concludes that Turkish cuisine recipes tend to combine dissimilar or contrasting flavors, favoring diversity over similarity in flavor molecules. Additionally, it was determined that categories such as soups, vegetable stews with meat, other vegetable dishes, olive oil dishes, rice dishes, salads, and pickles display positive flavor pairing, whereas categories such as meat dishes, meat-based legume stews, egg dishes, pasta and dumplings, pastries, breads, buns, flatbreads, desserts, stewed fruits (hoşaf), and beverages exhibit negative flavor pairing.