Exploring the food networks of Turkic peoples: A knowledge graph approach to transregional culinary connections
International Journal of Gastronomy and Food Science, vol.46, 2026 (SCI-Expanded, Scopus)
- Publication Type: Article / Article
- Volume: 46
- Publication Date: 2026
- Doi Number: 10.1016/j.ijgfs.2026.101643
- Journal Name: International Journal of Gastronomy and Food Science
- Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
- Keywords: Computational gastronomy, Culinary ontology, Food networks, Knowledge graph, Turkic cuisines, Turkic peoples
- Istanbul University Affiliated: Yes
Abstract
This article develops a knowledge-graph approach for analyzing culinary relations among Turkic peoples. Computational gastronomy has been conducted predominantly in English and has relied mainly on single-language datasets, while research on Turkish cuisine has focused on its post-migration Anatolian phase; both leave the broader Turkic cuisines under-examined. To address this underexamined relationship, the study assembles a corpus of traditional recipes from 15 Turkic cuisines, compiled across the multiple languages in which they are recorded and spanning four geographic zones of the Turkic cultural landscape: Southern and Eastern Siberia, Turkistan, the Volga–Ural region, and the Black Sea–Caucasus. The recipes, validated through comparative reading, are normalized through expert coding and modeled as a relational knowledge graph containing 2704 core triples and 181 semantic relations. Two custom relations extend existing food ontologies: isCognateOf links recipes with shared linguistic roots, while hasVariant links differently named recipes with similar ingredient and technique profiles. Multi-layer network analysis at regional, recipe, and ingredient levels identifies a dense shared culinary core structured around grain, dairy, meat, offal, and meat-dough combinations. Community detection yields low modularity across algorithms, suggesting continuity rather than sharply bounded clusters. Geographic drift shows a dual-scale pattern: regional ingredient profiles show a weak but statistically inconclusive decline in similarity with distance, whereas individual recipes show no significant geographic effect. Taken together, the findings suggest that recipes function as relatively conserved units of cultural connection across geographic distance. By rendering cross-linguistic relations analytically visible, the ontology-enriched knowledge graph demonstrates the transregional ties of Turkic cuisines and offers a methodological framework for analyzing culinary connections across languages and regions.