Abstract The relationship between naming traditions and geography is a complex subject that encompasses historical, sociocultural, and genetic dimensions. The intersection of naming traditions with geography and culture is a rich field of study, revealing much about human social structures, historical migrations, and cultural evolution. While qualitative studies have explored the sociocultural aspects of naming practices, the application of novel quantitative methods can enable identifying geospatial patterns. By employing qualitative methods on a dataset comprising Türkiye’s most common baby names in recent years, this study seeks to identify distinct spatiotemporal naming practices and socio-cultural patterns. To this end, firstly we extend our previous work, in which we introduced a basis of tool for demographic analysis. Enhancing and utilizing this tool we investigate baby names not only on geospatial scale but also in temporal dimension. Then, we employ K-means clustering as the unsupervised machine learning technique and show that, baby names are not randomly distributed but exhibit patterns attributed to geographic heterogeneity and reflect cultural dynamics. Employing data visualization techniques, we further trace cultural evolution in recent years. By combining K-means clustering along with Principal Component Analysis (PCA), we also show that name distributions across provinces exhibit semantic embeddings of baby names. Finally, based on the identified cultural clusters we show the correlation with politica ltrends, highlighting the potential of naming data as a proxy for broader social science research.
Emre Öner Tartan (Thu,) studied this question.