Insights into wheat science: A bibliometric review using unsupervised machine learning techniques

Abstract

Wheat (Triticum spp.) has been one of the most important cereal crops, serving as a source of protein and energy in the human diet. It remains a vital component of global food security, with extensive scientific literature dedicated to its study, although the large volume of literature often hinders global analysis. In this study, different unsupervised machine learning techniques, such as K-Nearest Neighbors (KNN) and Uniform Manifold Approximation and Projection (UMAP), text mining analyses, including word embeddings and statistical word analysis, and graph analysis methodologies, were applied to gain a deeper understanding of the wheat literature. The proposed bibliometric analysis was conducted and integrated with the Journal Citation Reports (JCR) to identify major wheat research trends in the PubMed literature. This analysis examined the evolution of these trends over time, evaluated the geographical distribution, impact, and research domains, and assessed author collaboration networks and the evolving relevance of different countries. Research on disease resistance, genetic modification, and dietary impact demonstrates a consistent increase in output, while interest in topics related to overcoming salt stress and enhancing animal feed appears to be diminishing. Interestingly, research on wheat germ agglutinin saw a surge in interest during the late 2000s, stabilizing thereafter. These trends underscore the dynamic nature of wheat research, driven by evolving priorities and technological advancements, particularly in genetics and omics tools. Moreover, the increasing significance of China in wheat research, including its size, impact, and networking, alongside longstanding leaders such as the United States, signals a shifting landscape in global wheat research.