Wine and grape fingerprinting based on chemical compounds and data analysis – insights from bibliometric analysis and text mining

Abstract

Abstract This paper explores the growing field of wine and grape authenticity based on chemical compounds and data analysis. A total of 468 studies published in the last 45 years were collected from the Web of Science. The results indicate that the geographic origin of wines is a key theme, involving authenticity, terroir, and fingerprinting. Studies were conducted on wines from around the world, with Europe being the most studied continent, followed by Asia, North America, South America, Australia and Africa. Some countries such as China, Spain, Italy, France, Portugal, Romania, Brazil, Argentina, Chile and Australia have been identified as regions with high wine production volume and scientific studies. The most commonly used terms in this area include exploratory data analysis, PCA, cluster analysis, ICP-MS, pattern recognition, and LDA. On the other hand, newer terms such as feature selection, support vector machines, neural networks, machine learning, and data mining have attracted the attention of authors and emerged as trendy and important terms in publications. The applied methodology provides a broad perspective on wine and grape authenticity research and is a guide for future projects.