Data Mining Techniques in Disease Classification: Descriptive Bibliometric Analysis and Visualization of Global Publications

Abstract

Many literature searches are required in scientific study, and these take a significant amount of time and effort. The bibliometric analysis is useful for locating research hotspots and gaining an understanding of research trends, according to the published literature. This bibliometric analysis was carried out with the assistance of the tools Bibliometric package in R (biblioshyny) and VOSviewer; the data used in this analysis was primarily derived from the Scopus repository to analyze the data mining approaches for disease classification (DMDC). A sample of 804 articles was selected by utilizing a query including essential key terms such as (”data mining” OR “data-mining”) AND (“disease classification” OR “disease identification” OR “disease prediction”). Overall, the findings of the study indicate the highest number of publications on applying DMDC published in 2019 with 141 research articles. As for individual researchers, the most productive authors are Jabbar MA, Li J, and Wang X. Jabbar MA in the field of DMDC, and 57 articles (1.2%) were written by one author, while the rest of the 747 articles were written by multiple authors. The Advances in Intelligent Systems and Computing journal (26 articles) has the greatest number of published articles connected to the DMDC field. Total of 804 articles in 474 distinct journals, there are 57 journals that have already published more than three papers, accounting for 12.03%. The USA was the most productive country, and the University of California is the affiliated university that comes from most top research in this field. The most cited research article published by Moore JH in 2006 included 489 citations. There are different types of diseases identified using data mining techniques such as heart disease, Brest cancer, liver disease, chronic kidney disease, Parkinson’s disease, diabetes mellitus, and Alzheimer’s disease. The most widely used algorithms in the research community include the random forest, decision tree, support vector machine, and naive bayes algorithms. In the future, the field of data mining could grow in many different directions and could be an effective way to increase the accuracy of disease prediction