A Comprehensive Bibliometric Analysis of Missing Value Imputation

Abstract

Data quality plays a crucial role in tasks, such as enhancing the accuracy of data analytics and avoiding the accumulation of redundant data. One of the significant challenges in data quality is dealing with missing data, which has been extensively explored by the scholarly community and has resulted in a significant increase in related publications. It is important to recognize that the landscape of missing data in computer science offers numerous opportunities for further research. However, upon closer examination of existing studies, it becomes evident that many have not fully utilized bibliometric analysis tools and software for comprehensive literature reviews. Therefore, this study aims to explore the essential characteristics, trends, and prevailing themes in the field of missing data imputation. Through a thorough bibliometric analysis, this study demonstrated the evolution of knowledge and key focal points in the field of missing data imputation. The analysis consisted of 352 journal papers in social sciences published between 2012 and 2023, all centered on missing data imputation. Among these publications, "IEEE Access" has become a highly respected source. To systematically explore various aspects of missing data imputation, a conceptual framework was used to uncover potential research directions and underlying themes. Ultimately, a thematic map serves as a valuable tool for providing a comprehensive understanding, categorizing significant concepts into basic or overarching, developing, or declining, central, highly developed, and isolated themes. These overarching and underlying themes offer valuable insights and pave the way for prospective directions and critical areas of study.