ANDez: An open-source tool for author name disambiguation using machine learning

• Published in: SoftwareX Vol. 26; p. 101719
• Main Authors: Kim, Jinseok, Kim, Jenna
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.05.2024; Elsevier
• Subjects: Author name disambiguation; Authority control; Bibliometrics; Machine learning; Science of science; Scientometrics; Scientometrics; Science of science; Bibliometrics; Author name disambiguation; Authority control; Machine learning
• ISSN: 2352-7110; 2352-7110
• DOI: 10.1016/j.softx.2024.101719

•ANDez consolidates multiple ML techniques for disambiguation.•Built using Python and popular ML libraries.•Provides a unified platform to evaluate and refine ML methods.•Assists scholars with limited ML expertise in bibliographic data analysis. Author name disambiguation in bibliographic data is challenging due to the same names of different authors and name variations of authors. Various machine learning (ML) methods address this, but a unified framework for comparing them is lacking. This study introduces ANDez, an open-source tool that integrates top-performing ML techniques for author name disambiguation. Developed in Python using popular ML libraries, ANDez provides a transparent system, merging complex procedures from different ML approaches. This promotes the assessment, modification, and benchmarking of ML techniques in author name disambiguation. ANDez's user-friendly design also helps researchers analyze ambiguous bibliographic data without needing advanced ML coding expertise.