Deciphering the Department-Discipline Relationships within a University through Bibliometric Analysis of Publications Aided with Multi-variate Techniques

• Published in: 2015 IIAI 4th International Congress on Advanced Applied Informatics pp. 468 - 471
• Main Author: Gautam, Pitambar
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2015
• Subjects: Animals; Bibliometrics; bibliometry; Chemistry; clustering; correspondence analysis; cross-disciplinarity; ESI fields; Mathematics; multivariate statistics; Physics; scientometry
• ISBN: 1479999571; 9781479999576
• DOI: 10.1109/IIAI-AAI.2015.212

This study explores a practical approach to decipher the department-discipline relationships between the organizational research units dedicated to natural science, technology, engineering & medical (STEM) fields and 22 disciplinary categories used in Essential Science Indicators database (ESI 22 fields), for a Japanese national university as seen in a set of peer-reviewed journal publications (articles & reviews) indexed in the Web of Science (WoS) Core Collection database for a 5-years period. The procedure involved several steps such as (i) identification of publications of each organizational research unit through disambiguation of the affiliation data, (ii) assigning each publication to the corresponding ESI field based on journal title, (iii) aggregating bibliometric information of all publications for each research unit and discipline, and (iv) performing multivariate analysis, e.g., Clustering and correspondence analysis, to extract proximity relationships and internal structures that enable regrouping the obtained data and visualizing them using two-dimensional plots and bar diagrams. This approach may be easily adapted for analysis using other available disciplinary (subject areas or categories) schemes. Moreover, such analysis can be further extended to lower hierarchical levels, such as research divisions or research teams comprising a complex multidisciplinary department. The proposed affiliation-based analysis is useful for initial understanding the disciplinary contribution of the university departments to overall research output, e.g., For analysis of ranking based on performance for past 5-6 years tracing past history. It can be easily adapted to the bottom-up research performance analysis (based on current researchers) required for research administration or research strategy formulation based on the research output of the immediate past.