Topic Modelling of Postgraduate Theses in the Field of Management Information Systems in Turkey with LDA Algorithm

Authors : Göktuğ İlisu, Nursal Arıcı

Pages : 1-13

View : 56 | Download : 21

Publication Date : 2025-06-30

Article Type : Research Paper

Abstract :Management Information Systems is a branch of science that deals with integrating information technology solutions and business processes in order to meet the information requirements of businesses and organisations at strategic, managerial and operational levels. In this respect, it is a multidisciplinary research field that draws upon a number of different disciplines, including computer science, management science, statistics, organisation theory and decision theory. The primary objective of this study is to investigate the reflections and improvement of the field of Management Information Systems as evidenced by graduate thesis topics in Turkey. To this end, 1,070 graduate theses (951 MSc-f and 119 PhD-f) prepared in the field of Management Information Systems between 2002 and 2023 and accessible via the YÖK National Thesis Centre website were included in the scope of the study. Topic modelling was performed with the LDA algorithm. The data set employed in the topic modelling process comprised the English abstracts of graduate theses. The thesis abstracts were subjected to text preprocessing and stemming. The resulting words were converted into nested lists and topic models were obtained by applying the LDA algorithm. Data visualisation was employed to create word clouds, topic clusters, word frequency histograms and document-topic distributions. It was established that the terms \\\"data,\\\" \\\"model,\\\" \\\"research,\\\" \\\"technology,\\\" and \\\"system\\\" were predominantly incorporated into the topic models.
Keywords : Yönetim bilişim sistemleri, konu modelleme, gizli dirichlet tahsisi algoritması

ORIGINAL ARTICLE URL