Analysis of Theses on 21st Century Skills in Türkiye Using Text Mining Method

Authors : Adem Uzun
Pages : 156-172
Doi:10.38015/sbyy.1791724
View : 69 | Download : 146
Publication Date : 2025-12-30
Article Type : Research Paper
Abstract :This study aims to identify the emphasized content in graduate theses focused on 21st-century skills in Türkiye. The research was conducted based on a qualitative research design and carried out using the document analysis method. Accordingly, the keyword \\\"21st-century skills\\\" was selected as the search criterion on the National Thesis Center website of the Council of Higher Education (CoHE). The permission status and thesis type options were set to \\\"all,\\\" and theses accessed between February 24, 2025, and March 6, 2025, were included in the study. The documents used in the research were analyzed using topic-based text analysis, a method of text mining applied with the R program. Through artificial intelligence-assisted topic algorithms, Latent Dirichlet Allocation (LDA) was used to identify the most frequently addressed themes in the thesis abstracts. The connections between topic titles and thesis summaries have been evaluated, and a comprehensive content analysis has been conducted. It has been observed that the words in the summary content generated by the GDA algorithm align with consistency tests. Additionally, to obtain more detailed findings on 21st-century skills, certain words were removed, and the algorithm was reapplied. When evaluating the results obtained from the research, it can be said that studies on 21st-century skills have increased in quantity in recent years, have formed themes around science and STEM subjects, and are mostly conducted through experimental studies. However, it has been observed that topics such as media literacy, digital literacy, and social sciences, on which the Ministry of National Education and the European Union are working regarding 21st-century skills, did not form a theme in this study. The obtained results are expected to serve as a reference for students and thesis advisors in selecting future thesis topics and identifying original research areas.
Keywords : İçerik analizi, gizli dirichlet tahsisi, metin madenciliği, döküman analizi, konu modelleme

ORIGINAL ARTICLE URL