Topic Modelling of Doctoral Theses Written on Lung Cancer in Türkiye Using LDA

Authors:
Publisher website
Google Scholar
Abstract
The aim of the study is to examine the research status, subject and content of doctoral theses on lung cancer in Türkiye. In December 2024, research documents are scanned using the text mining method in R software, employing topic-based text analysis. The search is conducted on the YOK National Thesis Centre page, selecting 'lung cancer', 'all', and 'doctorate'. The most frequently covered topics are found through the obtained thesis abstracts with the artificial intelligence-based 'Latent Dirichlet Allocation' algorithm. Content analysis is performed by examining the relationship between the subject headings and thesis abstracts. It is aimed to determine the most emphasized content in theses on lung cancer. As a result of the algorithm, the words are found to be compatible in the consistency test. The study shows that lung cancer research is mainly clinical and medical, but the data also has significant health management and health economics outputs. A detailed investigation of concepts like "quality of life, treatment process, cost, and value" identify areas for health policies and technology assessments. Latent Dirichlet Allocation (LDA) emerges as a tool to compare studies across databases, helping researchers choose topics and understand the subject density of theses conducted in Türkiye.

This publication has 14 references indexed in Scilit: