Abstract: The digital divide is one of the leading causes of social polarization in the era of digital transformation (Iivari et al., 2020). Economic inequality, such as income disparity, is causing complex social polarization during the fourth industrial revolution and the COVID-19 pandemic. A multidisciplinary understanding of the digital divide is necessary to address economic inequality related to quality of life, such as education, medical care, and gender and age issues (Martzoukou et al., 2020; Estacio et al., 2019; Adamczyk & Betlej, 2021). The HRD field was also directly affected by technology development, and digital HRD is being actively discussed (Thite, 2022). Understanding the digital divide in human resource development will significantly impact the education of organizational members while addressing the digital transformation of businesses. Furthermore, as digital technology spreads, the digital divide between enterprises leads to disparities in how the digital transformation is handled (Bennett & McWhorter, 2021). Therefore, this study analyzes the research trends, derives important topics, and provides primary data to gain a deeper understanding of the digital divide. The research questions considered in this study are as follows. RQ1: How are topics distributed in the context of research through the digital divide literature? RQ2: How should the derived topics be classified and strategically understood? An analysis method known as topic modeling is used to identify hidden topics in a document set and classify the content by similarity. This study aimed to model research abstracts regarding the digital divide using Latent Dirichlet Allocation (LDA), one of the most widely used topic modeling algorithms. LDA assigns keywords to one of k potential topics with a predetermined iteration value based on the appearance ratio of keywords in the document and the percentage of keywords in specific topics. For data collection, the title, abstract, publication year, and keywords were collected from the papers of interest using the bibliographic export function of Web of Science. The analysis included 5,558 articles with ‘digital divide*’ in their titles, abstracts, or keywords by April 7, 2022. There were 4,052 articles collected, 93 book reviews, six book chapters, 1,782 proceedings papers, 177 early access articles, 170 review articles, and 155 editorial materials. Only English-language studies were included in this study. This study preprocessed the data by removing stopwords in the abstract, such as copyright and publication information. This study performed LDA topic modeling analysis on nouns refined from 5,558 documents and using the ‘topicmodels’ package in R 4.0.3. Following parameter tuning, 21 topics were selected based on coherence and perplexity. The main topics derived were as follows. The first topic is on access to medical information and digital health literacy (most relevant terms: health, patient, portal, eHealth, medical, cancer, healthcare, providers, background, apps; composition ratio: 6.5%). The second topic focuses on the effectiveness of treatments or interventions and the digital divide (most relevant terms: participants, interventions, groups, results, outcomes, methods, conducted, test, questionnaire; composition ratio: 5.1%). The third topic focuses on the digital divide in schools and learning environments (most relevant terms: student, learning, education, educational, school, skills, teachers, university, computer, e-learning; composition ratio: 5.0%). The implications of this study are as follows. First, this study uses topic modeling to elucidate significant issues and present a multidisciplinary discussion that addresses the digital divide. Second, this study analyzed objective data to overcome the limitations of problem-solving alternatives based on subjective data, such as literature research and benchmarking. Finally, this study provides primary data that organizations can use to respond to digital transformation based on significant factors related to the digital divide. Keywords: Digital Divide, Topic Modeling, Research Trends
