Earticle

When a competent employee leaves an organization, the technical skills and know-how possessed by that employee also disappear, which may lead to various problems, such as a decrease in organizational morale and technology leakage. To address such problems, it is important to increase employees’ job satisfaction. Due to the advancement of both information and communication technology and social media, many former and current employees share information regarding companies in which they have worked or for which they currently work via job portal websites. In this study, a web crawl was used to collect reviews and job satisfaction ratings written by all and incumbent employees working in nine industries from Job Planet, a Korean job portal site. According to this analysis, regardless of the industry in question, organizational culture, welfare support, work system, growth capability and relationships had significant positive effects on job satisfaction, while time and attendance management, performance management, and organizational flexibility had significant negative effects on job satisfaction. With respect to the path difference between former and current employees, time and attendance management and organizational flexibility have greater negative effects on job satisfaction for current employees than for former employees. On the other hand, organizational culture, work system, and relationships had greater positive effects for current employees than for former employees.

해외건설 프로젝트를 기획하고 수행하는 과정에서 현지 시장의 상황을 신속하고 정확하게 파악하는 것은 수익성 창출에 매우 큰 영향을 미친다. 뉴스기사 데이터는 정치, 경제, 사회 등 다양한 관한 정보를 담고 있기 때문에 시장의 상황을 파악하는 데 사용할 수 있는 좋은 데이터이다. 텍스트의 형태로 존재하는 대량의 뉴스기사 데이터로부터 정보를 추출하고 내용을 요약하는 과정에서 인력, 비용, 시간의 소모를 줄이기 위해 텍스트마이닝 기술이 필요하다. 본 연구에서는 뉴스기사에 다양한 주제가 공존한다는 특성으로 인해 발생하는 정보 추출의 한계를 극복하기 위해 잠재 디리클레 할당(Latent Dirichlet Allocation) 방법론을 사용하여 토픽 모델링을 수행했다. 문서 집단에 존재하는 주제의 개수가 10개라고 가정했을 때, 이용자들의 편의 증진을 위한 프로젝트(2번 주제)와 아프리카 지역의 빈곤 문제를 해결하기 위한 민간 차원의 지원(4번 주제) 등의 주제 집단이 존재하는 것을 확인했다. 이와 같이 문서 집단의 주제를 구분함으로써 더욱 의미있는 정보를 추출하고, 요약 결과의 활용성을 높일 수 있다.

Sufficient understanding of oversea construction market status is crucial to get profitability in the international construction project. Plenty of researchers have been considering the news article as a fine data source for figuring out the market condition, since the data includes market information such as political, economic, and social issue. Since the text data exists in unstructured format with huge size, various text-mining techniques were studied to reduce the unnecessary manpower, time, and cost to summarize the data. However, there are some limitations to extract the needed information from the news article because of the existence of various topics in the data. This research is aimed to overcome the problems and contribute to summarization of market status by performing topic modeling with Latent Dirichlet Allocation. With assuming that 10 topics existed in the corpus, the topics included projects for user convenience (topic-2), private supports to solve poverty problems in Africa (topic-4), and so on. By grouping the topics in the news articles, the results could improve extracting useful information and summarizing the market status.

The COVID-19 pandemic has affected various sectors of society, including politics. The political changes have had both positive and negative impacts on people's lives. Different public discussions happened during that situation on social media. It is essential to understand those discussions to prepare for the same kind of situation in future. Therefore, this study aims to identify the topics discussed on Twitter regarding this influence. During March 2020 and December 2021, 10,658 Tweets were gathered through the Twitter application programming interface and preprocessed using Python libraries. After feature extraction using the bag-of-words method, both probabilistic latent semantic analysis (PLSA) and latent Dirichlet allocation (LDA) were used as topic modeling methods. As a result of the analysis, 15 topics by LDA and 25 topics by PLSA were extracted during the study and then grouped into five key themes: Government responses for managing the COVID-19 Pandemic, Government decisions for COVID-19, Public response to government measures for COVID-19, Social influence, and Vaccination. Through a comparative evaluation of the LDA and PLSA topic modeling techniques, the research identifies LDA as the superior method, providing more accurate and coherent results.

The study reported in this paper aimed to evaluate the topics and opinions of COVID-19 discussion found on Twitter. It performed topic modeling and sentiment analysis of tweets posted during the COVID-19 outbreak, and compared these results over space and time. In addition, by covering a more recent and a longer period of the pandemic timeline, several patterns not previously reported in the literature were revealed. Author-pooled Latent Dirichlet Allocation (LDA) was used to generate twenty topics that discuss different aspects related to the pandemic. Time-series analysis of the distribution of tweets over topics was performed to explore how the discussion on each topic changed over time, and the potential reasons behind the change. In addition, spatial analysis of topics was performed by comparing the percentage of tweets in each topic among top tweeting countries. Afterward, sentiment analysis of tweets was performed at both temporal and spatial levels. Our intention was to analyze how the sentiment differs between countries and in response to certain events. The performance of the topic model was assessed by being compared with other alternative topic modeling techniques. The topic coherence was measured for the different techniques while changing the number of topics. Results showed that the pooling by author before performing LDA significantly improved the produced topic models.

This study applies topic modeling analysis to AI?dance research articles indexed in SCOPUS and KCI to identify the major themes and knowledge structures of international AI?dance scholarship, and, through comparison with domestic research, to suggest future directions for AI?dance research in Korea. First, while both international and domestic AI?dance research have experienced rapid growth since the pandemic, they exhibit distinct differences in research perspectives and methodologies. International research focuses on the structures, algorithms, and creative systems of AI technologies?such as generative AI, learning, movement modeling, neural networks, and human?computer interaction?thereby conceptualizing dance as a data-driven object of creation and analysis. In other words, AI is understood as a technological medium that performs practical creative functions within dance. In contrast, domestic research largely approaches AI as a tool for education, classroom practice, and applied case studies, remaining at an exploratory stage of potential utilization. AI is positioned not as a component of creative systems but as part of an educational discourse. Consequently, while international research has already entered a convergent and multidisciplinary research ecosystem, domestic research still remains in a foundational stage focused on educational introduction and preliminary application. To bridge this gap, it is necessary to strengthen the core competencies required for AI research, pursue more ambitious and forward-looking studies grounded in a deeper understanding of international scholarship, and expand domestic research into specialized educational topics as well as culturally and identity-based themes.

Screenwriting manuals are comprehensive guides designed to instruct and aid individuals in the art of screenwriting. A total of 60 screenwriting manual books were systematically analyzed with topic modeling methods to identify a topographical overview of the field along with similarity analysis using Word2Vec. The analysis employed TF-IDF, LDA, and Word2Vec techniques to identify key themes and semantic structures. All 60 screenwriting manual books contained similar contents focusing on four categories - story, film, audience, and industry. The core theme running through these categories was identified as 'character-centered story'. Clustering analysis revealed limited differentiation in content among manuals despite their proliferation in the market. Dual characteristics of screenwriting, such as in-between art and craft, writing and filmmaking, and autonomous and disciplinary, were detected through semantic analysis. This study makes methodological contributions to screenwriting research by demonstrating how computational text analysis could complement traditional qualitative approaches. Findings of this study can inform the development of new screenwriting manuals that address gaps in current offerings and serve as a framework for evaluating screenwriting pedagogy. This paper can help us design a new screenwriting manual with a difference and make a methodological contribution to the field of screenwriting research.

Nowadays, users' reviews and feedback on e-commerce sites stored in text create a huge source of information for analyzing customers' experience with goods and services provided by a business. In other words, collecting and analyzing this information is necessary to better understand customer needs. In this study, we first collected a corpus with 99,322 customers' comments and opinions in English. From this corpus we chose the best number of topics (K) using Perplexity and Coherence Score measurements as the input parameters for the model. Finally, we conducted an experiment using the latent Dirichlet allocation (LDA) topic model with K coefficients to explore the topic. The model results found hidden topics and keyword sets with high probability that are interesting to users. The application of empirical results from the model will support decision-making to help businesses improve products and services as well as business management and development in the field of hotel services.

Games are widely used in many fields, but not all games are successful. Then what makes games successful? The question gave us the motivation of this paper, which is to identify critical factors for successful games with topic modeling technique. It is supposed that game reviews written by experts sit on abundant insights and topics of how games succeed. To excavate these insights and topics, latent Dirichlet allocation, a topic modeling analysis technique, was used. This statistical approach provided words that implicate topics behind them. Fifty topics were inferred based on these words, and these topics were categorized by stimulation-response-desiregoal (SRDG) model, which makes a streamlined flow of how players engage in video games. This approach can provide game designers with critical factors for successful games. Furthermore, from this research result, we are going to develop a model for immersive game experiences to explain why some games are more addictive than others and how successful gamification works.

This study aimed to identify key topics related to the consumption of second-hand fashion products by analyzing reviews of fashion products on a C2C second-hand trading platform. Reviews of second-hand fashion stores in Bungaejangter were collected and analyzed using text mining and Latent Dirichlet Allocation (LDA), a topic modeling technique. The topic modeling analysis resulted in the extraction of eight topics: customer service and satisfaction, wearing experience and satisfaction, product condition and customer expectations, interaction in the transaction process, gap between expectations and reality, seller friendliness and trust, overall satisfaction, quality and value assessment. The findings of this study provide valuable guidelines for developing strategies to meet customer expectations and demands in C2C-based online secondhand fashion transactions. They will contribute to enabling secondhand fashion sellers and platform operators to deliver customer-centric services and pursue sustainable growth.

Purpose: This study quantitatively examined the portrayal of physician assistant (PA) nurses in Korean media by integrating text network analysis with latent Dirichlet allocation (LDA) topic modeling. Methods: A total of 3,564 news articles published by nine major Korean media outlets between 2020 and 2024 were analyzed. Content analysis was conducted using term frequency-inverse document frequency calculations, network centrality analysis, and LDA topic modeling to extract key terms, map discourse structures, and identify latent topics. Results: The analysis identified four primary topics in Korean media discourse: "healthcare workforce expansion policies" (30.4%), "hospital clinical practice and operational management" (23.5%), "institutionalization of the PA nursing role" (17.8%), and "COVID-19 response and public health crisis management" (28.3%). High-centrality keywords included "hospital," "medical," "patient," "physician," "government," and "nurse," indicating that the discourse primarily focused on clinical settings. Topic modeling revealed a major shift from pandemic-centered coverage in 2020 to a focus on healthcare workforce policy and PA nurse institutionalization in 2024, coinciding with the passage of the Nursing Act. Conclusion: This study provides empirical evidence suggesting that the portrayal of PA nurses in Korean media discourse evolved from a peripheral regulatory issue to a central healthcare delivery solution, particularly in the contexts of workforce management, clinical practice, and crisis response. Our findings suggest that PA nurse institutionalization received broader attention when positioned as part of systemic healthcare improvements addressing concrete clinical needs. These results offer valuable insights for policymakers and administrators in framing and implementing workforce policy reforms.

This study explored issues related to the library in the COVID-19 era in YouTube videos in Korea. This study performed social network analysis and topic modeling analysis by collecting 479 YouTube videos, 20,545 words, and 8,379 channels related to COVID-19 and the library from 2019 to 2020. The study results confirmed that YouTube, a social media platform, was used as an important medium to connect users and physical libraries and provide/promote online library services. In the study, major topics and keywords such as quarantine, vlog, and library identity during the COVID-19 pandemic, library services and functions, and introductions and user guides of libraries were derived. Additionally, it was identified that videos about COVID-19 and the library are being produced by various actors (news and media channels, libraries, government agencies, librarians, and individual users). However, the study also identified that the actor network is fragmented through the channel network, showing a low density or weak linkage, and that the centrality of the library in the actor network is weak.

This study presents an automated content analysis of 2,379 abstracts of research articles published in four major journals from 1997 to 2017. A total of 45 topics were extracted using the structural topic model (STM), a statistical text analytic method that allows us to infer latent topics from abstracts and to conduct statistical significance tests regarding the relationships between topic prevalence and abstract features. We found that while some topics (e.g., public health or online social movement) demonstrate increasing trends, others (e.g., third-person effect and communication strategies) reveal decreasing trends; authors in non-US-based institutions are more likely than those affiliated with US-based institutions to focus on the societal consequences of media (e.g., agenda-setting effect or public deliberation). Additionally, interesting relationships were found between topic prevalence and the number of article authors (e.g., popular topics among multi-author papers are lab-based experiments or large-scale campaign studies). Finally, topic-journal associations were identified.

This study investigates how digital discourse surrounding Yuseong-gu public libraries is structured and how it informs evidence-based policy architecture. Online text data were collected from major Korean portals (Naver and Daum) between July 2022 and June 2025 using "Yuseong-gu public libraries" as the core search query. The corpus was analyzed using text mining, keyword network analysis, and latent Dirichlet allocation (LDA) topic modeling. Word frequency and TF-IDF results indicate that place-anchored identifiers (e.g., Yuseong-gu, Daejeon) and culture-related vocabulary constitute the discourse backbone, while managerial and operational terms such as integration, support, and homepage signal demand for coordinated governance and enhanced digital accessibility. N-gram analysis further emphasizes the demand for an integrated information and participation portal, most clearly reflected in the recurrent sequence "Yuseong-gu-integrated-library-homepage." Network analysis reveals a high-density structure with a short average path length, confirming strong thematic interconnectedness; the node "library" functions as the primary hub and is directly linked to "culture," indicating the library's discursive positioning as a cultural platform. The findings support strategic policy directions, including a hub-satellite spatial system embedded across neighborhood life zones, cross-sectional programming integrating education, culture, and community participation, a mobile-first integrated digital portal, and institutionalized partnerships with schools and local cultural institutions.

Artificial intelligence (AI), an essential skill of the Fourth Industrial Revolution, is being actively taught in higher education; however, AI education is only in the preparatory stage in elementary, middle, and high schools. Investigating various newspaper articles related to AI education to date can aid in basic data collection, which is an important process in the preparatory stage. Accordingly, 13,378 newspaper articles were collected from a total of 21 newspapers, and five topics were extracted using the latent Dirichlet allocation (LDA)-based topic model along with frequency analysis. Newspaper articles from the early 2000s expanded to technologies related to the Fourth Industrial Revolution. Accordingly, education in AI fields should be linked with education in AI-based technology. In addition, efforts should be made to secure the continuity and sequence of AI education in cooperation with related higher institutions and companies.

Purpose: The purpose of this study was to identify the supportive care needs of parents of preterm children in South Korea using text data from a portal site. Methods: In total, 628 online newspaper articles and 1,966 social network service posts published between January 1 and December 31, 2019 were analyzed. The procedures in this study were conducted in the following order: keyword selection, data collection, morpheme analysis, keyword analysis, and topic modeling. Results: The term "yirundung-yi", which is a native Korean word referring to premature infants, was confirmed to be a useful term for parents. The following four topics were identified as the supportive care needs of parents of preterm children: 1) a vague fear of caring for a baby upon imminent neonatal intensive care unit discharge, 2) real-world difficulties encountered while caring for preterm children, 3) concerns about growth and development problems, and 4) anxiety about possible complications. Conclusion: Supportive care interventions for parents of preterm children should include general parenting methods for babies. A team composed of multidisciplinary experts must support the individual growth and development of preterm children and manage the complications of prematurity using highly accessible media.

Purpose: This study analyzed research trends in infant and toddler rearing behavior among family caregivers over a 10-year period (2010-2021). Methods: Text network analysis and topic modeling were employed on data collected from relevant papers, following the extraction and refinement of semantic morphemes. A semantic-centered network was constructed by extracting words from 2,613 English-language abstracts. Data analysis was performed using NetMiner 4.5.0. Results: Frequency analysis, degree centrality, and eigenvector centrality all revealed the terms ''scale," ''program," and ''education" among the top 10 keywords associated with infant and toddler rearing behaviors among family caregivers. The keywords extracted from the analysis were divided into two clusters through cohesion analysis. Additionally, they were classified into two topic groups using topic modeling: "program and evaluation" (64.37%) and "caregivers' role and competency in child development" (35.63%). Conclusion: The roles and competencies of family caregivers are essential for the development of infants and toddlers. Intervention programs and evaluations are necessary to improve rearing behaviors. Future research should determine the role of nurses in supporting family caregivers. Additionally, it should facilitate the development of nursing strategies and intervention programs to promote positive rearing practices.

본 연구는 지금까지 한국의 영어 교육 분야에서 발표된 챗봇 관련 연구들의 주제 동향을 탐색하고, 시간의 흐름에 따라 연구 주제의 변화를 확인하고자 실시되었다. 본 연구의 목표를 달성하기 위해 학술지 논문을 중심으로 관련 문헌을 선별하여 수집하였다. 그 결과 93편의 학술 논문이 최종 분석 대상으로 선정되었다. 아울러, 분석 대상 논문에서 제목과 초록을 추출하였으며, 이를 바탕으로 토픽 모델링 기법을 활용하여 챗봇 관련 연구들의 주제 동향 및 시간에 따른 주제 변화 추이를 관찰하였다. 본 연구 결과를 정리하면 다음과 같다. 첫째, 한국의 영어 교육 분야에서 실시된 챗봇 관련 연구들은 주로 학생들을 대상으로 하고 있으며, 사용된 챗봇 유형은 당시의 기술 동향이 반영되고 있는 것으로 나타났다. 둘째, 토픽 모델링 분석을 통해 "다양한 목적을 위한 챗봇의 다면적 응용(주제 1)", "챗봇 사용의 효과와 사용자 인식 이해(주제 2)", "챗봇 중심 과업, 과업 별 챗봇 개발 및 학습자-챗봇 상호작용(주제 3)"의 세 가지 연구 주제 동향이 관찰되었다. 셋째, 해당 주제들의 시간적 변화 추이를 살펴본 결과, 주제1의 비중은 시간이 지남에 따라 증가하는 추세를 보였고, 주제2는 반대로 감소하는 추세를 보였다. 주제3에서는 시간의 흐름에 상관없이 일관되게 낮은 관찰 빈도가 목격되었다. 본 연구는 한국의 영어 교육 맥락을 바탕으로 수행되었으나, 한국을 포함하여 유사한 외국어 학습 배경을 공유하는 학문 공동체에 역시 영어 교육에서의 효과적인 챗봇 통합을 위해 필요한 향후 연구 주제에 대해 다양한 방향과 지침을 제공해줄 수 있다.

This study aimed to investigate the specific topics of scholarly inquiries related to a chatbot within the domain of English pedagogy in South Korea and to track the changes in these topics over time. To achieve this goal, rigorous criteria were established to screen and gather relevant literature, resulting in the selection of 93 scholarly articles. From these articles, titles and abstracts were extracted and analyzed using topic modeling techniques. The findings of this investigation are as follows: first, it was observed that most chatbot-related scholarly articles in Korea predominantly involved students as subjects, and the types of chatbots used reflect sensitivity to current technological trends. Secondly, the topic modeling analysis identified three main topics: "the multifaceted application of chatbots for various purposes (Topic 1)," "understanding the impacts and perceptions of chatbot use (Topic 2)," and "chatbot-centric tasks, the development of task-specific chatbots, and learner-chatbot interactions (Topic 3)." Lastly, the temporal trajectory of these topics revealed a noticeable increase in the proportion of Topic 1, a decrease in Topic 2, and a consistently low frequency of Topic 3. Despite being limited to the context of South Korean English pedagogy, these insights provide several academic implications. Specifically, the findings can help researchers better understand and identify future research areas within and beyond the South Korean English education context.

본 수영은 대중적인 생활체육으로 인식되고 있다. 하지만 아파트, 호텔 등에 설치된 실내수영장에서 종종익수사고가 발생하고 있다. 따라서 본 연구에서는 쇼셜미디어, 뉴스, 잡지, 학술논문에서 수집한 데이터를 대상으로 텍스트마이닝 기법을 활용하여 수영장 익수사고를 중심으로 키워드 분석을 수행하였다. 이를 통해 수영장에서 발생하는 사고의 원인, 결과와 밀접한 키워드를 분석하여 향후 사고예방을 위한 기술적 연구에 활용하는데 연구의 목적이 있다. 본 연구에서는 최근 5년간 소셜 미디어 동영상(102개)과 뉴스기사(30개)를 활용하여 키워드 및 토픽 모델링 분석을 수행하였다. 그 결과 첫째, 아동 익수 사고와 구조 활동, 둘째로 부주의와 그에 따른 법적 책임, 셋째, 수영장, 워터파크, 주거시설 내 수영시설에서의 아동 안전사고 위험으로 분류되었다. 네트워크 분석 결과는 ‘수영활동’과 ‘조사’와 같은 키워드가 매개 역할을 하는 것으로 나타났다. 본 연구 결과는 향후 수영장 익수사고의 인과관계 분석, 기술적 해결방안을 마련하는 기초자료로 활용성이 높을 것으로 판단된다.

Swimming is widely recognized as a popular recreational sport. However, drowning accidents frequently occur in indoor swimming pools located in apartments, hotels, and other facilities. Therefore, this study employs text mining techniques to analyze drowning accidents in swimming pools using data collected from social media, news articles, magazines, and academic papers. The purpose of the study is to identify keywords closely related to the causes and consequences of these accidents and to provide insights that can inform future technological research aimed at accident prevention. This study utilizes 102 social media videos and 30 news articles from the past five years to conduct keyword analysis and topic modeling. The results reveal three main themes: (1) child drowning accidents and rescue activities, (2) negligence and related legal responsibilities, and (3) child safety risks in swimming pools, water parks, and residential swimming facilities. Network analysis further indicates that keywords such as “water activity” and “investigation” serve as mediating nodes within the keyword network. These findings are expected to serve as valuable foundational data for future causal analyses of swimming pool drowning accidents and for developing technical solutions to enhance safety in swimming environments.

In this paper, we proposed a multi-subject-level sentiment analysis model for user reviews using the Latent Dirichlet Allocation (LDA) method targeting user-generated content (UGC). Data were collected from users' online reviews of hotels in major tourist cities in the world, and 30 hotel-related topics were extracted using the entire user reviews through the LDA technique. Six major hotel-related themes (Cleanliness, Location, Rooms, Service, Sleep Quality, and Value) were selected from the extracted themes, and emotions were evaluated for sentences corresponding to six themes in each user review in the proposed sentiment analysis model. Sentiment was analyzed using a dictionary. In addition, the performance of the proposed sentiment analysis model was evaluated by comparing the emotional values for each subject in the user reviews and the detailed scores evaluated by the user directly for each hotel attribute. As a result of analyzing the values of accuracy and recall of the proposed sentiment analysis model, it was analyzed that the efficiency was high.