Earticle

This study investigates how a government-funded research institute facilitated the growth of a high-potential spin-off firm through a combination of equity investment and life-cycle support. Drawing on theories of absorptive capacity and innovation intermediation, the research examines how ETRI’s support mechanisms enabled Firm B to transform external technologies into tangible performance outcomes. The case demonstrates three distinct stages of technology absorption support: (1) short-term learning through consultations and seminars, (2) intensive problem-solving for commercialization difficulty, and (3) long-term researcher on-site dispatch to embed knowledge into product design and manufacturing. These sequential supports illustrate the transformation of potential absorptive capacity into realized absorptive capacity. Beyond technology absorption, ETRI’s commercialization division provided an integrated package of consulting, infrastructure, branding, and government project linkages, thereby acting as an innovation intermediary that connected technology, capital, and markets. As a result, Firm B achieved significant outcomes, including the localization of technology, revenue growth, U.S. patent registrations, expanded R&D capabilities, and enhanced enterprise value. The findings provide empirical evidence that government-funded research institutes can effectively accelerate firm growth by simultaneously enabling absorptive capacity and mediating innovation, offering both theoretical contributions and practical implications for institute-led growth models.

This study aims to explore strategies for revitalizing the aerospace and defense industry ecosystem in South Korea’s Jinju-Sacheon region. Although Jinju is a key hub for the nation’s aerospace and defense industry, with major corporations like Korea Aerospace Industries (KAI) located there, the startup ecosystem and regional innovation network are underdeveloped. The lack of collaboration and technology transfer between large corporations and small and medium-sized enterprises (SMEs) limits opportunities for growth, highlighting the need to revitalize the startup ecosystem and promote technological innovation and market entry. To address this, the study conducted a topic analysis on news data related to the aerospace and defense industry. Based on these findings, this research proposes a collaboration model among large corporations, SMEs, research institutes, and universities. The results are expected to diagnose the collaborative structure within Jinju’s aerospace and defense industry and contribute to the sustainable development of the regional industry by providing practical strategies to strengthen the startup ecosystem and regional innovation network.

This study investigated psychological factors (intrinsic/extrinsic motivation for major choice, major satisfaction, academic adjustment, and other‘s influence on university choice) impacting undergraduate dropout intention (N=331). It specifically compared lower-grade (1st/2nd year) and upper-grade (3rd year and above) students using multi-group Structural Equation Modeling (SEM), particularly considering the ‘sophomore Syndrome’ context. Scalar invariance across groups was established, ensuring valid comparisons. Results revealed no overall statistically significant differences in the structural path model between the academic year groups. However, path analyses for each group showed that major satisfaction and academic adjustment consistently and significantly reduced dropout intention for both lower and upper-year students. Notably, intrinsic motivation unexpectedly exhibited a positive association (or tendency) with dropout intention for both groups. Some subtle path variations were observed, such as extrinsic motivation appearing more influential on major satisfaction for lower-year students. These findings suggest that while core dropout mechanisms might be largely similar across academic stages, the role of intrinsic motivation is complex and potentially paradoxical. Universities may therefore need nuanced student support strategies that address these specific dynamics to effectively mitigate dropout.

This Paper builds an international trade model of endogenous innovation more consistent with empirical evidences on R&D productivity and theoretically revisits how international trade affects economic growth. Specifically, to reconcile the shortcomings of the conventional model with the empirical evidences of both decreasing returns to R&D and the limited international knowledge spillovers of international trade on R&D productivity, as well as the costly adoption of technology to the production process, this paper makes the R&D productivity subject to the congestion externality from previous R&D and also introduces the technology specific training for production workers as the firm’s investment for the adoption of technology. Main findings are as follow. Trade affects growth positively, and the mechanism for this positive effect is the resource allocation rather than the costless knowledge spillover. Thus, the greater the dissimilarity among trading countries, the greater the growth effect, but trade between countries with the identical economic structure has no growth effect.

This study aims to collect and analyze unstructured customer review data to analyze user experiences on global fandom platforms like Weverse, Bubble, and Mubeat. The research aims to analyze user review data on fandom platforms using a text mining technique based on Latent Dirichlet Allocation (LDA). Considering user experience reviews, the study aims to derive insights for service improvement and platform competitiveness. The analysis revealed that “communication”, “convenience”, and “fandom identity” were the main themes identified in positive reviews. The negative reviews identified “login and account errors”, “translation and subtitle quality issues”, and “decreased usability after updates” as key issues. This study identified technical stability and translation quality improvements as urgent challenges for global fandom platforms. Therefore, platform operators must ensure stable service quality and provide customized features for global users. This can contribute to the continued growth of global fandoms and strengthen the competitiveness of the K-content industry.

This study presents a comprehensive analysis of Deleuze's theory of film color and identifies it as a philosophical operator within his taxonomy of images. By adopting Deleuze's philosophical concepts as an analytical framework, the research traces the evolution of the function of color through three distinct stages. First, it examines the affection-image in Michelangelo Antonioni’s films, then explores the logic of sensation in Jean-Luc Godard’s works, and finally analyzes the crystal-image in Alfred Hitchcock’s Vertigo. The study argues that color acts as a critical agent that facilitates the transition from the movement-image regime, which is governed by the sensory-motor schema, to the time-image regime. This research posits that the rise of Generative AI completes Deleuze's philosophical trajectory. The study analyzes AI-generated videos through specific criteria such as probabilistic color distribution, AI-specific any-space-whatever, and temporal hallucination. Through this analysis, it demonstrates that AI images are decoupled from any sensory-motor schema and therefore function as the purest form of the logic of sensation and the crystal-image. Deleuze's color theory provides a vital theoretical tool for elucidating the aesthetic and ontological nature of AI-generated video in the field of contemporary media studies.

This study aims to solve the 'semantic interference' problem that occurs in B2B recommender systems when processing over 17,000 product categories within a single embedding space. To address this, we propose a two-stage Hierarchical Multimodal Recommendation Model (Hierarchical Model). The proposed model first predicts one of 238 mid-level categories in the first stage, and then performs detailed item recommendation on a refined candidate pool within that category in the second stage. Comparative experiments against a single model (Baseline Model) using the same CLIP-based architecture showed that the Hierarchical Model achieved a consistent and significant performance improvement of 4.5-5.0%p on average across all key metrics, including Precision@K, MAP@K, and NDCG@K. This paper empirically demonstrates that this performance enhancement is not merely due to search space reduction, but is based on clear theoretical justifications: (1) 'semantic disambiguation' through the specialization of embedding spaces, (2) the efficiency of a 'learned cascade' structure, and (3) the mitigation of data sparsity via the 'information bottleneck' principle.

This study analyzed patent trends in decision support systems for worker health management to provide foundational data for future development of worker decision support systems. A total of 989 patents retrieved from the KIPRIS website, spanning from 1997 to 2025, were analyzed using Python-based Jupyter Notebook with keywords including "worker" and IPC codes (G16H/G06N/H04W). The analysis revealed four distinct periods of technological evolution: Period 1 exhibited a concentrated hub network centered on a few core technologies; Period 2 showed keyword diversification with the emergence of platform and smart technologies; Period 3 entered a convergent stage with the proliferation of AI and big data; and Period 4 demonstrated the convergence of mega-topics based on hyper-connected platforms and generative AI technologies, illustrating temporal changes across periods. Future research could verify whether shifts in network centrality correlate with actual industrial competitiveness, potentially serving as a strategic decision-making tool.

This study empirically investigates the impact of cloud service adoption on firm performance and examines the mediating roles of digital capability and organizational agility. While prior studies have primarily focused on adoption factors (e.g., relative advantage, complexity, organizational readiness) or on direct adoption-performance relationships, only few studies have provided an integrated analysis of how adoption strengthens internal capabilities and translates into performance outcomes. Drawing on the Technology-Organization- Environment (TOE) framework and the Dynamic Capability View (DCV), this study conducted a survey of 464 employees from firms in China and analyzed the data using PLS-SEM. The results show that relative advantage, top management support, and competitive pressure have significant positive effects on both digital capability and organizational agility, whereas complexity negatively influences organizational agility, and organizational readiness and regulatory support are not significant. Moreover, both digital capability and organizational agility positively influence firm performance, with organizational agility exhibiting a stronger effect. By clarifying the adoption-capability-performance mechanism, this study contributes to the extension of theoretical insights and offers practical implications for firms formulating digital transformation strategies.

This study aims to analyze user perceptions of public information systems in the publishing industry and suggest ways to improve their quality. Drawing on existing research on information systems, we identified quality factors for publishing information systems and conducted an Importance-Performance Analysis(IPA). The results revealed that satisfaction with information quality was relatively low compared to its importance. Satisfaction with some aspects of service quality was also low, while system quality and transaction quality showed relatively high levels of both importance and satisfaction. This study is significant in that it systematically identified user perceptions of publishing information systems through IPA analysis. The findings suggest the need for improving information and service quality.