Development of an Authorship Identification Model Using Contextual Embeddings and Metric Learning
Abstract : With the recent rapid growth of domestic and international academic databases, the problem of author name disambiguation to distinguish between homonyms and synonyms has emerged as an essential task to improve the accuracy of various information services using academic information. Since existing rule-basedsimple statistical matching methods have limitations in capturing the contextual meaning of language, this paper proposes a new disambiguation model that precisely measures the fine-grained semantic similarity between authors by applying contrastive learning based on a pre-trained Korean language model. The proposed model converts the metadata of papers (author name, affiliation, title, abstract, etc.) into integrated text features and maps them into a high-dimensional vector space to determine whether they are the same author. In addition, the scalability of the system was secured by applying an incremental learning mechanism that issues new unique IDs to new authors in real time. As a result of the experiment, the proposed model demonstrated high performance with an accuracy of 0.962, precision of 0.958, recall of 0.962, and an F1-score of 0.957.
Keyword : Author Name Disambiguation, Researcher Name Disambiguation, Contrastive Learning, Incremental Learning
http://dx.doi.org/10.29056/jncist.2026.06.01
Analyzing Ray Sampling and Layer Normalization Contributions in NeRF-based CT Reconstruction
Abstract : CT enables precise examination of three-dimensional anatomical structures inside the human body, but it also exposes patients to high radiation because it requires many X-ray projections. This has increased the demand for sparse-view CT, which reconstructs volumesfewer images. NeRF has opened a promising route for efficient sparse-view CT reconstruction, and recent work seeks improvements not only via network design but also through training choices such as ray sampling and normalization. Here, we compared several importance-based ray sampling strategies under identical settings and found that preserving global coverage in ray selection is critical for reconstruction quality. We further show that applying layer normalization across sampling conditions stabilizes internal feature scales, improving training stability and final reconstruction performance.
Keyword : sparse-view CT reconstruction, NeRF, ray sampling, Layer Normalization
http://dx.doi.org/10.29056/jncist.2026.06.02
Abstract : This study examines how player experience elements and recommendation judgments vary by playtime at the time of review in the narrative-centered game Disco Elysium. A total of 53,099 English Steam reviews were collected through the Steam API, and 42,878 valid reviews were used after preprocessing. The analysis focused on review text, recommendation status, and playtime at review. K-Means-based BERTopic was applied to classify the reviews into 25 topics. The results show that short-playtime reviews were mainly associated with boredom, slow pacing, save-related issues, and unexpected failure experiences. In contrast, reviews written after more than 10 hours of play emphasized worldbuilding, characters, dialogue, RPG/CRPG genre experience, and long-term positive evaluation. Recommended reviews were linked to artistic quality, narrative, worldbuilding, and character attachment, whereas non-recommended reviews were also associated with game-external issues related to production entities and rights relations. The high proportion of such topics in non-recommended reviews with over 60 hours of playtime suggests that Steam recommendation status may reflect a complex judgment combining game experience and external context. This study demonstrates that player experience in narrative-centered games is organized around different evaluative focuses over time.
Keyword : Narrative-centered game, Player experience, BERTopic, Steam reviews, Playtime
http://dx.doi.org/10.29056/jncist.2026.06.03
An Integrative Review on Developmental Transformations and Neuro-Linguistic Programming
Abstract : This study is a literature review that examines the theoretical commonalities and integration potential between the Developmental Transformations (DvT) and Neuro-Linguistic Programming (NLP). DvT is a drama therapy approach that facilitates the experiential exploration of dynamic human developmental processes through bodily movement and improvisational play. NLP is a psychotherapeutic methodology that explains the formation and transformation of human experience through the interaction of neurological processes, linguistic structures, and behavioral patterns. This study compares and analyzes the relational structures and cognitive neuroscientific mechanisms of the two approaches by focusing on the core concepts of DvT—embodiment, encounter, and playspace—and the NLP techniques of rapport and modeling. The findings suggest that changes in developmental levels and the growth of embodied experience, as emphasized in DvT in response to clients’ actions, are closely linked to NLP’s techniques of rapport and modeling, thereby facilitating clients’ ive awareness of themselves and others and promoting metacognitive abilities. These findings indicate the theoretical possibility of an integrative and complementary relationship between DvT and NLP within a cognitive neuroscientific framework. By presenting an integrative perspective on DvT and NLP that has not been previously discussed, this study contributes as foundational material for theoretical expansion and integrative practice in the fields of drama therapy and psychotherapy.
Keyword : Developmental Transformations, Neuro-Linguistic Programming, Rapport·Modeling, Embodiment·Encounter·Playspace, Cognitive Neuroscientific Mechanism
http://dx.doi.org/10.29056/jncist.2026.06.04
Abstract : This study aims to examine the structural relationships among self-efficacy, psychological sense of community, and subjective well-being perceived by performing artists, and further to investigate the moderating effect of COVID-19 constraints. To achieve the purpose of the study, a survey was conducted with professional performing artists aged 20older, and 287 out of 300 distributed questionnaires were used for analysis. For hypothesis testing, structural equation modeling was performed using SPSS Amos 20.0. The results supported all proposed hypotheses: Hypothesis 1, that self-efficacy has a positive effect on psychological sense of community; Hypothesis 2, that psychological sense of community has a positive effect on subjective well-being; and Hypothesis 3, that COVID-19 constraints have a moderating effect on the relationship between self-efficacy and psychological sense of community. In addition, a comparison of path coefficients between groups with high and low levels of COVID-19 constraints revealed that the effect of self-efficacy on psychological sense of community was weaker in the high-constraint group than in the low-constraint group. These findings provide meaningful academic and practical implications. Furthermore, by examining the moderating effect of COVID-19 constraints on the relationship between self-efficacy and psychological sense of community perceived by performing artists, this study contributes to the theoretical development of research in the field of performing arts.
Keyword : Performing artist, Self-efficacy, Psychological Sense of Community, Subjective Well-Being, COVID-19 Constraints
http://dx.doi.org/10.29056/jncist.2026.06.05
Abstract : The purpose of this study is to understand the research trends on leisure psychology (leisure motivation, leisure experience, leisure attitude, leisure commitment, and leisure addiction) published in domestic tourism journals.1993 to 2025, a content analysis was conducted centering on a total of 105 papers published in the journal registered (candidate) in the field of tourism. Research trends were identified centering on eight analysis topics, including publication year, research topic, research scene, data collection methods, subject of investigation, research methodology, statistical analysis, and survey area. As a result of the analysis, research related to leisure psychology has increased since 2010, and shows high diversity in terms of research topics and survey subjects. This study will be of theoretical value in that it systematically analyzed the research trends of leisure psychology in journals in the field of tourism1993 to 2025. In addition, unlike previous studies on leisure psychology that were limited to identifying the frequencysimple trends of individual variables, this study holds academic significance in that it structured and analyzed the independent, moderator, mediator, and dependent variables studied together with the five core variables of leisure psychology. Based on the analysis results, this study derived the characteristics of leisure psychology research and suggested future research directions and limitations.
Keyword : Leisure Psychology, Research Trends, Content Analysis
http://dx.doi.org/10.29056/jncist.2026.06.06
An Analysis of the User Experience Structure of Digital Pet Content: A Case Study of Tamagotchi
Abstract : Tamagotchi is a representative digital pet franchise that develops through user care and interaction. In recent years, it has expanded beyond traditional digital pet experiences through character merchandise, pop-up stores, and collaborative content. This research examined the user experience structure of contemporary Tamagotchi experiences through a text-mining-based network analysis of online community data collected between May 2025 and April 2026. LDA topic modeling and CONCOR analysis were applied to preprocessed text data to identify major semantic structures and user experience clusters associated with Tamagotchi experiences. The results revealed two primary topics: Offline Experience and Product and Character Consumption Experiences. In addition, the CONCOR analysis identified four clusters: offline experiential experiences, character and design experiences, emotional and nostalgic experiences, and play and nurturing experiences. The findings suggest that digital pet content maintains its traditional nurturing-based play experience while expanding into a multifaceted experiential structure that combines character consumption, emotional engagement, and offline experiential activities. This research provides empirical evidence on the user experience structure of digital pet content based on online community data and identifies the key experiential factors that constitute contemporary Tamagotchi experiences.
Keyword : Digital Pet, Tamagotchi, User Experience, Emotional Consumption, Experiential Consumption
http://dx.doi.org/10.29056/jncist.2026.06.07
A Study on an AI-Service Design Thinking Integrated Policy Living Lab for Solving Community Problems
Abstract : The public sector today faces complex local challenges such as traffic congestion, older adults living alone, and environmental pollution. In reality, traditional top-down policymaking and simple citizen-participation living labs are not enough for rapid and effective responses. At the same time, the rapid growth of Generative AI offers new possibilities for public-sector innovation. However, many current AI policy cases are technology-centered. As a result, citizen participation and democratic decision-making are often not sufficiently ensured. This study explores how Service Design Thinking, a human-centered approach, can be integrated with AI. Based on this, it proposes a policy living lab model for solving community problems. It also suggests a workshop program that can be linked to real local policy processes. The study aims to present a new direction for public-sector innovation in addressing local issues. It is expected to serve as basic reference material for community innovation and public service design in the AI transformation (AX) era.
Keyword : AI, Living Lab, Policy Lab, Service Design Thinking, Social Innovation, Community Problem Solving
http://dx.doi.org/10.29056/jncist.2026.06.08
Abstract : This study empirically analysed the impact of realisable absorptive capacity and outbound open innovation on financial performance, along with the moderating effects of location environment and transaction costs, across 317 deep-tech start-ups. The findings confirmed that absorptive capacity-an internal capability-and external open innovation activities are key determinants of corporate survival and performance enhancement. Specifically, the location environment acts as a powerful catalyst in converting absorptive capacity into performance. However, it was found that an excellent location without accompanying innovation activities can trigger a 'strategic mismatch' phenomenon. This occurs when the inherent advantages of knowledge and talent accessibility are not leveraged, and the opportunity costs invested in securing the initial location negatively impact performance. This highlights the importance of a 'capabilities first, location second' strategy. Furthermore, the transaction cost mitigation (efficiency) mechanism exhibits a significant moderating effect only within the open innovation pathway. It was confirmed that even in environments with high potential for transaction costs arisingtechnological uncertainty, a leverage function maximises innovation outcomes when supported by effective cost-reduction measures. The findings of this study indicate that future policies for SMEs and venture companies should shift'universal distribution' to 'selective concentration'. Furthermore, policy support aimed at mitigating transaction costs is necessary to ensure responsiveness and sustainable corporate competitiveness, even amidst external environmental changes. To achieve this, a one-stop support system should be strengthened, centred on innovation intermediary organisations based on the quadruple helix model, to unify fragmented support functions. This is particularly crucial for establishing the self-sustaining competitiveness of deep tech technology start-up ecosystems outside metropolitan areas and building a sustainable system for disseminating outcomes.
Keyword : absorption capacity, open innovation, location environment, transaction costs, quadruple helix
http://dx.doi.org/10.29056/jncist.2026.06.09
Abstract : This study examines instructional and learning support for students with disabilities in higher education not at the level of individual experiences, butthe perspective of structural conditions and regional disparities in support infrastructure. Using public datasets, including records on special education support centers and special schools, disability welfare centers, and disability activity support services-this study analyzes the regional distribution of support resources that constitute the surrounding support ecosystem for higher education institutions. In addition, accessibility facility indicators by school level (primary, secondary, and special schools) are reviewed to contextualize institutional governance and baseline accessibility conditions. A descriptive statistical and regional comparative analysis was conducted to examine (1) the distribution of disability-related support institutions by region and (2) the density of activity support infrastructure (institutions and workforce relative to beneficiaries) across regions. The results indicate substantial regional disparities in support infrastructure, including large gaps in the absolute number of institutions and in supply-demand-adjusted density measures. These disparities suggest that universities' capacity to support students with disabilities may be structurally constrainedenabled by regional support ecosystems. The findings imply that instructional and learning support in higher education should be designed as a regionally linked support system rather than relying solely on institutional capacity. By empirically examining structural and environmental contexts using public data, this study provides foundational evidence for developing integrated, regionally connected support models and informing future policy directions.
Keyword : students with disabilities, instructional and learning support, support infrastructure, regional disparity, public data analysis
http://dx.doi.org/10.29056/jncist.2026.06.10
Graph-Based Portfolio Selection via Filtering and Centrality Measures
Abstract : This study investigates the impact of graph filtering methods and centrality measures on portfolio stock selection. Two graph filtering methods - Minimum Spanning Tree (MST) and Triangulated Maximally Filtered Graph (TMFG)-are combined with three centrality measures-Degree, Eigenvector, and Subgraph centrality-to identify peripheral assets in financial networks. The analysis is conducted on 29 stocksthe Dow Jones Industrial Average with no constituent changes. Portfolios are constructed by selecting the least central assets at varying portfolio sizes(5,10,15, and 20 assets) and applying equal weight allocation. The empirical results indicate that the combination of TMFG and Eigenvector centrality shows the highest Sharpe ratio of 2.65 with 10 selected assets, and maintains high performance with 5 and 15 assets. Degree centrality exhibits improving performance as portfolio size increases across both filtering methods ; for MST with Degree centrality, the Sharpe ratio rises1.24 with 5 assets to 2.55 with 20 assets. These findings clarify the impact of the choice of filtering method and centrality measure on portfolio performance, demonstrating that portfolios composed of peripheral assets achieve higher risk-adjusted returns.
Keyword : Business Analytics, Quantitative Finance, Financial Mathematics, Network Analysis, Risk Management
http://dx.doi.org/10.29056/jncist.2026.06.11
Abstract : This study analyzes research trends in digital twin-based information security through bibliometric network analysis and BERTopic topic modeling. A total of 308 journal articles indexed in Scopus were collected using search terms related to digital twin and security, and their publication and citation trends, journal distribution, country-level collaboration networks, keyword co-occurrence networks, and topic structures were examined. Related publications first appeared in 2018 and grew rapidly through 2025, recording a CAGR of approximately 93.07%. Most highly cited journals were IEEE publications, reflecting the field's close ties to engineering and information and communications technology. In the collaboration network, China and the United States held central positions, with China leading in publication volume and the United States acting as the primary mediator of cross-national connections. Keyword analysis identified digital twin and cybersecurity as the core keywords, with blockchain, IoT, CPS, and AI forming key surrounding clusters, and the 9 clusters suggest that research topics are increasingly differentiated by application domain such as smart cities, ICS, healthcare, and energy infrastructure. BERTopic modeling yielded 11 distinct topics covering smart grid security, CPS and ICS cyber threats, healthcare data privacy, autonomous system security, and vehicular network security. These findings indicate that the field is moving toward technical vulnerability detection, threat management, and policy responses as digital twins are applied across various industries.
Keyword : digital twin, information security, cybersecurity, bibliometric analysis, BERTopic, keyword network
http://dx.doi.org/10.29056/jncist.2026.06.12
Table of Contents