Kwang-Seong Shin

Prof. Kwang-Seong Shin


Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 January 2024 in Applied Sciences
Verified
Journal Article
Open access
PDF file
CSChuanxiang Song
Chuanxiang Song
SSSeong-Yoon Shin
Seong-Yoon Shin
KSKwang Seong Shin
Kwang Seong Shin
This study introduces a novel approach named the Dynamic Feedback-Driven Learning Optimization Framework (DFDLOF), aimed at personalizing educational pathways through machine learning technology. Our findings reveal that this framework significantly enhances student engagement and learning effectiveness by providing real-time feedback and personalized instructional content tailored to individual learning needs. This research demonstrates the potential of leveraging advanced technology to create more effective and individualized learning environments, offering educators a new tool to support each student’s learning journey. The study thus contributes to the field by showcasing how personalized education can be optimized using modern technological advancements.

21 November 2023 in Applied Sciences
Verified
Journal Article
Open access
PDF file
CSChuanxiang Song
Chuanxiang Song
SSSeong-Yoon Shin
Seong-Yoon Shin
KSKwang Seong Shin
Kwang Seong Shin
With the widespread application of virtual reality (VR) in education, optimizing foreign language learning in VR has become a focal point of research. This paper introduces a comprehensive theoretical framework (VR-CCL) based on constructivism and cognitive load theory to enhance foreign language learning in VR. Through a literature review, we explore the applications of VR in education, foreign language learning theories, and prior works on technology-assisted language learning. We further detail the three main components of the VR-CCL framework and validate its effectiveness through two case studies: Duolingo VR and Rosetta Stone VR. Finally, we discuss the strengths and limitations of the framework and its implications for educators and developers.

28 June 2023 in Applied Sciences
Verified
Journal Article
Open access
PDF file
CSChuanxiang Song
Chuanxiang Song
SSSeong-Yoon Shin
Seong-Yoon Shin
KSKwang Seong Shin
Kwang Seong Shin
This study presents an insightful examination of the conceptual and practical facets of the Metaverse by establishing a novel theoretical framework underpinned by an empirical case study of the Sandbox platform. Anchored in the principles of legality, virtual-reality integration, technological affinity, and community-driven innovation, the paper elucidates the inherent characteristics and potentialities of the Metaverse. Through meticulous research, the paper investigates the antecedents and evolution of the Metaverse, postulating an open, decentralized, and self-regulating ecosystem predicated on user-generated content and engagement. Furthermore, an in-depth case study of the Sandbox elucidates the practical applications, challenges, and opportunities associated with the operationalization of the Metaverse. The study showcases how avant-garde technologies such as blockchain, virtual reality, and artificial intelligence are instrumental in fostering immersive experiences, safeguarding virtual asset ownership, and facilitating tailored services. Moreover, the paper accentuates the indispensable role of community engagement and continuous innovation in cultivating a flourishing Metaverse environment. The analysis exposes that the burgeoning development of the Metaverse is intrinsically linked to the amalgamation of the virtual and the tangible, extending the frontiers of the digital economy. While shedding light on the virtues of the Metaverse, the study recognizes its nascent state and encourages further scholarly inquiry to comprehend and navigate its complexities. This research contributes significantly to the academic and practical understanding of the Metaverse, serving as a cornerstone for future investigations and technological advancements in this paradigm-shifting domain.

18 October 2024 in Applied Sciences
Verified
Journal Article
Open access
PDF file
YXYina Xia
Yina Xia
SSSeong-Yoon Shin
Seong-Yoon Shin
KSKwang Seong Shin
Kwang Seong Shin
This study introduces the Data-Driven Personalized Learning Model (DDPLM), a sophisticated framework designed to enhance foreign language acquisition through the integration of big data analytics. Implemented within the educational platforms Edmodo and Duolingo, DDPLM utilizes real-time data processing to tailor learning paths and content dynamically to individual learner needs. Our findings indicate significant improvements in language learning efficiency, engagement, and retention. The model’s adaptability across different digital environments showcases its potential scalability and effectiveness in various educational contexts. Additionally, the research addresses the critical role of personalized feedback and adaptive challenges in maintaining learner motivation and promoting deeper linguistic comprehension. The outcomes suggest that DDPLM significantly transforms traditional language education, making it more personalized, efficient, and aligned with individual learning preferences.

2 October 2020 in Electronics
Verified
Journal Article
Open access
PDF file
KSKwang Seong Shin
Kwang Seong Shin
HKHowon Kim
Howon Kim
JLJeong-Gon Lee
Jeong-Gon Lee
Dongsik JoDongsik Jo
Dongsik Jo
With continued technological innovations in the fields of mixed reality (MR), wearable type MR devices, such as head-mounted display (HMD), have been released and are frequently used in various fields, such as entertainment, training, education, and shopping. However, because each product has different parts and specifications in terms of design and manufacturing process, users feel that the virtual objects overlaying real environments in MR are visualized differently, depending on the scale and color used by the MR device. In this paper, we compare the effect of scale and color parameters on users’ perceptions in using different types of MR devices to improve their MR experiences in real life. We conducted two experiments (scale and color), and our experimental study showed that the subjects who participated in the scale perception experiment clearly tended to underestimate virtual objects, in comparison with real objects, and overestimated color in MR environments.

20 February 2026 in Electronics
Verified
Journal Article
Open access
PDF file
DWDong-Hyun Won
Dong-Hyun Won
KSKwang Seong Shin
Kwang Seong Shin
SYSungkwan Youm
Sungkwan Youm
Class imbalance in tabular datasets poses a challenge for machine learning classification tasks, often leading to biased models that underperform in predicting minority class instances. This study presents a comparative analysis of synthetic data generation methods for addressing class imbalance in tabular data. We evaluate four augmentation approaches—Synthetic Minority Over-sampling Technique (SMOTE), Gaussian Copula, Tabular Variational Autoencoder (TVAE), and Conditional Tabular Generative Adversarial Network (CTGAN)—using the University of California Irvine (UCI) Bank Marketing dataset, which exhibits a class imbalance ratio of approximately 7.88:1. Our experimental framework assesses each method across three dimensions: statistical fidelity to the original data distribution evaluated through four complementary metrics (marginal numerical similarity, categorical distribution similarity, correlation structure preservation, and Kolmogorov–Smirnov test), machine learning utility measured through classification performance, and minority class detection capability. Results indicate that all augmentation methods achieved statistically significant improvements over the baseline (p<0.05). SMOTE achieved the highest recall (54.2%, a 117.6% relative improvement over the baseline) and F1-Score (0.437, +22.4% over the baseline) for minority class detection, while Gaussian Copula provided the highest composite fidelity score (0.930) with competitive predictive performance. A weak negative correlation (ρ=−0.30) between composite fidelity and classification performance was observed, suggesting that higher statistical fidelity does not necessarily translate to better downstream task performance. Deep learning-based methods (TVAE, CTGAN) showed statistically significant improvements over the baseline (recall: +58% to +63%) but underperformed compared to simpler methods under default configurations, suggesting the need for larger training samples or more extensive hyperparameter tuning. These findings offer reference points for practitioners working with moderately imbalanced tabular data with limited minority class samples, supporting the selection of generation strategies based on specific requirements regarding data fidelity and classification objectives.