eISSN : 3022-7011
ISSUER : KIPS
 
After the Korea Information Processing Society (KIPS) Transactions journal was founded in 1994, it was reorganized into the KIPS Transactions: Computer and Communication Systems(2287-5891/2734-049X ) and the KIPS Transactions: Software and Data Engi neering(2287-5905/2734-0503) in 2012. Through the KIPS official meeting on January 8th, 2024, the new KIPS Transaction journal was founded by integrating two KIPS Journals, KIPS Transactions: Computer and Communication Systems and KIPS Transactions: Software and Data Engineering. The new journal aims to realize social value and contribute to the development of South Korea’s science and technology with support from the lottery fund of the Ministry of Strategy and Finance and the science/technology promotion fund of the Ministry of Science and ICT. It is indexed in the Korea Science Academic Database, Korea Citation Index (KCI), and EBSCO.

HighlightsMore

Diffusion-based Audio-to-Visual Generation for High-Quality Bird Images

Adel Toleubekova  Joo Yong Shim  XinYu Piao  Jong-Kook Kim
Accurately identifying bird species from their vocalizations and generating corresponding bird images is still a challenging task due to limited training data and environmental noise in audio data. To address this limitation, this paper introduces a...

Bambda: A Framework for Preventing Function Invocation Condition-Based Attacks in Serverless Environments

Shin Chang Hee  Lee Seung Soo
Serverless computing is rapidly emerging as a new paradigm in cloud computing, offering automatic scalability, cost efficiency, and ease of operation. However, its two core characteristics—IAM-based privilege management and event-driven execution—ca...

PEFT Methods for Domain Adaptation

Lee You Jin  Yoon Kyung Koo  Chung Woo Dam
This study analyzed that the biggest obstacle in deploying Large Language Models (LLMs) in industrial settings is incorporating domain specificity into the models. To mitigate this issue, the study compared model performance when domain knowledge wa...

Label Differential Privacy Study for Privacy Protection in Multimodal Contrastive Learning Model

Youngseo Kim  Minseo Yu  Younghan Lee  Ho Bae
Recent advancements in multimodal deep learning have garnered significant attention from both academia and industry due to their exceptional accuracy and ability to learn rich knowledge representations. In particular, contrastive learning based appr...

Latest Publication   (Vol. 14, No. 9, Sep.  2025)

Windows Malware Family Dataset Construction and Classification
Kim Tae Young  Doo Seop Choi  Eul Gyu Im
Malware family classification is a critical task for enhancing the efficiency of threat analysis and enabling rapid response strategies. However, accurate classification remains challenging due to behavioral similarities among different families and the ambiguous boundaries between variants. Moreover, most previous studies rely on outdated datasets, limiting their ability to reflect the latest trends in malware. To address these issues, this study constructs a new dataset of 3,357 Windows malware samples collected in 2024, with high label reliability ensured through cross-verification. Using this dataset, we applied a hybrid feature approach that combines static and dynamic features to a Random Forest model, achieving a maximum classification accuracy of 92.14%. An analysis of misclassified samples revealed that classification errors were often caused by shared API call sequences among certain malware families, leading to confusion, or by premature termination of malware execution, which hindered the collection of sufficient dynamic information. Based on these findings, we suggest the need for more sophisticated behavior-based feature extraction and improvements to the dynamic analysis environment to prevent early termination. This study is expected to make a practical contribution to enhancing the accuracy and reliability of future malware detection systems.
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 651-661, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.651
Windows Malware Family Classification Dataset Construction Static and Dynamic analysis Hybrid Approach
Machine Learning-Based Cooperative Detection Method for Effectively Mitigating Availability Attacks in Wireless Network Environments
Min-Ji Cho  So-Eun Jeon  Il-Gu Lee
With the rapid increase in the number of Internet of Things (IoT) devices such as smartphones, wearables, and smart appliances, data transmission over wireless networks has significantly grown. Consequently, availability attacks targeting wireless networks—such as jamming, flooding, blackhole, and grayhole attacks—are also rising. Conventional attack detection methods rely on statistical indicators such as packet delivery rate, received signal strength, and timestamps. However, these approaches depend heavily on resource-constrained IoT devices, resulting in low detection accuracy and degradation of overall network performance. Recently, machine learning-based local detection methods have been explored, but their high computational complexity makes them unsuitable for deployment on lightweight devices. This paper proposes a cooperative detection framework that distributes machine learning-based training and detection tasks between resource-rich access points (APs) and lightweight IoT devices. Experimental results show that the proposed model improves detection accuracy by an average of 38% and data transmission success rate by 17.75% compared to conventional local learning-based detection models.
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 662-667, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.662
Wireless network security IoT security Cooperative detection Machine learning. Availability attack
Improving Object Detection via Lightweight Cross-Attention-based Semantic Alignment
Hyungseop Lee  Jiho Lee  Woochul Kang
Accurate detection of objects with varying scales requires effective multi-scale feature representation learning. To this end, most modern object detectors adopt Feature Pyramid Network (FPN)-based feature fusion strategies. However, due to differences in semantic granularity and information content across feature levels, direct fusion often results in semantic misalignment, leading to increased false positives and limited detection performance. In this paper, we propose a lightweight cross-attention-based semantic alignment module that aligns adjacent feature levels prior to fusion. The module leverages semantically weak low-level features as queries and semantically rich high-level features as keys and values, enabling effective modeling of inter-level semantic relationships. To ensure computational efficiency and real-time applicability, the sequence length is constrained based on the lowest-resolution feature map. We integrate the proposed module into both conventional and real-time object detectors and evaluate it on the MS COCO and PASCAL VOC datasets. Experimental results demonstrate consistent improvements in AP and AP50 metrics, validating the effectiveness and generality of our approach.
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 668-676, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.668
Deep Learning Object Detection Multi-scale feature fusion Cross-attention Real-Time Inference
TokenROC: A ROC-Based Evaluation Method for XAI Techniques in Software Vulnerability Token Detection
Jaehong Kim  Sebeom Cheon  Sungmin Han  Sangkyun Lee
Recently, deep learning-based models have advanced automated software vulnerability detection. However, most provide only coarse information, such as vulnerability presence or affected code lines, and cannot precisely identify the actual vulnerable tokens. To address this, explainable artificial intelligence (XAI) techniques have been applied to generate token-level explanations. Yet, systematic methods for quantitatively evaluating their accuracy remain lacking. This paper proposes TokenROC, a framework that measures how accurately XAI methods detect security-critical tokens. It compares attribution maps with ground-truth labels and evaluates performance using ROC curves and AUC scores. We apply TokenROC to VulBERTa and VulDeBERT using four XAI methods: TIS, AttCAT, Rollout, and Grad-SAM. Results are also compared with the performance of the StagedVulBERT model. Experiments show that XAI methods can meaningfully identify vulnerable tokens. TokenROC offers a practical framework for evaluating and improving explainability-driven vulnerability analysis.
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 677-686, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.677
Software Vulnerability Detection Vulnerable Token Identification Explainable Artificial Intelligence
A Study on Optimizing Korean Performance of Quantized Multilingual LLMs through Language-Specific Neuron Preservation
Jaeyoung Lee  Jin Young Choi
Quantization of large language models (LLMs) is an effective technique for reducing model size and computational overhead. However, most existing methods are designed with an English-centric perspective, which leads to performance degradation in non-English languages. To address this limitation, this study proposes an improved quantization approach that enhances the AWQ framework by identifying Korean-specific neurons using the Language Activation Probability Entropy (LAPE) metric and preserving their weights during quantization. Furthermore, we introduce an optimization strategy that focuses the analysis on deeper layers where language-specific activation patterns tend to emerge, thereby improving both computational efficiency and model performance. Ex perimental results demonstrate that the proposed method significantly improves Korean performance by up to 17.9% on the Llama-3.2-3B-Instruct model, while keeping the model size virtually unchanged.
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 687-694, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.687
LLM quantization Artificial intelligence Natural Language Processing
Generative AI-based Explainable Personalized Hybrid Skincare Recommendation System
Jin Hui Jeong  Seo Young Kim  Seon Ho Choi  Hyon Hee Kim
Recently, in the cosmetics market, there is a remarkable trend of smart consumption in which consumers directly compare and analyze products suitable for their skin characteristics. However, due to the difficulty of interpreting ingredients and the limitations of information, consumers need reliable customized product recommendations and intuitive descriptions of products in the purchasing decision-making process. Therefore, this study designed and implemented a hybrid AI recommendation system that combines item-based collaborative filtering and content-based filtering by utilizing the user’s purchase history, rating pattern, product review, and ingredient information. In addition, by combining Generative AI, the system generates ingredient-based reason for recommendations, review-based keywords, and product advertisement images, thereby enhancing the explainability of recommendation results and ensuring user satisfaction. Our results showed that the proposed hybrid recommendation system achieved about 19% improvement in MAP@8(0.5199) over the conventional single filtering model and exhibiting notable advantages in both precision and explainability of recommendation. This study is expected to contribute to advancement of personalized services, the strengthening of platform competitiveness, and the improvement of purchase conversion rates in future digital commerce environments
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 695-703, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.695
Generative AI prompt engineering hybrid recommendation system Collaborative Filtering Content-based Filtering Explainable Recommendation Personalized Recommendation
xDBTune: eXplainable Database Tuning Framework
Okjoo Choi  Wonsun Shin
Database systems serve as critical infrastructure for efficiently processing large-scale data across various industries such as finance and manufacturing. However, as data complexity increases and query workloads become more diverse, database performance optimization has emerged as an essential task to ensure system stability and service quality. Recently, with the advancement of artificial intelligence, machine learning–based automated database tuning techniques have been proposed as alternatives. Nevertheless, most existing approaches rely on black-box models, leading to a fundamental limitation in the interpretability of tuning results. To address this issue, this study proposes xDBTune, a database tuning framework that integrates explainable artificial intelligence (XAI) techniques. xDBTune automatically recommends optimized tuning parameter configurations based on machine learning–based performance prediction, and employs the SHAP (Shapley Additive Explanations) algorithm to visually and quantitatively explain the impact of each parameter on performance. This enables both DBAs and general users to clearly understand and verify the tuning results with confidence. The proposed framework was evaluated using the TPC-H benchmark dataset in a MySQL environment. Experimental results show that the recommended tuning configuration reduced the average query execution time by approximately 22.2% compared to the default setting. Furthermore, SHAP-based visualizations effectively conveyed the contribution of each tuning parameter, demonstrating the framework’s ability to improve both interpretability and user trust in database tuning.
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 704-712, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.704
database database management system DB Tuning XAI(Explainable AI) SHAP
Sex-Specific Prompt Engineering for OSA Severity Classification Using Large Language Models
Seungyeon Ryu  Hyeon Jin Kim  Younghan Lee
Obstructive Sleep Apnea (OSA) is a prevalent but underdiagnosed sleep disorder associated with serious health risks. Although polysomnography (PSG) is the diagnostic gold standard, its cost and clinical burden limit scalability. This study explores the use of Large Language Models (LLMs), particularly the lightweight GPT-4o-mini, to classify OSA severity (Normal, Moderate, Severe) based solely on non-invasive demographic and sleep questionnaire data, without access to PSG features. We develop a prompt engineering pipeline that incorporates sex-specific feature importance extracted from Random Forest classifiers to design three types of prompts: Basic, Role-Based, and Chain-of-Thought (CoT). Experimental results show that Chain-of-Thought (CoT) prompts with feature emphasis tailored to male and female subgroups achieve F1 scores of up to 0.60–0.62, outperforming non-sex-specific (generic) prompts and matching the performance of Random Forest baselines. These findings highlight the effectiveness of sex-aware, few-shot prompting in improving diagnostic performance. Our results suggest that prompt-based LLMs can serve as interpretable, low-resource alternatives to traditional supervised models, with potential for scalable and equitable OSA screening.
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 713-721, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.713
Obstructive Sleep Apnea (OSA) prompt engineering Chain-of-Thought Prompting Large Language Model
Fast Booting for BMC
Kyung Eun Oh  Ji Man Hong
BMC is an essential embedded system responsible for remote management and monitoring of servers and data centers, where improved boot speed is crucial for rapid recovery. This paper analyzes bottlenecks in the booting process of conventional BMC operating systems and proposes optimizations through kernel light-weighting, bootloader optimization, and hibernation-based boot techniques. Additionally, it compares the performance of compression methods, applies the optimal scheme, and verifies it experimentally. A comparative analysis between the optimized system and the conventional system demonstrates a significant reduction in boot time. Through the proposed Fast Booting technique, this paper improves the boot performance of BMC systems and further enhances system availability by reducing maintenance and recovery time.
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 722-724, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.722
Baseboard Management Controller Embedded Linux Kernel Optimization Hibernation
Scheduling Performance Analysis of SMP-based RTEMS
Jeon Hyeon Soo  Park Jun Yong  Kim Hyeong Ho  Jang Joon Hyouk  Jung Jin Man
With the proliferation of multi-core embedded systems, the SMP scheduling policy of real-time operating systems significantly impacts system performance. In high-reliability environments such as aerospace and defense, it is essential to optimize scheduling configurations under limited hardware resources. While previous studies have discussed the general characteristics of Global and Partitioned Queues, detailed performance comparisons remain insufficient. This study simulates the GR740 architecture using SIS and compares the performance of the two scheduling approaches in the RTEMS environment. The experimental setup uses a 4-core SMP architecture, where the operating system recognizes each core as an independent CPU for scheduling. The results show that for cache-friendly tasks, the Partitioned Queue demonstrates superior performance due to better cache locality, whereas for CPU-bound tasks that require load balancing, the Global Queue performs better. These findings can serve as a guideline for selecting appropriate scheduling strategies in SMP-based RTEMS systems according to application requirements.
The Transactions of the Korea Information Processing Society, Vol. 14, No. 9, pp. 725-731, Sep. 2025
https://doi.org/10.3745/TKIPS.2025.14.9.725
RTEMS RTOS SMP Scheduling Scheduling Performance Analysis

Forms

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)

Old Journals

Indexing

All publications of TKIPS is indexed in DOI, EBSCO, Google Scholar, Crossref, and CrossCheck.

TKIPS is also selected as the Journal for Accreditation by NRF (National Research Foundation of Korea).

Related Journals