• Journal of the Korea Society of Computer and Information
• /
• v.28 no.10
• /
• pp.67-76
• /
• 2023

Although deep learning models are making innovative achievements in the field of computer vision, the problem of vulnerability to adversarial examples continues to be raised. Adversarial examples are attack methods that inject fine noise into images to induce misclassification, which can pose a serious threat to the application of deep learning models in the real world. In this paper, we propose a model that detects adversarial examples using differences in predictive values between edge-learned classification models and underlying classification models. The simple process of extracting the edges of the objects and reflecting them in learning can increase the robustness of the classification model, and economical and efficient detection is possible by detecting adversarial examples through differences in predictions between models. In our experiments, the general model showed accuracy of {49.9%, 29.84%, 18.46%, 4.95%, 3.36%} for adversarial examples (eps={0.02, 0.05, 0.1, 0.2, 0.3}), whereas the Canny edge model showed accuracy of {82.58%, 65.96%, 46.71%, 24.94%, 13.41%} and other edge models showed a similar level of accuracy also, indicating that the edge model was more robust against adversarial examples. In addition, adversarial example detection using differences in predictions between models revealed detection rates of {85.47%, 84.64%, 91.44%, 95.47%, and 87.61%} for each epsilon-specific adversarial example. It is expected that this study will contribute to improving the reliability of deep learning models in related research and application industries such as medical, autonomous driving, security, and national defense.

• Journal of Service Research and Studies
• /
• v.13 no.4
• /
• pp.191-205
• /
• 2023

Artificial Intelligence (AI), especially in the domain of text-generative services, has witnessed a significant surge, with forecasts indicating the AI-as-a-Service (AIaaS) market reaching a valuation of $55.0 Billion by 2028. This research set out to explore the quality dimensions characterizing synthetic text media software, with a focus on four key players in the industry: ChatGPT, Writesonic, Jasper, and Anyword. Drawing from a comprehensive dataset of over 4,000 reviews sourced from a software evaluation platform, the study employed the Latent Dirichlet Allocation (LDA) topic modeling technique using the Gensim library. This process resulted the data into 11 distinct topics. Subsequent analysis involved comparing these topics against established AI service quality dimensions, specifically AICSQ and AISAQUAL. Notably, the reviews predominantly emphasized dimensions like availability and efficiency, while others, such as anthropomorphism, which have been underscored in prior literature, were absent. This observation is attributed to the inherent nature of the reviews of AI services examined, which lean more towards semantic understanding rather than direct user interaction. The study acknowledges inherent limitations, mainly potential biases stemming from the singular review source and the specific nature of the reviewer demographic. Possible future research includes gauging the real-world implications of these quality dimensions on user satisfaction and to discuss deeper into how individual dimensions might impact overall ratings.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.11
• /
• pp.29-42
• /
• 2023

In this paper, we propose a method to enhance the prediction accuracy of solar irradiance for three major South Korean cities: Seoul, Busan, and Incheon. Our method entails the development of five generative models-vanilla GAN, CTGAN, Copula GAN, WGANGP, and TVAE-to generate independent variables that mimic the patterns of existing training data. To mitigate the bias in model training, we derive values for the dependent variables using random forests and deep neural networks, enriching the training datasets. These datasets are integrated with existing data to form comprehensive solar irradiance prediction models. The experimentation revealed that the augmented datasets led to significantly improved model performance compared to those trained solely on the original data. Specifically, CTGAN showed outstanding results due to its sophisticated mechanism for handling the intricacies of multivariate data relationships, ensuring that the generated data are diverse and closely aligned with the real-world variability of solar irradiance. The proposed method is expected to address the issue of data scarcity by augmenting the training data with high-quality synthetic data, thereby contributing to the operation of solar power systems for sustainable development.

• Analyses & Alternatives
• /
• v.7 no.3
• /
• pp.27-59
• /
• 2023

This study aims to serve as a critical comparison of the currently controversial 'new cold war' discourse. It took three triggers for the 'new cold war' discourse to emerge as a major issue in the media and academia and to have real political impact. With the launch of China's 'Belt and Road' project and Russia's annexation of Crimea leading to the 'Ukraine crisis,' the 'new cold war' discourse has begun to take shape. Trump's U.S.-China trade spat has brought the 'new cold war' debate to the forefront. The 'new cold war' debate is currently being intensified by the Biden administration's framing of "democracy versus authoritarianism" and Putin's invasion of Ukraine. Currently, there is no consensus among scholars on whether the controversial 'new cold war' is a new version, or a continuation of the historically defined concept of the Cold War. The term 'New Cold War' is less of an analytical concept and more of a topical term that has yet to achieve analytical status, let alone a theoretical validation and systematization, and the related debate remains at the level of assertion or discourse. Through this comparative analysis, I will argue that the ongoing discourse of the 'New Cold War' does not have the instrumental explanatory power to analyze the transitional phenomena of the world order today.

• Journal of Animal Science and Technology
• /
• v.65 no.6
• /
• pp.1124-1150
• /
• 2023

This study aimed to analyze the leading research materials and research trends related to livestock food in Asia in recent years and propose future research agendas to ultimately contribute to the development of related livestock species. On analyzing more than 200 relevant articles, a high frequency of studies on livestock species and products with large breeding scales and vast markets was observed. Asia possesses the largest pig population and most extensive pork market, followed by that of beef, chicken, and milk; moreover, blood and egg markets have also been studied. Regarding research keywords, "meat quality" and "probiotics" were the most common, followed by "antioxidants", which have been extensively studied in the past, and "cultured meat", which has recently gained traction. The future research agenda for meat products is expected to be dominated by alternative livestock products, such as cultured and plant-derived meats; improved meat product functionality and safety; the environmental impacts of livestock farming; and animal welfare research. The future research agenda for dairy products is anticipated to include animal welfare, dairy production, probiotic-based development of high-quality functional dairy products, the development of alternative dairy products, and the advancement of lactose-free or personalized dairy products. However, determining the extent to which the various research articles' findings have been applied in real-world industry proved challenging, and research related to animal food laws and policies and consumer surveys was lacking. In addition, studies on alternatives for sustainable livestock development could not be identified. Therefore, future research may augment industrial application, and multidisciplinary research related to animal food laws and policies as well as eco-friendly livestock production should be strengthened.

• Smart Structures and Systems
• /
• v.32 no.4
• /
• pp.267-279
• /
• 2023

Data modelling and interpretation for structural health monitoring (SHM) field data are critical for evaluating structural performance and quantifying the vulnerability of infrastructure systems. In order to improve the data modelling accuracy, and extend the application range from data regression analysis to out-of-sample forecasting analysis, an improved most likely heteroscedastic Gaussian process (iMLHGP) methodology is proposed in this study by the incorporation of the outof-sample forecasting algorithm. The proposed iMLHGP method overcomes this limitation of constant variance of Gaussian process (GP), and can be used for estimating non-stationary typhoon-induced response statistics with high volatility. The first attempt at performing data regression and forecasting analysis on structural responses using the proposed iMLHGP method has been presented by applying it to real-world filed SHM data from an instrumented cable-stay bridge during typhoon events. Uncertainty quantification and correlation analysis were also carried out to investigate the influence of typhoons on bridge strain data. Results show that the iMLHGP method has high accuracy in both regression and out-of-sample forecasting. The iMLHGP framework takes both data heteroscedasticity and accurate analytical processing of noise variance (replace with a point estimation on the most likely value) into account to avoid the intensive computational effort. According to uncertainty quantification and correlation analysis results, the uncertainties of strain measurements are affected by both traffic and wind speed. The overall change of bridge strain is affected by temperature, and the local fluctuation is greatly affected by wind speed in typhoon conditions.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.12
• /
• pp.41-48
• /
• 2023

In this paper, we propose an analysis of recent trends in the NFT (Non-Fungible Token) industry using topic modeling techniques, focusing on their universal application across various industrial fields. For this study, patent data was utilized to understand industry trends. We collected data on 371 domestic and 454 international NFT-related patents registered in the patent information search service KIPRIS from 2017, when the first NFT standard was introduced, to October 2023. In the preprocessing stage, stopwords and lemmas were removed, and only noun words were extracted. For the analysis, the top 50 words by frequency were listed, and their corresponding TF-IDF values were examined to derive key keywords of the industry trends. Next, Using the LDA algorithm, we identified four major latent topics within the patent data, both domestically and internationally. We analyzed these topics and presented our findings on NFT industry trends, underpinned by real-world industry cases. While previous review presented trends from an academic perspective using paper data, this study is significant as it provides practical trend information based on data rooted in field practice. It is expected to be a useful reference for professionals in the NFT industry for understanding market conditions and generating new items.

• Journal of Gastric Cancer
• /
• v.23 no.4
• /
• pp.609-621
• /
• 2023

Purpose: Determination of optimal treatment strategies for HER2-positive advanced gastric cancer (AGC) in randomized trials is necessary despite difficulties in direct comparison between trastuzumab deruxtecan (T-DXd) and nivolumab as third or later-line treatments. Materials and Methods: This single-institution, retrospective study aimed to describe the real-world efficacy and safety of T-DXd and nivolumab as ≥ third line treatments for HER2-positive AGC between March 2016 and May 2022. Overall, 58 patients (median age, 64 years; 69% male) were eligible for the study (T-DXd group, n=20; nivolumab group, n=38). Results: Most patients exhibited a HER2 3+ status (72%) and presented metastatic disease at diagnosis (66%). The response rates of 41 patients with measurable lesions in the T-DXd and nivolumab groups were 50% and 15%, respectively. The T-DXd and nivolumab groups had a median progression-free survival of 4.8 months (95% confidence interval [CI], 3.3, 7.0) and 2.3 months (95% CI, 1.5, 3.5), median overall survival (OS) of 10.8 months (95% CI, 6.9, 23.8) and 11.7 months (95% CI, 7.6, 17.1), and grade 3 or greater adverse event rates of 50% and 2%, respectively. Overall, 64% patients received subsequent treatment. Among 23 patients who received both regimens, the T-DXd-nivolumab and nivolumab-T-DXd groups had a median OS of 14.0 months (95% CI, 5.0, not reached) and 19.3 months (95% CI, 9.5, 25.1), respectively. Conclusions: T-DXd and nivolumab showed distinct efficacy and toxicity profiles as ≥ third line treatments for HER2-positive AGC. Considering the distinct features of each regimen, they may help clinicians personalize optimal treatment approaches for these patients.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.5
• /
• pp.597-603
• /
• 2023

The paradigm of military operations has evolved from platform-centric warfare to network-centric warfare and further to information-centric warfare, driven by advancements in information technology. In recent years, with the development of cutting-edge technologies such as big data, artificial intelligence, and the Internet of Things (IoT), military operations are transitioning towards knowledge-centric warfare (KCW), based on artificial intelligence. Consequently, the military places significant emphasis on integrating advanced information and communication technologies (ICT) to establish reliable C4I (Command, Control, Communication, Computer, Intelligence) systems. This research emphasizes the need to apply data mining techniques to analyze and evaluate various aspects of C4I systems, including enhancing combat capabilities, optimizing utilization in network-based environments, efficiently distributing information flow, facilitating smooth communication, and effectively implementing knowledge sharing. Data mining serves as a fundamental technology in modern big data analysis, and this study utilizes it to analyze real-world cases and propose practical strategies to maximize the efficiency of military command and control systems. The research outcomes are expected to provide valuable insights into the performance of C4I systems and reinforce knowledge-centric warfare in contemporary military operations.

• Journal of Intelligence and Information Systems
• /
• v.29 no.4
• /
• pp.31-49
• /
• 2023

With the growth of the food-catering industry, consumer preferences and the number of dine-in restaurants are gradually increasing. Thus, personalized recommendation services are required to select a restaurant suitable for consumer preferences. Previous studies have used questionnaires and star-rating approaches, which do not effectively depict consumer preferences. Online reviews are the most essential sources of information in this regard. However, previous studies have aggregated online reviews into long documents, and traditional machine-learning methods have been applied to these to extract semantic representations; however, such approaches fail to consider the surrounding word or context. Therefore, this study proposes a novel review textual-based restaurant recommendation model (RT-RRM) that uses deep learning to effectively extract consumer preferences from online reviews. The proposed model concatenates consumer-restaurant interactions with the extracted high-level semantic representations and predicts consumer preferences accurately and effectively. Experiments on real-world datasets show that the proposed model exhibits excellent recommendation performance compared with several baseline models.