• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.8
• /
• pp.4285-4299
• /
• 2019

Adversarial attack is a technique that causes a malfunction of classification models by adding noise that cannot be distinguished by humans, which poses a threat to a deep learning model. In this paper, we propose an efficient method to detect adversarial images using Gaussian process regression. Existing deep learning-based adversarial detection methods require numerous adversarial images for their training. The proposed method overcomes this problem by performing classification based on the statistical features of adversarial images and clean images that are extracted by Gaussian process regression with a small number of images. This technique can determine whether the input image is an adversarial image by applying Gaussian process regression based on the intermediate output value of the classification model. Experimental results show that the proposed method achieves higher detection performance than the other deep learning-based adversarial detection methods for powerful attacks. In particular, the Gaussian process regression-based detector shows better detection performance than the baseline models for most attacks in the case with fewer adversarial examples.

• The Journal of Information Systems
• /
• v.27 no.4
• /
• pp.149-165
• /
• 2018

Purpose As Korean economy has been sluggish in recent years, firms' interest in publicly financed projects has increased due to the relatively increasing proportion in the economy. Since 1999, publicly financed projects in Korea need to undergo preliminary feasibility study to evaluate economic efficiency and policy quality if they are larger than a certain scale. The benefits of public projects are one of the most important factors in the preliminary feasibility study but are difficult to estimate due to their nature. Design/methodology/approach This study estimates the benefits of the agricultural soil information database, a public database in Korea. The method used in the study is the stated preference method which is formally used in Korea's preliminary feasibility study. Data are collected through surveys and a logit model is constructed to be estimated by the maximum likelihood estimation method. Findings As the first study evaluating a public database, this study can be used as a baseline in all public database projects developed in the future. In addition, this study can contribute to improving the understanding of both private companies and public organizations who are interested in the cost-benefit analysis and estimation procedure for the publicly financed projects.

• The Korean Journal of Applied Statistics
• /
• v.31 no.6
• /
• pp.761-769
• /
• 2018

Simulations are important for survival analyses that deal with censored data. Cox models are widely used in survival analyses, therefore, we investigate how to generate censored data that can simulate the Cox model. Bender et al. (Statistics in Medicine, 24, 1713-1723, 2005) provided a parametric method for generating survival times, but we need to generate censoring times as well as survival times to simulate the censored data. In addition to the parametric method for generating censored data, a nonparametric method is also proposed and applied to a real data set.

• Korean Journal of Remote Sensing
• /
• v.37 no.1
• /
• pp.139-151
• /
• 2021

By the end of December 2020, an earthquake with Mw about 6.4 hit Sisak-Moslavina County, Croatia. The town of Petrinja was the most affected region with major power outage and many buildings collapsed. The damage also affected neighbor countries such as Bosnia and Herzegovina and Slovenia. As a light of this devastating event, a deformation map due to this earthquake could be generated by using remote sensing imagery from Sentinel-1 SAR data. InSAR could be used as deformation map but still affected with noise factor that could problematize the exact deformation value for further research. Thus in this study, 17 SAR data from Sentinel-1 satellite is used in order to generate the multi-temporal interferometry utilize Stanford Method for Persistent Scatterers (StaMPS). Mean deformation map that has been compensated from error factors such as atmospheric, topographic, temporal, and baseline errors are generated. Okada model then applied to the mean deformation result to generate the modeled earthquake, resulting the deformation is mostly dominated by strike-slip with 3 meter deformation as right lateral strike-slip. The Okada sources are having 11.63 km in length, 2.45 km in width, and 5.46 km in depth with the dip angle are about 84.47° and strike angle are about 142.88° from the north direction. The results from this modeling can be used as learning material to understand the seismic activity in the latest 2020 Petrinja, Croatia Earthquake.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.4
• /
• pp.1166-1186
• /
• 2022

Artistic font design has become an integral part of visual media. However, without prior knowledge of the font domain, it is difficult to create distinct font styles. When the number of characters is limited, this task becomes easier (e.g., only Latin characters). However, designing CJK (Chinese, Japanese, and Korean) characters presents a challenge due to the large number of character sets and complexity of the glyph components in these languages. Numerous studies have been conducted on automating the font design process using generative adversarial networks (GANs). Existing methods rely heavily on reference fonts and perform font style conversions between different fonts. Additionally, rather than capturing style information for a target font via multiple style images, most methods do so via a single font image. In this paper, we propose a network architecture for generating multilingual font sets that makes use of geometric structures as content. Additionally, to acquire sufficient style information, we employ multiple style images belonging to a single font style simultaneously to extract global font style-specific information. By utilizing the geometric structural information of content and a few stylized images, our model can generate an entire font set while maintaining the style. Extensive experiments were conducted to demonstrate the proposed model's superiority over several baseline methods. Additionally, we conducted ablation studies to validate our proposed network architecture.

• Journal of Information Processing Systems
• /
• v.18 no.5
• /
• pp.614-627
• /
• 2022

Aspect-based sentiment analysis is to discover the sentiment polarity towards an aspect from user-generated natural language. So far, most of the methods only use the implicit position information of the aspect in the context, instead of directly utilizing the position relationship between the aspect and the sentiment terms. In fact, neighboring words of the aspect terms should be given more attention than other words in the context. This paper studies the influence of different position embedding methods on the sentimental polarities of given aspects, and proposes a position embedding interactive attention network based on a long short-term memory network. Firstly, it uses the position information of the context simultaneously in the input layer and the attention layer. Secondly, it mines the importance of different context words for the aspect with the interactive attention mechanism. Finally, it generates a valid representation of the aspect and the context for sentiment classification. The model which has been posed was evaluated on the datasets of the Semantic Evaluation 2014. Compared with other baseline models, the accuracy of our model increases by about 2% on the restaurant dataset and 1% on the laptop dataset.

• Korean Journal of Organic Agriculture
• /
• v.30 no.2
• /
• pp.151-170
• /
• 2022

The demand for commercial kimchi has increased continuously for the past 20 years due to the increase in eating out demand. Although Korean kimchi industry has expanded significantly, it is still small and a large portion of domestic demand is dependent on Chinese kimchi. Chinese kimchi imports has markedly increased over the last 20 years. However, kimchi imports from China in 2021 significantly reduced due to the recently released video showing a naked man making Kimchi. Korean government has decided to apply HACCP to all imported Kimchi from October 2021 in order to improve the safety of imported kimchi. This study analyzed the effect of changes in the amount of kimchi imports due to the introduction of HACCP on the kimchi industry by using a dynamic partial equilibrium model. According to the analysis result, if imports decreased by 20% compared to the Baseline, domestic kimchi production increased from 1.8% to a maximum of 4.8%, but kimchi consumption decreased from 3.1% to 5.2%. In particular, consumption away from home decreased from 3.3% to 5.7%. It is expected that the results of this study would be used as useful data in the decision-making process of market participants and policy makers related to the kimchi industry.

• Journal of Information Technology Services
• /
• v.21 no.4
• /
• pp.75-90
• /
• 2022

From the past to the present, reviews have had much influence on consumers' purchasing decisions. Companies are making various efforts, such as introducing a review incentive system to increase the number of reviews. Recently, as various types of reviews can be left, reviews have begun to be recognized as interesting new content. This way, reviews have become essential in creating loyal customers. Therefore, research and utilization of reviews are being actively conducted. Some studies analyze reviews to discover customers' needs, studies that upgrade recommendation systems using reviews, and studies that analyze consumers' emotions and attitudes through reviews. However, research that predicts the future using reviews is insufficient. This study used a dataset consisting of two reviews written in pairs with differences in usage periods. In this study, the direction of consumer product evaluation is predicted using KoBERT, which shows excellent performance in Text Deep Learning. We used 7,233 reviews collected to demonstrate the excellence of the proposed model. As a result, the proposed model using the review text and the star rating showed excellent performance compared to the baseline that follows the majority voting.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.10
• /
• pp.3230-3255
• /
• 2022

Causality mining in NLP is a significant area of interest, which benefits in many daily life applications, including decision making, business risk management, question answering, future event prediction, scenario generation, and information retrieval. Mining those causalities was a challenging and open problem for the prior non-statistical and statistical techniques using web sources that required hand-crafted linguistics patterns for feature engineering, which were subject to domain knowledge and required much human effort. Those studies overlooked implicit, ambiguous, and heterogeneous causality and focused on explicit causality mining. In contrast to statistical and non-statistical approaches, we present Bidirectional Encoder Representations from Transformers (BERT) integrated with Multi-level Feature Networks (MFN) for causality recognition, called BERT+MFN for causality recognition in noisy and informal web datasets without human-designed features. In our model, MFN consists of a three-column knowledge-oriented network (TC-KN), bi-LSTM, and Relation Network (RN) that mine causality information at the segment level. BERT captures semantic features at the word level. We perform experiments on Alternative Lexicalization (AltLexes) datasets. The experimental outcomes show that our model outperforms baseline causality and text mining techniques.

• Journal of Korea Trade
• /
• v.24 no.2
• /
• pp.68-82
• /
• 2020

Purpose - This study investigates the impact of inland port development in China on the promotion of bilateral trade flows between China and South Korea. Design/methodology - The probable association between the establishment of inland ports and Sino-Korea trade was estimated using gravity models. In this regards, two sets of data were collected. The first dataset consists of the baseline variables of a gravity model, while the second one includes variables of logistics infrastructure development. The indicators of logistics infrastructure development include inland ports, the amount of government expenditure on transport infrastructure, the lengths of roads and railways, the number of trucks and the number of logistics industry workforce. Findings - The results show that inland port development has a positive impact on facilitating bilateral trade between China and South Korea. However, the positive association holds only for Chinese regions with a large trade volume and a proximity to seaports. In other regions, the impact of inland ports is not statistically significant. Originality/value - To the best knowledge of the authors, this study is the first attempt to explore the economic impact of inland ports in China. In addition, the findings in this paper provide both policy and managerial implications for the future development of inland ports, such as the strategic location of inland ports and integrated intermodal operations.