Search

The pulmonate snail Melampus sincaporensis L. Pfeiffer, 1855 (Gastropoda: Ellobiidae) lives in extreme environments such as salt marshes with limited oxygen. Here, we characterized the complete mitochondrial genome of M. sincaporensis from South Korea. It is 14,962 bp in length and includes 13 protein-coding genes (PCGs), two genes, 21 tRNA genes lacking one tRNA-Ser, and two A + T rich regions. Among these 36 genes, 24 were encoded on the heavy strand and 12 on the light strand. A maximum likelihood tree constructed using nucleotide sequences of the 13 PCGs did not support the monophyly of Ellobiidae. This study could provide useful information for exploring phylogenetic relationships among ellobiids and their related species.

The chiton (Polyplacophora) occupies a significant position in molluscan evolutionary history as one of the most primitive groups within the phylum Mollusca. Acanthochitona defilippii (Tapparone-Canefri 1874) (Chitonida: Acanthochitonidae) is a commonly found intertidal chiton species in South Korea. In this study, we characterized the complete mitochondrial genome of A. defilippii (14,999 bp long), comprising 13 protein-coding genes (PCGs), 22 transfer RNA genes, two ribosomal RNA genes, and an A + T rich region (166 bp). The base composition is as follows: 31.82% for A, 11.63% for C, 16.69% for G, and 39.86% for T. We reconstructed a maximum likelihood (ML) tree to elucidate phylogenetic relationships among the eight chitonid families using the nucleotide sequences of all PCGs. The ML tree revealed that A. defilippii clustered with Acanthochitona avicula (BP 100) within the family Acanthochitonidae. Acanthochitonidae formed a sister group with Mopaliidae. The results could provide a valuable understanding the phylogenetic relationships of chitonid species.

The ground beetle Synuchus nitidus (Motschulsky, 1861) (Carabidae: Harpalinae: Sphodrini) is one of the most common species in the forests of South Korea, which has the potential to be utilized as an environmental indicator. Here, we characterized the complete mitochondrial genome (mitogenome) of S. nitidus, which is the first in the harpaline tribe Sphodrini. Its genome is 16,392 bp in length and composed of 13 protein-coding genes (PCGs), 22 tRNA genes, two rRNA genes, and an A + T rich region. In addition, we reconstructed a maximum likelihood tree to elucidate the phylogenetic position of Sphodrini among the seven harpaline tribes using nucleotide sequences of the 13 PCGs. The ML tree supported a monophyletic clade of the subfamily Harpalinae and showed a close relationship between Sphodrini and Lebinii with a low bootstrap value. The complete mitogenome of S. nitidus could be helpful for molecular species identification and exploring phylogenetic relationships among carabids.

Rehabilitation of gait function in post-stroke hemiplegic patients is critical for improving mobility and quality of life, requiring a comprehensive understanding of individual gait patterns. Previous studies on gait analysis using unsupervised clustering often involve manual feature extraction, which introduces limitations such as low accuracy, low consistency, and potential bias due to human intervention. This cross-sectional study aimed to identify and cluster gait patterns using an end-to-end deep learning approach that autonomously extracts features from joint angle trajectories for a gait cycle, minimizing human intervention. A total of 74 sub-acute post-stroke hemiplegic patients with lower limb impairments were included in the analysis. The dataset comprised 219 sagittal plane joint angle and angular velocity trajectories from the hip, knee, and ankle joints during gait cycles. Deep temporal clustering was employed to cluster them in an end-to-end manner by simultaneously optimizing feature extraction and clustering, with hyperparameter tuning tailored for kinematic gait cycle data. Through this method, six optimal clusters were selected with a silhouette score of 0.2831, which is a relatively higher value compared to other clustering algorithms. To clarify the characteristics of the selected groups, in-depth statistics of spatiotemporal, kinematic, and clinical features are presented in the results. The results demonstrate the effectiveness of end-to-end deep learning-based clustering, yielding significant performance improvements without the need for manual feature extraction. While this study primarily utilizes sagittal plane data, future analysis incorporating coronal and transverse planes as well as muscle activity and gait symmetry could provide a more comprehensive understanding of gait patterns.

Recent advancements in offline reinforcement learning (RL) have showcased the potential for leveraging static datasets to train optimal policies. However, real-world applications often face challenges due to missing or incomplete state information caused by imperfect sensor performance or intentional interlaces. We propose the Dimension-Wise Drop Decision Transformer (D3T), a novel framework designed to address dimension-wise data loss in sensor observations, enhancing the robustness of RL algorithms in real-world scenarios. D3T innovatively incorporates dimension-wise drop information embeddings within the Transformer architecture, facilitating effective decision-making even with incomplete observations. Our evaluation in the D4RL MuJoCo domain demonstrates that D3T significantly outperforms existing methods such as the Decision Transformer, particularly with substantial dimension-wise drops of observations. These results confirm D3T’s capability in managing real-world imperfections in state observations and illustrate its potential to substantially expand the applicability of RL in more complex and dynamic environments.

IntroductionSea slater, in the genus Ligia, is widespread in rocky shore habitats, and the taxon is easily isolated due to its limited dispersal capacity. Therefore, most Ligia species exhibit an allopatric distribution, but Ligia exotica and L. cinerascens exhibit an overlapping distribution distribution in East Asia. Previous studies on both species have confirmed the existence of highly divergent lineages based on 16S rRNA.MethodsIn the present study, 282 Ligia individuals were collected at ten, three, and three sites in South Korea, Japan, and Vietnam, respectively, and 41 haplotypes were observed based on 16S rRNA.Results and discussionThe results of phylogeny, phylogenetic network, and TCS network, Principal Coordinates Analysis, and four Molecular Species Delimitation Analyses revealed that six genetic lineages including L. cinerascens, Lineages N and S of L. exotica, Ligia sp. 1, sp.2 and sp.3 were present. The three genetic lineages, including L. cinerascens, Lineage N of L. exotica, and Lineage S of L. exotica, were also identified in the phylogeny based on a nuclear gene of the sodium–potassium ATPase α-subunit (Nak). Phylogeographic analysis revealed that L. cinerascens and Lineage N of L. exotica were distributed overlappingly in South Korea, Japan, and the northern region of China. Generally, the two lineages of L. exotica were distributed allopatrically, which was more evident along the coastline of mainland China than that of Japan. The results of time-calibrated phylogeny suggested that the speciation events of Ligia species might be associated with Japanese mainland formation from Oligocene to Miocene (approximately 30-5 million years ago, Mya). The results of the present study provide insights that could facilitate the understanding of the evolutionary history of Ligia, tracking of geological processes, and evolutionary effects of palaeogeographical events at the population level.

Abstract The blotched nerite Nerita albicilla (Linnaeus 1758) is distributed in intertidal areas of the Indo-Pacific. In South Korea, it has been found only in the southernmost region of Jeju Island so far. Owing to its limited distribution, it can be a promising intertidal species helpful for monitoring global climate change effects in the Korean Peninsula. We performed population genetic analyses based on 393 COI haplotypes from 697 N. albicilla, including 167 from this study and 530 from public databases. The results showed that there are two distinct genetic lineages in N. albicilla: PAIO (Palearctic, Australasia, Indo-Malay, and Oceania) and Afrotropic lineages. DNA barcoding gap analyses indicated that the two lineages could be differentiated into two different species: N. albicilla (PAIO) and N. originalis sp. nov. (Afrotropic) (3.96%). Additionally, it was revealed that their divergence time was ca. 5.96 Ma and dramatic diversification of COI haplotypes occurred during the late Pliocene and Pleistocene. The results of MDA, BSP, and neutrality test implied recent population size expansion, which was estimated to be ca. 250 Ka. Finally, we discussed whether the observation of N. originalis sp. nov. in South Korea is due to the northward migration through ocean currents caused by global warming or due to artificial activity through marine transportation.

Abstract As society continues to age, it is becoming increasingly important to monitor drug use in the elderly. Social media data have been used for monitoring adverse drug reactions. The aim of this study was to determine whether social network studies (SNS) are useful sources of drug side effects information. We propose a method for utilizing SNS data to plot the known side effects of geriatric drugs in a dosing map. We developed a lexicon of drug terms associated with side effects and mapped patterns from social media data. We confirmed that well-known side effects may be obtained by utilizing SNS data. Based on these results, we propose a pharmacovigilance pipeline that can be extended to unknown side effects. We propose the standard analysis pipeline Drug_SNSMiner for monitoring side effects using SNS data and evaluated it as a drug prescription platform for the elderly. We confirmed that side effects may be monitored from the consumer’s perspective based on SNS data using only drug information. SNS data were deemed good sources of information to determine ADRs and obtain other complementary data. We established that these learning data are invaluable for AI requiring the acquisition of ADR posts on efficacious drugs.

Abstract Background The generation of liver organoids recapitulating parenchymal and non-parenchymal cell interplay is essential for the precise in vitro modeling of liver diseases. Although different types of multilineage liver organoids (mLOs) have been generated from human pluripotent stem cells (hPSCs), the assembly and concurrent differentiation of multiple cell types in individual mLOs remain a major challenge. Particularly, most studies focused on the vascularization of mLOs in host tissue after transplantation in vivo. However, relatively little information is available on the in vitro formation of luminal vasculature in mLOs themselves. Methods The mLOs with luminal blood vessels and bile ducts were generated by assembling hepatic endoderm, hepatic stellate cell-like cells (HscLCs), and endothelial cells derived entirely from hPSCs using 96-well ultra-low attachment plates. We analyzed the effect of HscLC incorporation and Notch signaling modulation on the formation of both bile ducts and vasculature in mLOs using immunofluorescence staining, qRT-PCR, ELISA, and live-perfusion imaging. The potential use of the mLOs in fibrosis modeling was evaluated by histological and gene expression analyses after treatment with pro-fibrotic cytokines. Results We found that hPSC-derived HscLCs are crucial for generating functional microvasculature in mLOs. HscLC incorporation and subsequent vascularization substantially reduced apoptotic cell death and promoted the survival and growth of mLOs with microvessels. In particular, precise modulation of Notch signaling during a specific time window in organoid differentiation was critical for generating both bile ducts and vasculature. Live-cell imaging, a series of confocal scans, and electron microscopy demonstrated that blood vessels were well distributed inside mLOs and had perfusable lumens in vitro. In addition, exposure of mLOs to pro-fibrotic cytokines induced early fibrosis-associated events, including upregulation of genes associated with fibrotic induction and endothelial cell activation (i.e., collagen I, α-SMA, and ICAM) together with destruction of tissue architecture and organoid shrinkage. Conclusion Our results demonstrate that mLOs can reproduce parenchymal and non-parenchymal cell interactions and suggest that their application can advance the precise modeling of liver diseases in vitro.

Most textual analysis-based trading approaches in cryptocurrency (crypto) involve lexical, rule-based methods for extracting news sentiments. Furthermore, language models (LMs) are not always suitable for the crypto domain due to jargon that is not covered in general-purpose texts. This study answers the question of “Is it possible that the LMs can profit by effectively applying the sentiment score of the natural language processing task with chart score in the BTC trading system?” by focusing on the effectiveness of both scores, which significantly affect the profit of the trading system. We introduce CBITS: Cryptocurrency BERT Incorporated Trading System based on pre-trained LMs for Korean crypto sentiment analysis to aid Bitcoin (BTC) trading models. We pre-trained crypto-specific LMs, which are transformer encoder-based architectures. Along with our pre-trained LMs, we also present our custom fine-tuning dataset used to train our LMs on the BTC sentiment classifier and show that using sentiment scores along with BTC chart data boosts the performance of BTC trading models and also allows us to create a market-neutral trading strategy.

In this study, we conduct a pioneering and comprehensive examination of ChatGPT’s (GPT-3.5 Turbo) capabilities within the realm of Korean Grammatical Error Correction (K-GEC). Given the Korean language’s agglutinative nature and its rich linguistic intricacies, the task of accurately correcting errors while preserving Korean-specific sentiments is notably challenging. Utilizing a systematic categorization of Korean grammatical errors, we delve into a meticulous, case-specific analysis to identify the strengths and limitations of a ChatGPT-based correction system. We also critically assess influential parameters like temperature and specific error criteria, illuminating potential strategies to enhance ChatGPT’s efficacy in K-GEC tasks. Our findings offer valuable contributions to the expanding domain of NLP research centered on the Korean language.

Although COVID-19 has severely affected the global economy, information technology (IT) employees managed to perform most of their work from home. Telecommuting and remote work have promoted a demand for IT services in various market sectors, including retail, entertainment, education, and healthcare. Consequently, computer and information experts are also in demand. However, producing IT, experts is difficult during a pandemic owing to limitations, such as the reduced enrollment of international students. Therefore, researching increasing software productivity is essential; this study proposes a code similarity determination model that utilizes augmented data filtering and ensemble strategies. This algorithm is the first automated development system for increasing software productivity that addresses the current situation—a worldwide shortage of software dramatically improves performance in various downstream natural language processing tasks (NLP). Unlike general-purpose pre-trained language models (PLMs), CodeBERT and GraphCodeBERT are PLMs that have learned both natural and programming languages. Hence, they are suitable as code similarity determination models. The data filtering process consists of three steps: (1) deduplication of data, (2) deletion of intersection, and (3) an exhaustive search. The best mating (BM) 25 and length normalization of BM25 (BM25L) algorithms were used to construct positive and negative pairs. The performance of the model was evaluated using the 5-fold cross-validation ensemble technique. Experiments demonstrate the effectiveness of the proposed method quantitatively. Moreover, we expect this method to be optimal for increasing software productivity in various NLP tasks.

Numerical reasoning in machine reading comprehension (MRC) has demonstrated significant performance improvements in the past few years. However, due to the process being restricted to specific languages, low-resource languages are not considered, and MRC studies on such languages are limited. In addition, the methods that rely on existing information extracted within the span of a paragraph have limitations in responding to questions requiring actual reasoning. To overcome these shortcomings, this study establishes a dataset for learning Korean Question and Answering (QA) models that not only answer within the span of passages but also perform numerical reasoning on passages and questions. Its efficacy was verified by training the model. We recruited eight annotators to tag the ground truth label, and they annotated datasets with 920, 115, and 115 passages in the train, dev, and test, respectively. A simple yet sophisticated automatic inter-annotation tool was created by effectively reducing the possibility of inaccuracy and error entailed by humans in the data construction process. This tool used common KoBERT and KoELECTRA. We defined four general conditions, and six conditions humans must inspect and fine-tune the pre-trained language models with numerically aware architecture. The KoELECTRA and NumNet+ with KoELECTRA were fine-tuned, and experiments in identical hyperparameter settings showed that compared with other models, the performance of NumNet+ with KoELECTRA was higher by more than 1.3 points. Our research contributes to the Korean MRC research and suggests potential and insight into MRC models capable of numerical reasoning.

Abstract Background Fungal infection secondary to a penetrating plant thorn injury in upper extremities is infrequently reported especially in immunocompetent hosts. Alternaria is a dematicaceous hyphomycete, commonly found in decay and plant pathogens, and Alternaria alternata has been regarded as the most frequent species among more than 400 s of its species. This case is the first report of thorn-induced Alternaria alternata infection of the hand in an immunocompetent host. Case presentation A 47-year-old male patient was admitted to our institution with persistent pain and swelling of the right hand even after a prior surgical removal of a previous thorn injury. Upon impression of abscess, chronic extensor tenosynovitis, and septic arthritis at the 3rd metacarpophalangeal joint based on advanced imaging, the patient underwent surgical incision and drainage. Intraoperative culture, biopsy, and gene molecular sequencing results revealed fungal infection with Alternaria alternata. Postoperatively, the patient was treated with oral itraconazole (200 mg q 12 h) for nine consecutive months. Conclusions We report the first case of chronic extensor tenosynovitis and septic arthritis of the hand with Alternaria alternata after a thorn injury in an immunocompetent host. Despite rare incidences of fungal extensor tenosynovitis and septic arthritis, the current case strongly suggests a careful examination of social history and surgical debridement along with a prolonged use of appropriate anti-fungal agents in thorn injuries.

As the use of renewable energy is continuously increasing, power systems are currently exposed to greater uncertainty and variability, which can lead to severe power system stability issues. Therefore, a power system analysis tool should be devised to assess the impact of renewable energy integration along with an accurate modeling of their stochastic characteristics. In this study, an advanced probabilistic power flow (PPF) method is developed using vine copulas that captures the complex dependency of the stochastic wind power generated from multiple wind sites. The proposed method also involves the use of a function for selecting the probability models of wind speeds by regions in a sophisticated manner. The effectiveness of the proposed method is tested on an IEEE bus system as well as, on a South Korean power system with thousands of buses and transmission lines using PSS/E with Python API. The simulations demonstrate that the proposed method can more accurately evaluate the power system risks with the sophisticated modeling of wind power in multiple sites as compared to the deterministic approach or the PPF with independent sampling.

A white-spotted flower chafer Protaetia brevitarsis seulensis widely distributed in Asian countries is traditionally used in oriental medicine. This study explored gene expression abundance with respect to wing development and metamorphosis in P. b. seulensis based on the large-scale RNA-seq data. The transcriptome assembly consists of 23,551 high-quality transcripts which are approximately 96.7% covered. We found 265 wing development genes, 19 metamorphosis genes, and 1,314 candidates. Of the 1,598 genes, 1,594 are included exclusively in cluster 4 with similar gene co-expression patterns. The network centrality analyses showed that wing development- and metamorphosis-related genes have a high degree of betweenness centrality and are expressed most highly in eggs, moderately in pupa and adults, and lowest in larva. This study provides some meaningful clues for elucidating the genetic modulation mechanism of wing development and metamorphosis in P. b. seulensis.

OBJECTIVES The reported effects of digital media overuse on physical activity among adolescents are inconsistent. This study examined the association between hours of digital media use and lack of moderate-intensity physical exercise (mPE) according to the type of digital media. METHODS This study included 1,837 middle school students from the iCURE (Internet user Cohort for Unbiased Recognition of gaming disorder in Early Adolescence) study conducted in Korea. Hours spent using digital media were measured by self-reported daily usage time for Internet games, messengers, social media, and watching game streaming on weekdays. Lack of mPE was defined as performing a minimum of 30 minutes at a time less than twice weekly. Multivariable logistic regression analysis stratified by sex was performed. RESULTS Among male students, the group with the highest hours of using either Internet games or watching game streaming was more likely to lack mPE than each non-user group. In contrast, among male students, the group using either messengers or social media had a higher rate of mPE compared to each non-user group. Female students showed no association between hours spent using Internet games, messengers, social media, or watching game streaming and a lack of mPE. CONCLUSIONS Among male middle school students in Korea, the excessive use of Internet games or watching game streaming was associated with a lack of mPE. Thus, guidelines should be established regarding adolescent use of internet games and watching game streaming.

Cytochrome P450 (CYP) reaction phenotyping has become crucial for predicting drug reactions and side effects. Single nucleotide polymorphisms (SNPs) in CYP genes alter drug metabolism capacity and cause unexpected drug-related reactions. Here, we established two human induced pluripotent stem cell (hiPSC) lines with pharmacologically important SNPs in CYP2D6 in conjunction with CYP2C19 or CYP3A5 genes. These hiPSC lines can serve as valuable resources for expanding our understanding of the relationships between genotypes and drug reactions.

Dialogue relation extraction identifies semantic relations between entity pairs in dialogues. This research explores a methodology harnessing the potential of prompt-based fine-tuning paired with a trigger-generation approach. Capitalizing on the intrinsic knowledge of pre-trained language models, this strategy employs triggers that underline the relation between entities decisively. In particular, diverging from the conventional extractive methods seen in earlier research, our study leans towards a generative manner for trigger generation. The dialogue-based relation extraction (DialogeRE) benchmark dataset features multi-utterance environments of colloquial speech by multiple speakers, making it critical to capture meaningful clues for inferring relational facts. In the benchmark, empirical results reveal significant performance boosts in few-shot scenarios, where the availability of examples is notably limited. Nevertheless, the scarcity of ground-truth triggers for training hints at potential further refinements in the trigger-generation module, especially when ample examples are present. When evaluating the challenges of dialogue relation extraction, combining prompt-based learning with trigger generation offers pronounced improvements in both full-shot and few-shot scenarios. Specifically, integrating a meticulously crafted manual initialization method with the prompt-based model—considering prior distributional insights and relation class semantics—substantially surpasses the baseline. However, further advancements in trigger generation are warranted, especially in data-abundant contexts, to maximize performance enhancements.

Abstract The chiton Liolophura japonica (Lischke 1873) is distributed in intertidal areas of the northwestern Pacific. Using COI and 16S rRNA, we found three genetic lineages, suggesting separation into three different species. Population genetic analyses, the two distinct COI barcoding gaps albeit one barcoding gap in the 16S rRNA, and phylogenetic relationships with a congeneric species supported this finding. We described L. koreana, sp. nov. over ca. 33°24′ N (JJ), and L. sinensis, sp. nov. around ca. 27°02′–28°00′ N (ZJ). We confirmed that these can be morphologically distinguished by lateral and dorsal black spots on the tegmentum and the shape of spicules on the perinotum. We also discuss species divergence during the Plio-Pleistocene, demographic expansions following the last interglacial age in the Pleistocene, and augmentation of COI haplotype diversity during the Pleistocene. Our study sheds light on the potential for COI in examining marine invertebrate species discrimination and distribution in the northwestern Pacific.

Abstract The land snail Ellobium chinense (L. Pfeiffer, 1855) (Eupulmonata, Ellobiida, Ellobiidae), which inhabits the salt marshes along the coastal areas of northwestern Pacific, is an endangered species on the IUCN Red List. Over recent decades, the population size of E. chinense has consistently decreased due to environmental interference caused by natural disasters and human activities. Here, we provide the first assessment of the genetic diversity and population genetic structures of northwestern Pacific E. chinense. The results analyzed with COI and microsatellites revealed that E. chinense population exhibit metapopulation characteristics, retaining under the influence of the Kuroshio warm currents through expansion of the Late-Middle and Late Pleistocene. We also found four phylogenetic groups, regardless of geographical distributions, which were easily distinguishable by four unidirectional and stepwise adenine-to-guanine transitions in COI (sites 207–282–354–420: A–A–A–A, A–A–G–A, G–A–G–A, and G–G–G–G). Additionally, the four COI hotspots were robustly connected with a high degree of covariance between them. We discuss the role of these covariate guanines which link to form four consecutive G-quadruplexes, and their possible beneficial effects under positive selection pressure.

Tokenization is a significant primary step for the training of the Pre-trained Language Model (PLM), which alleviates the challenging Out-of-Vocabulary problem in the area of Natural Language Processing. As tokenization strategies can change linguistic understanding, it is essential to consider the composition of input features based on the characteristics of the language for model performance. This study answers the question of “Which tokenization strategy enhances the characteristics of the Korean language for the Named Entity Recognition (NER) task based on a language model?” focusing on tokenization, which significantly affects the quality of input features. We present two significant challenges for the NER task with the agglutinative characteristics in the Korean language. Next, we quantitatively and qualitatively analyze the coping process of each tokenization strategy for these challenges. By adopting various linguistic segmentation such as morpheme, syllable and subcharacter, we demonstrate the effectiveness and prove the performance between PLMs based on each tokenization strategy. We validate that the most consistent strategy for the challenges of the Korean language is a syllable based on Sentencepiece.

The project Carbon-Free Island Jeju by 2030 promoted by the Republic of Korea aims to expand the renewable energy sources centered on wind power in Jeju Island and supply electric vehicles for eco-friendly mobility. However, the increased penetration rate of electric vehicles and expansion of variable renewable energy sources can accelerate the power demand and uncertainty in the power generation output. In this paper, power system analysis is performed through electric vehicle charging demand and wind power outputs prediction, and an electric vehicle charging decentralization algorithm is proposed to mitigate system congestion. In order to predict electric vehicle charging demand, the measurement data were analyzed, and random sampling was performed by applying the weight of charging frequency for each season and time. In addition, wind power outputs prediction was performed using the ARIMAX model. Input variables are wind power measurement data and additional explanatory variables (wind speed). Wind power outputs prediction error (absolute average error) is about 9.6%, which means that the prediction accuracy of the proposed algorithm is high. A practical power system analysis was performed for the scenario in which electric vehicle charging is expected to be higher than the wind power generation due to the concentration of electric vehicle charging. The proposed algorithm can be used to analyze power system problems that may occur due to the concentration of electric vehicle charging demand in the future, and to prepare a method for decentralizing electric vehicle charging demand to establish a stable power system operation plan.

Reasoning, a trait of cognitive intelligence, is regarded as a crucial ability that distinguishes humans from other species. However, neural networks now pose a challenge to this human ability. Text-to-image synthesis is a class of vision and linguistics, wherein the goal is to learn multimodal representations between the image and text features. Hence, it requires a high-level reasoning ability that understands the relationships between objects in the given text and generates high-quality images based on the understanding. Text-to-image translation can be termed as the visual thinking of neural networks. In this study, our model infers the complicated relationships between objects in the given text and generates the final image by leveraging the previous history. We define diverse novel adversarial loss functions and finally demonstrate the best one that elevates the reasoning ability of the text-to-image synthesis. Remarkably, most of our models possess their own reasoning ability. Quantitative and qualitative comparisons with several methods demonstrate the superiority of our approach.

The proportion of renewable energy generation is expanding worldwide with the goal of reducing greenhouse gas. According to the 8th Basic Plan for Long-term Electricity Supply and Demand in South Korea, South Korea reduces traditional energy generation such as nuclear and coal plants and achieves 20% (58.5GW) of renewable energy generation by 2030. Wind Generating Resources (WGRs) are affected by meteorological variables such as temperature, wind speed and wind direction. Specifically, WGRs have uncertainty and variability issues depending on temporal and spatial characteristics. In this paper, we propose the probabilistic estimation of wind generating resources based on the spatiotemporal penetration scenarios for power grid expansion. The data of WGRs are analyzed based on clustering method considering the spatiotemporal penetration scenarios, and the potential scenarios are estimated using Monte Carlo simulation by selecting a representative power distribution probability for each cluster. The proposed estimation model of WGRs will play a key role to develop the hedging strategies of investment decision on power grid expansion planning with high wind power penetrations.

A giant water bug Lethocerus deyrollei (Hemiptera: Belostomatidae) is a large, predatory, and nocturnal hemipteran insect, which has been considered threatened and thus enrolled as an endangered species in South Korea and Japan. Here, we characterized the complete mitochondrial genome of L. deyrollei, which has a circular form with 19,295 bp in length, which is the longest when compared to those of the 111 hemipteran species reported so far. Its longest genome size is due to the extremely extended CR (4686 bp), which is much longer than those of China and Japan. It consisted of a total of 37 genes (13 PCGs, 22 tRNA genes, and two rRNA genes) and one control region (CR). The genome composition and gene order were identical to those previously reported from the same species of China and Japan with over 99.7% sequence similarities except for CR and trnI. The nucleotide composition was highly A + T biased, accounting for 71% of the whole mitochondrial genome, as in other species of Nepoidea. Based on the aa sequences of 13 PCGs, we reconstructed a maximum likelihood tree, which indicated that the three mitochondrial genomes of L. deyrollei from South Korea, China, and Japan are grouped, and also Lethocerus, Belostomatidae, Nepoidea, Nepomorpha, Heteroptra are strong monophyletic groups, respectively.

The complete mitochondrial genome of a two-spotted cricket Gryllus bimaculatus (Orthoptera: Gryllidae) from South Korea is determined and characterized in this study. The circular genome is 16,075 bp long, which consists of 13 protein-coding genes (PCGs), 2 ribosomal RNA genes, 22 transfer RNA genes, and an A + T−rich region. It has a base composition of A (40.35%), G (9.09%), C (16.80%), and T (33.76%). The gene order is identical to the ancestral gene arrangement pattern generally shown in arthropods, with the exception of an inversion of trnN-trnS1-trnE into trnE-trnS1-trnN. The maximum likelihood (ML) tree supports that G. bimaculatus is a distinct member of the monophyletic family Gryllidae.

Machine reading comprehension is a question answering mechanism in which a machine reads, understands, and answers questions from a given text. These reasoning skills can be sufficiently grafted into the Korean College Scholastic Ability Test (CSAT) to bring about new scientific and educational advances. In this paper, we propose a novel Korean CSAT Question and Answering (KCQA) model and effectively utilize four easy data augmentation strategies with round trip translation to augment the insufficient data in the training dataset. To evaluate the effectiveness of KCQA, 30 students appeared for the test under conditions identical to the proposed model. Our qualitative and quantitative analysis along with experimental results revealed that KCQA achieved better performance than humans with a higher F1 score of 3.86.

The complete mitochondrial genome sequence from the Asian crested ibis, Nipponia nippon (Aves, Pelecaniformes, Threskiornithidae), was determined and characterized in detail. This mitochondrial genome is 16,813 bp long, and consists of 13 PCGs, 22 tRNAs, and 2 rRNAs. The nucleotide composition is slightly biased with A + T contents of 53.79% (A, T, C, and G was 30.37%, 23.42%, 31.99%, and 14.22%, respectively). 11 PCGs are initiated by ATN codons, except for cox1 and cox2 with GTG instead. The phylogenetic relationships based on the maximum-likelihood and Bayesian methods showed that the placement of N. nippon within the order Pelecaniformes, with forming the monoclade of the family Threskiornithidae.

In an age overflowing with information, the task of converting unstructured data into structured data are a vital task of great need. Currently, most relation extraction modules are more focused on the extraction of local mention-level relations—usually from short volumes of text. However, in most cases, the most vital and important relations are those that are described in length and detail. In this research, we propose GREG: A Global level Relation Extractor model using knowledge graph embeddings for document-level inputs. The model uses vector representations of mention-level ‘local’ relation’s to construct knowledge graphs that can represent the input document. The knowledge graph is then used to predict global level relations from documents or large bodies of text. The proposed model is largely divided into two modules which are synchronized during their training. Thus, each of the model’s modules is designed to deal with local relations and global relations separately. This allows the model to avoid the problem of struggling against loss of information due to too much information crunched into smaller sized representations when attempting global level relation extraction. Through evaluation, we have shown that the proposed model yields high performances in both predicting global level relations and local level relations consistently.