• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.1
• /
• pp.1-8
• /
• 2021

In the political field of news articles, there are polarized and biased characteristics such as conservative and liberal, which is called political bias. We constructed keyword-based dataset to classify bias of news articles. Most embedding researches represent a sentence with sequence of morphemes. In our work, we expect that the number of unknown tokens will be reduced if the sentences are constituted by subwords that are segmented by the language model. We propose a document embedding model with subword tokenization and apply this model to SVM and feedforward neural network structure to classify the political bias. As a result of comparing the performance of the document embedding model with morphological analysis, the document embedding model with subwords showed the highest accuracy at 78.22%. It was confirmed that the number of unknown tokens was reduced by subword tokenization. Using the best performance embedding model in our bias classification task, we extract the keywords based on politicians. The bias of keywords was verified by the average similarity with the vector of politicians from each political tendency.

• Research in Plant Disease
• /
• v.26 no.4
• /
• pp.210-221
• /
• 2020

This research was executed to select beneficial antagonists from digestive organ of Allomyrina dichotoma larva that can be put on environment friendly control against phytopathogenic fungi. We screened 38 bacterial strains inhibiting mycelial growth against eight plant pathogens through dual culture assay. The 10 strains among 38 bacterial strains were selected as beneficial microbes showing antifungal activity against Botrytis cinerea, Plasmodiophora brassicae, Colletotrichum acutatum and Phytophthora capsici through under greenhouse pot trials. The 10 bacterial strains that shown strongest antifungal activity were classified into 3 genera and 10 species, and identified as the genus Bacillus (DM146, DM152, DH2, and DH16), Paenibacillus (DF30, DH14, and DM142) and Streptomyces (DF137, DM48, and DH92) by morphological characteristics and 16s rRNA gene sequence. The 10 bacterial strains had solubilizing activity of insoluble phosphates, production of IAA (indole-3-acetic acid), β-1,3-glucanase and protease. Among the 10 bacterial strains, DM152 strain was produced significant enhancement of all growth parameters of chili pepper and tomato seedlings under greenhouse condition. Thus, this study demonstrated that gut microbes of Allomyrina dichotoma larva will be useful as a potential biocontrol agent against plant pathogens and biofertilizer.

• Research in Plant Disease
• /
• v.26 no.4
• /
• pp.256-263
• /
• 2020

Incidence of Fusarium wilt was surveyed in fields of summer radish in Gangwon province in Korea in 2018 and 2019. The disease started in early July and spread rapidly in hot summer of late July and August and in severe case, reached up to 80% in a field in Gangneung area. Symptoms in the seedling stage include poor growth and browning of internal tissue of root. During mid-growth, the leaves of diseased plant turned yellow over time, the surface of the roots changed from white to blackish, and the vascular tissues turned brown. A total of 23 isolates was obtained from the diseased plants and identified as Fusarium oxysporum f. sp. raphani by elongation factor-1α and intergenic spacer sequence analysis. Pathogenicity of the isolates was tested by artificial inoculation to the radish and other plants. All the isolates tested were pathogenic to radish plant, although there were differences in virulence on radish 11 cultivars. However, the isolates were not virulent to other plants except some cruciferous vegetables including Brussels sprouts, rocket, stock, and turnip. The results of pathogenicity test showed that it is necessary to rotate with crops other than cruciferous vegetables in order to prevent Fusarium wilt from radish fields.

• Conservation Science in Museum
• /
• v.25
• /
• pp.9-26
• /
• 2021

This paper describes the production methods that were originally used for an arrow bundle excavated from a Bronze Age residential area in Auraji in Jeongseon, Gangwon-do Province and the conservation treatment process that it subsequently underwent. An arrow conventionally consists of an arrowhead and a shaft. It is rare to excavate a shaft along with an arrowhead in a complete form since the shaft is made of organic materials. Notably, the arrow bundle from the Auraji site is of great significance as it shows traces of tangless stone arrowheads attached to charred shafts and offers an important case of the split end of a piece of a tree being inserted into an arrowhead. For a further examination of the characteristics of the arrows from Auraji, microscopic investigation was conducted and the type of wood used for the arrow shafts was examined. The sequence and direction of processing and the particle sizes of the grinding tools were revealed through the analysis of traces of grinding on the stone arrowheads. The shaft is presumed to have been made from a green length of three-year-old willow (Salix spp.). A curing agent with a high degree of waterproofing and reversibility was used during the on-site curing process according to demands of the surrounding environment, and a technique that the authors call the "Bridge" method was used for emergency collection of the relics. Once the bundle was transferred to the conservation treatment lab, reinforcing materials were carefully chosen as it was important not to damage the relics during the process of turning them for the repair of their reverse sides. For this purpose, artificial clay was selected since it can safely bear a load and has excellent physical properties. Finally, detached parts were rejoined, the relics and their surrounding materials were cleaned, and the bottom sides were finished with epoxy resin prior to the display of the relics at the museum.

• Journal of Korea Water Resources Association
• /
• v.55 no.5
• /
• pp.383-392
• /
• 2022

The restoration of the Nakdong River estuary is one of the most important projects of the Ministry of Environment, Republic of Korea. A real-scale experiment of gate operation was executed from 2019 to 2020, and a pilot operation was performed in 2021. The gate of Nakdong River Estuary Barrier (NEB) is supposed to be continuously opened based on the experiment results. Many critical decisions should be made immediately during the experiment based on the real-time measured data and numerical analysis considering the seawater inflows. The decision-making sequence was made systematically with the accurate estimation of seawater inflow. The estimation of seawater inflow is the main research objective and the equations of seawater inflow were developed, reflecting the structural characteristics of NEB. The inflow equations were developed in two forms, overflow and underflow. ADCP (Acoustic Doppler Current Profiler) was used to measure seawater inflow, check the accuracy of the developed equations, and derive the flow coefficient. The comparison error of the developed equations was about 3% compared to the measured data.

• Journal of Bio-Environment Control
• /
• v.31 no.3
• /
• pp.152-162
• /
• 2022

This study developed simulation model for predicting the greenhouse interior environment using artificial intelligence machine learning techniques. Various methods have been studied to predict the internal environment of the greenhouse system. But the traditional simulation analysis method has a problem of low precision due to extraneous variables. In order to solve this problem, we developed a model for predicting the temperature inside the greenhouse using machine learning. Machine learning models are developed through data collection, characteristic analysis, and learning, and the accuracy of the model varies greatly depending on parameters and learning methods. Therefore, an optimal model derivation method according to data characteristics is required. As a result of the model development, the model accuracy increased as the parameters of the hidden unit increased. Optimal model was derived from the GRU algorithm and hidden unit 6 (r² = 0.9848 and RMSE = 0.5857℃). Through this study, it was confirmed that it is possible to develop a predictive model for the temperature inside the greenhouse using data outside the greenhouse. In addition, it was confirmed that application and comparative analysis were necessary for various greenhouse data. It is necessary that research for development environmental control system by improving the developed model to the forecasting stage.

• Science of Emotion and Sensibility
• /
• v.25 no.1
• /
• pp.29-40
• /
• 2022

This study uses the bidet 4D double diamond design process model to propose an improvement for "senior-oriented bidet product service design" that reflects the characteristics and needs of seniors. This study was based on the product service system concept. To this end, qualitative research on seniors was conducted to derive user value factors, and, based on this, product service ideas were discovered, and a prototype reflecting the usefulness review of a working-level expert group was proposed. First, a "smart application service for user-customized function setting guide" was proposed. A bidet incorporating Internet of Things technology and a smart phone are linked to provide an app service that automatically interprets user characteristic information and information on bidet products to guide customized functions. Second, a control panel and remote control user interface to "user-oriented product service interface" was proposed. In consideration of the usability and cognitive ability of seniors, a simple and intuitive physical user interface such as a configuration centered on main functions, button arrangement according to task sequence, and a touch screen remote control was presented. Third, we proposed a "bidet care service linked with products and health/hygiene care" that provides a wide range of services such as user health and hygiene, cleanliness, entertainment, etc., in addition to regular bidet product service. This study proposed a product-based service design methodology that can improve user experience and relationship quality by discovering and improving the pain points and needs of users (seniors) in the process of using bidet products (before, during, and after use).

• Journal of Korean Elementary Science Education
• /
• v.41 no.2
• /
• pp.307-324
• /
• 2022

This study examined the content diversity of seven authorized science textbooks by comparing the characteristics of the science concept description and the contents of inquiry activities in the "weight of objects" unit. For each textbook, the flow of concept description content and the uniqueness of the concept description process were analyzed, and the number of nodes and links and words with high connections were determined using language network analysis. In addition, for the inquiry activities described in each textbook, the inquiry subject, inquiry type, science process skill, and uniqueness were investigated. Results showed that the authorized textbooks displayed no more diversity than expected in their scientific concept description method or their inquiry activity composition. The learning elements, inclusion of subconcepts, and central words were similar for each textbook. The comparison of inquiry activities showed similarities in their contents, inquiry types, and scientific process skills. Specifically, these textbooks did not introduce any research topics or experimental methods that were absent in previous textbooks. However, despite the fact that the authorized textbook system was developed based on the same curriculum, some efforts were made to make use of its strengths. Since the sequence of subconcepts to explain the core contents differed across textbooks, this explanation process was divided into several types, and although the contents of inquiry activities were the same, the materials for inquiry activities were shown differently for each textbook to improve and overcome the difficulties in the existing experiments. These findings necessitate the continuation of efforts to utilize the strengths of certified textbooks.

• Composites Research
• /
• v.36 no.1
• /
• pp.53-58
• /
• 2023

The PIC design method is assigning different stacking sequences for each shell element through the preliminary FE analysis. In previous study, machine learning was applied to the PIC design method in order to assign the region efficiently, and the training data is labeled by dividing each region into tension, compression, and shear through the preliminary FE analysis results value. However, since buckling is not considered, when buckling occurs, it can't be divided into appropriate loading type. In the present study, it was proposed PIC-NTL (PIC design using novel technique for analyzing load type) which is method for applying a novel technique for analyzing load type considering buckling to the conventional PIC design. The stress triaxiality for each ply were analyzed for buckling analysis, and the representative loading type was designated through the determined loading type within decision area divided into two regions of the same size in the thickness direction of the elements. The input value of the training data and label consisted in coordination of element and representative loading type of each decision area, respectively. A machine learning model was trained through the training data, and the hyperparameters that affect the performance of the machine learning model were tuned to optimal values through Bayesian algorithm. Among the tuned machine learning models, the SVM model showed the highest performance. Most effective stacking sequence were mapped into PIC tube based on trained SVM model. FE analysis results show the design method proposed in this study has superior external loading resistance and energy absorption compared to previous study.

• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.107-119
• /
• 2023

The technology that recognizes a soldier's motion and movement status has recently attracted large attention as a combination of wearable technology and artificial intelligence, which is expected to upend the paradigm of troop management. The accuracy of state determination should be maintained at a high-end level to make sure of the expected vital functions both in a training situation; an evaluation and solution provision for each individual's motion, and in a combat situation; overall enhancement in managing troops. However, when input data is given as a timer series or sequence, existing feedforward networks would show overt limitations in maximizing classification performance. Since human behavior data (3-axis accelerations and 3-axis angular velocities) handled for military motion recognition requires the process of analyzing its time-dependent characteristics, this study proposes a high-performance data-driven classifier which utilizes the long-short term memory to identify the order dependence of acquired data, learning to classify eight representative military operations (Sitting, Standing, Walking, Running, Ascending, Descending, Low Crawl, and High Crawl). Since the accuracy is highly dependent on a network's learning conditions and variables, manual adjustment may neither be cost-effective nor guarantee optimal results during learning. Therefore, in this study, we optimized hyperparameters using Bayesian optimization for maximized generalization performance. As a result, the final architecture could reduce the error rate by 62.56% compared to the existing network with a similar number of learnable parameters, with the final accuracy of 98.39% for various military operations.