• Korean small business review
• /
• v.42 no.4
• /
• pp.309-326
• /
• 2020

There often is information asymmetry between start-ups and the investors, which is because start-up companies in the early stages do not have track records. Meanwhile, since the government grants programs go through a fair and the intense competition process, the government grants can provide a more objective information for start-ups in the early stages and perform a signal function that guarantees a company's capabilities and potential. This study confirms the quantitative relationship between government grants and investment attraction by using the hurdler model. We found that, although there is the proportionate relationship between the scale of government grants and that of external funds, more than a certain amount of government grants is required for technology-based start-ups to exceed the stage of attracting their first external funds. Our findings suggest that it is necessary to consider the hurdles structure in the study of signaling theory perspective, as the mechanisms for determining whether or not to attract external funds are different from determining the level of external funds. In addition, differentiated policy support is needed to help early-stage technology start-ups go beyond the threshold of investment attraction-the creation of a 'threshold effect'.

• Journal of the Korean Society for Library and Information Science
• /
• v.56 no.4
• /
• pp.101-127
• /
• 2022

Article 31(1) of the Libraries Act(Act No. 18547), which was completely revised on December 7, 2021, stipulates that "the head of a local government or the superintendent of a city/provincial office of education must formulate a plan for the establishment and operation of a public library in advance and obtain the pre-evaluation of the feasibility of establishing a public library from the Minister of Culture, Sports and Tourism." Through the preliminary feasibility evaluation at the construction stage of the public library, it is possible to adjust distribution to improve the adequacy of scale and resolve regional imbalances and gaps. In addition, it is expected to increase service satisfaction and operational enhancement by inducing faithful securing of core infrastructure (librarians, collection, facilities, systems, etc.) in terms of balanced regional development and public library construction. The purpose of this study is to develop and present the basic direction and feasibility evaluation model for establishment of public libraries. The proposed evaluation model is expected to secure the legal basis and institutional legitimacy of the pre-evaluation system for public library establishment and to prevent waste of tax due to poor construction and operation of public libraries.

• Journal of Korea Water Resources Association
• /
• v.55 no.12
• /
• pp.1011-1020
• /
• 2022

Previous studies for the assessment of submarine groundwater discharge (SGD) were perfomed for areas where a large amount of SGD was observed. Newly developed assessment methods were proposed that was based on an analytic solution using sharp interface model. The proposed mathematical equations used the existing observed groundwater level and hydrogeological data of Jeju Island as input data. The quantitatively assessed FSGD values were compared to the basin-scale recharge estimation values in Seong-San area in eastern Jeju. As a result of the study, it was estimated that the amount of FSGD in the Seongsan area ranges from about 2.65 to 9.15% of the amount of areal-recharge. Through the analysis of the FSGD combined with the analytic model, it is to be provided as a scientific tool to establish a more reasonable coastal water resource management plan.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.25 no.2
• /
• pp.23-38
• /
• 2022

This study investigates the characteristics and physical habitat requirements for each Bufo gargarizans life history through a literature survey. After deriving variables for each component of Bufo gargarizans, in order to reduce regional deviations from eight previously studied literature research areas for deriving the criteria for variables, a total of 12 natural habitats of Bufo gargarizanss are selected as spatial ranges by selecting four additional sites such as Umyeonsan Ecological Park in Seoul, Wonheungibangjuk in Cheongju in the central region, Changnyeong Isan Reservoir in the southern region, and Mangwonji in Daegu. This study presents Bufo gargarizans SI, a species endemic to Korea, whose population is rapidly declining due to large-scale housing site development and road development, and develops a Bufo gargarizans HSI model accordingly to improve the function of the damaged Bufo gargarizans habitat and to present an objective basis for site selection of alternative habitat. At the same time, it provides basic data for adaptive management and follow-up monitoring. The three basic habitat requirements of amphibians, the physical habitat requirements of Bufo gargarizans, synthesized with shelter, food, and water, and the characteristics of each life history, are classified into five components by adding space and threats through literature research and expert advice. Variables are proposed by synthesizing and comparing the general characteristics of amphibians, among the previously studied single species of amphibians, the components of HSI of goldfrogs and Bufo gargarizans, and the ecological and physical environmental characteristics of Bufo gargarizans. Afterwards, through consultation with an amphibian expert, a total of 10 variables are finally presented by adjacent forest area(ha), the distance between spawning area and the nearest forest land(m), the soil, the distance from the wetland(m), the forest layered structure, the low grassland space, the permanent wetland area(ha), shoreline slope(%), PH, presence of predators, distance from road(m), presence or absence of obstacles. n order to derive the final criteria for each of the 10 variables, the criteria(alternative) for each variable are presented through geographic information analysis of the site survey area and field surveys of the previously studied literature research area. After a focus group interview(FGI) of 30 people related to the Bufo gargarizans colony in Cheongju, a questionnaire and in-depth interviews with three amphibians experts are conducted to verify and supplement the criteria for each final variable. Based on the finally developed Bufo gargarizans HSI, the Bufo gargarizans habitat model is presented through the SI graph model and the drawing centering on the Bufo gargarizans spawning area

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.8
• /
• pp.325-330
• /
• 2022

This paper conducts the collection and experiment of datasets for deep learning research on sign language such as sign language recognition, sign language translation, and sign language segmentation for Korean sign language. There exist difficulties for deep learning research of sign language. First, it is difficult to recognize sign languages since they contain multiple modalities including hand movements, hand directions, and facial expressions. Second, it is the absence of training data to conduct deep learning research. Currently, KETI dataset is the only known dataset for Korean sign language for deep learning. Sign language datasets for deep learning research are classified into two categories: Isolated sign language and Continuous sign language. Although several foreign sign language datasets have been collected over time. they are also insufficient for deep learning research of sign language. Therefore, we attempted to collect a large-scale Korean sign language dataset and evaluate it using a baseline model named TSPNet which has the performance of SOTA in the field of sign language translation. The collected dataset consists of a total of 11,402 image and text. Our experimental result with the baseline model using the dataset shows BLEU-4 score 3.63, which would be used as a basic performance of a baseline model for Korean sign language dataset. We hope that our experience of collecting Korean sign language dataset helps facilitate further research directions on Korean sign language.

• Journal of the Korean Society for Library and Information Science
• /
• v.57 no.1
• /
• pp.93-114
• /
• 2023

Information extraction can facilitate the intensive analysis of documents by providing semantic triples which consist of named entities and their relations recognized in the texts. However, most of the research so far has been carried out separately for named entity recognition and relation extraction as individual studies, and as a result, the effective performance evaluation of the entire information extraction systems was not performed properly. This paper introduces two models of end-to-end information extraction that can extract various entity names in clinical records and their relationships in the form of semantic triples, namely pipeline and joint models and compares their performances in depth. The pipeline model consists of an entity recognition sub-system based on bidirectional GRU-CRFs and a relation extraction module using multiple encoding scheme, whereas the joint model was implemented with a single bidirectional GRU-CRFs equipped with multi-head labeling method. In the experiments using i2b2/VA 2010, the performance of the pipeline model was 5.5% (F-measure) higher. In addition, through a comparative experiment with existing state-of-the-art systems using large-scale neural language models and manually constructed features, the objective performance level of the end-to-end models implemented in this paper could be identified properly.

• Smart Media Journal
• /
• v.11 no.10
• /
• pp.65-75
• /
• 2022

Deep learning is used as a creative tool that could overcome the limitations of existing analysis models and generate various types of results such as text, image, and music. In this paper, we propose a method necessary to preprocess audio data using the Niko's MIDI Pack sound source file as a data set and to generate music using Bi-LSTM. Based on the generated root note, the hidden layers are composed of multi-layers to create a new note suitable for the musical composition, and an attention mechanism is applied to the output gate of the decoder to apply the weight of the factors that affect the data input from the encoder. Setting variables such as loss function and optimization method are applied as parameters for improving the LSTM model. The proposed model is a multi-channel Bi-LSTM with attention that applies notes pitch generated from separating treble clef and bass clef, length of notes, rests, length of rests, and chords to improve the efficiency and prediction of MIDI deep learning process. The results of the learning generate a sound that matches the development of music scale distinct from noise, and we are aiming to contribute to generating a harmonistic stable music.

• Geomechanics and Engineering
• /
• v.32 no.3
• /
• pp.307-319
• /
• 2023

Ground fissures have a huge effect on the integrity of surface structures. In high-intensity ground fissure regions, however, land resource would be wasted and city building and economic development would be limited if the area avoiding principle was used. In view of this challenge, to reveal the seismic response and seismic failure characteristics of ground fissure sites, a shaking table test on model soil based on a 1:15 scale experiment was carried out. In the test, the spatial distribution characteristics of acceleration response and Arias intensity were obtained for a site exposed to earthquakes with different characteristics. Furthermore, the failure characteristics and damage evolution of the model soil were analyzed. The test results indicated that, with the increase in the earthquake acceleration magnitude, the crack width of the ground fissure enlarged from 0 to 5 mm. The soil of the hanging wall was characterized by earlier cracking and a higher abundance of secondary fissures at 45°. Under strong earthquakes, the model soil, especially the soil near the ground fissure, was severely damaged and exhibited reduced stiffness. As a result, its natural frequency also decreased from 11.41 Hz to 8.05 Hz, whereas the damping ratio increased from 4.8% to 9.1%. Due to the existence of ground fissure, the acceleration was amplified to nearly 0.476 m/s², as high as 2.38 times of the input acceleration magnitude. The maximum of acceleration and Arias intensity appeared at the fissure zone, which decreased from the main fissure toward both sides, showing hanging wall effects. The seismic intensity, duration and frequency spectrum all had certain effects on the seismic response of the ground fissure site, but their influence degrees were different. The seismic response of the site induced by the seismic wave that had richer low-frequency components and longer duration was larger. The discrepancies of seismic response between the hanging wall and the footwall declined obviously when the magnitude of the earthquake acceleration increased. The research results will be propitious to enhancing the utilizing ratio of the limited landing resource, alleviation of property damages and casualties, and provide a good engineering application foreground.

• Journal of Korea Water Resources Association
• /
• v.55 no.spc1
• /
• pp.1283-1293
• /
• 2022

The purpose of this study is to predict residual chlorine in order to maintain stable residual chlorine concentration in sedimentation basin by using artificial intelligence algorithms in water treatment process employing pre-chlorination. Available water quantity and quality data are collected and analyzed statistically to apply into mathematical multiple regression and artificial intelligence models including multi-layer perceptron neural network, random forest, long short term memory (LSTM) algorithms. Water temperature, turbidity, pH, conductivity, flow rate, alkalinity and pre-chlorination dosage data are used as the input parameters to develop prediction models. As results, it is presented that the random forest algorithm shows the most moderate prediction result among four cases, which are long short term memory, multi-layer perceptron, multiple regression including random forest. Especially, it is result that the multiple regression model can not represent the residual chlorine with the input parameters which varies independently with seasonal change, numerical scale and dimension difference between quantity and quality. For this reason, random forest model is more appropriate for predict water qualities than other algorithms, which is classified into decision tree type algorithm. Also, it is expected that real time prediction by artificial intelligence models can play role of the stable operation of residual chlorine in water treatment plant including pre-chlorination process.

• The Journal of Engineering Geology
• /
• v.32 no.1
• /
• pp.59-72
• /
• 2022

This study employed a 3-D numerical analysis based on the distinct element method to estimate the strength and deformability of a Cretaceous biotite granitic rock mass at Gijang, Busan, Korea. A workflow was proposed to evaluate the scale effect and the representative elementary volume (REV) of mechanical properties for fractured rock masses. Directional strength and deformability parameters such as block strength, deformation modulus, shear modulus, and bulk modulus were estimated for a discrete fracture network (DFN) in a cubic block the size of the REV. The size of the mechanical REV for fractured rock masses in the study area was determined to be a 15 m cube. The mean block strength and mean deformation modulus of the DFN cube block were found to be 52.8% and 57.7% of the intact rock's strength and Young's modulus, respectively. A constitutive model was derived for the study area that describes the linear-elastic and orthotropic mechanical behavior of the rock mass. The model is expected to help evaluate the stability of tunnels and underground spaces through equivalent continuum analysis.