• Journal of the Korea Society of Computer and Information
• /
• v.27 no.2
• /
• pp.15-23
• /
• 2022

In the field of patents, as NLP(Natural Language Processing) is a challenging task due to the linguistic specificity of patent literature, there is an urgent need to research a language model optimized for Korean patent literature. Recently, in the field of NLP, there have been continuous attempts to establish a pre-trained language model for specific domains to improve performance in various tasks of related fields. Among them, ELECTRA is a pre-trained language model by Google using a new method called RTD(Replaced Token Detection), after BERT, for increasing training efficiency. The purpose of this paper is to propose KorPatELECTRA pre-trained on a large amount of Korean patent literature data. In addition, optimal pre-training was conducted by preprocessing the training corpus according to the characteristics of the patent literature and applying patent vocabulary and tokenizer. In order to confirm the performance, KorPatELECTRA was tested for NER(Named Entity Recognition), MRC(Machine Reading Comprehension), and patent classification tasks using actual patent data, and the most excellent performance was verified in all the three tasks compared to comparative general-purpose language models.

• KNOM Review
• /
• v.23 no.1
• /
• pp.34-42
• /
• 2020

Network Function Virtualization (NFV) environment can deal with dynamic changes in traffic status with appropriate deployment and scaling of Virtualized Network Function (VNF). However, determining and applying the optimal VNF deployment is a complicated and difficult task. In particular, it is necessary to predict the situation at a future point because it takes for the process to be applied and the deployment decision to the actual NFV environment. In this paper, we randomly generate service requests in Multiaccess Edge Computing (MEC) topology, then obtain training data for machine learning model from an Integer Linear Programming (ILP) solution. We use the simulation data to train the machine learning model which predicts the optimal VNF deployment in a predefined future point. The prediction model shows the accuracy over 90% compared to the ILP solution in a 5-minute future time point.

• KNOM Review
• /
• v.23 no.1
• /
• pp.43-50
• /
• 2020

As General Purpose Graphics Processing Unit (GPGPU) recently plays an essential role in high-performance computing, several cloud service providers offer GPU service. Most cluster orchestration platforms in a cloud environment using containers allocate the integer number of GPU to jobs and do not allow a node shared with other jobs. In this case, resource utilization of a GPU node might be low if a job does not intensively require either many cores or large size of memory in GPU. GPU virtualization brings opportunities to realize kernel concurrency and share resources. However, performance may vary depending on characteristics of applications running concurrently and interference among them due to resource contention on a node. This paper proposes GPU container co-execution framework with multiple server creation and execution based on Kubernetes, container orchestration platform for measuring interference which may be occurred by sharing GPU resources. Performance changes according to scheduling policies were investigated by executing several jobs on GPU. The result shows that optimal scheduling is not possible only considering GPU memory and computing resource usage. Interference caused by co-execution among applications is measured using the framework.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.35 no.4
• /
• pp.197-205
• /
• 2022

In this paper, a retaining wall system, developed using building Information modeling (BIM), is presented. Based on the information from a literature review, elementary technologies for the system were defined and developed. First, for the elementary technology, BIM libraries were constructed using standards and previous study results to achieve versatility and reusability. Second, methods for determining the quantity take-off (QTO) of a retaining wall were reviewed for an earth-work calculating system. Additionally, inverse distance weighting interpolation was used to generate topography. Finally, four formulas for calculating the QTO were proposed and devised for each element. After its development, the BIM system was analyzed and verified through comparison with a two-dimensional drawing-based QTO. The proposed system is deemed to be practical for determining the QTO of retaining walls and earth works. The contributions and limitations of the research are discussed in this paper.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.11
• /
• pp.67-80
• /
• 2022

In this paper, we propose a system which automatically generates the ShEx schemas to describe and validate RDF knowledge graphs constructed by RML mapping. ShEx schemas consist of constraints. The proposed system generates most of the constraints by converting the RML mapping rules. The schemas consisting only of constraints obtained from mapping rules can help users to figure out the structure of the graphs generated by RML mapping, but they are not sufficient for sophisticated validation purposes. For users who need a schema for validation, the proposed system is also able to provide the schema with added constraints generated from metadata extracted from the input data sources for RML mapping. The proposed system has the ability to handle CSV, XML, JSON or RDBMS as input data sources. Testing results from 297 cases show that the proposed system can be applied for RDF graph validation in various practical cases.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.35 no.5
• /
• pp.309-316
• /
• 2022

This study proposes a finite element analysis method that can analyze the vibration of a beam by considering the inertia effect of moving masses in a vertical direction. The proposed method is effective when a precise interaction analysis is not required. The inertial effects of the moving masses are included in the equation of motion, and the interaction forces between the masses and the beam are considered only as external loads. Time domain analyses were performed using Abaqus, a general-purpose finite element analysis software, and an implementation method using multi-point constraints wais presented to link the displacements of the beam element nodes and moving rigid masses. The proposed method was verified by comparing its solution with that obtained using an existing analytical method, and the analysis results for continuous beam vibrations under dynamic gait loadings were used to examine the mass effect of pedestrians.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.65 no.3
• /
• pp.29-42
• /
• 2023

Universal Soil Loss Equation (USLE) is suggested and employed in the policy to conserve soil resources and to manage the impact of development, since soil loss is very essential to nonpoint source pollution management. The equation requires only five factors to estimate average annual potential soil loss, USLE is simplicity provides benefits in use of the equation. However, it is also limitation of the model, since the estimated results are very sensitive to the five factors. There is a need to examine the application procedures. Three approaches to estimate potential soil loss were examined, In the first approach, all factors were prepared with raster data, soil loss were computed for each cell, and sum of all cell values was determined as soil loss for the watersheds. In the second approach, the mean values for each factor were defined as representing USLE factors, and then the five factors were multiplied to determine soil loss for the watersheds. The third approach was same as the second approach, except that the Vegetative and Mechanical measure was used instead of the Cover and management factor and Support practice factor. The approaches were applied in 38 watersheds, they displayed significant difference, moreover no trends were detected for the soil loss at watersheds with the approaches. Therefore, it was concluded that there is a need to be developed and provided a typical guideline or public systems so that soil loss estimations have consistency with the users.

• Journal of the Society of Disaster Information
• /
• v.19 no.3
• /
• pp.689-697
• /
• 2023

Purpose: The purpose of this study is to automatically extract information on underground facilities using a general-purpose smartphone in the process of applying the mini-trenching method. Method: Data sets for image learning were collected under various conditions such as day and night, height, and angle, and the object detection algorithm used the R-CNN algorithm. Result: As a result of the study, F1-Score was applied as a performance evaluation index that can consider the average of accurate predictions and reproduction rates at the same time, and F1-Score was 0.76. Conclusion: The results of this study showed that it was possible to extract information on underground buried materials based on smartphones, but it is necessary to improve the precision and accuracy of the algorithm through additional securing of learning data and on-site demonstration.

• Applied Chemistry for Engineering
• /
• v.34 no.4
• /
• pp.383-387
• /
• 2023

The quartz crystal microbalance (QCM), commonly used in high vacuum deposition, becomes difficult to use when a thick film is deposited on the quartz, affecting the crystal's inherent vibration. In this study, a non-destructive optical measurement method was developed to measure the film's deposition rate during the in-situ film deposition process. By measuring the scattered laser intensity caused by the dimer in the parylene gas passing through the gas flow path, it was successfully confirmed that the ratio of the dimer in the parylene gas increases as the pyrolysis temperature decreases. Additionally, it was noted that the film's thickness and haze increase as the pyrolysis temperature decreases by confirming the characteristics of the visible parylene films. Through the research results, we aim to utilize the stable in-situ film deposition rate measurement system to control the precise film deposition rate of parylene films.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.206-208
• /
• 2022

As the big data industry has recently developed significantly, interest in privacy violations caused by personal information leakage has increased. There have been attempts to automate this through named entity recognition in natural language processing. In this paper, named entity recognition data is constructed semi-automatically by identifying sentences with de-identification information from de-identification information in Korean Wikipedia. This can reduce the cost of learning about information that is not subject to de-identification compared to using general named entity recognition data. In addition, it has the advantage of minimizing additional systems based on rules and statistics to classify de-identification information in the output. The named entity recognition data proposed in this paper is classified into twelve categories. There are included de-identification information, such as medical records and family relationships. In the experiment using the generated dataset, KoELECTRA showed performance of 0.87796 and RoBERTa of 0.88.