• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.4
• /
• pp.157-163
• /
• 2022

Recently, with the development of automation in the industrial field, research on anomaly detection is being actively conducted. An application for anomaly detection used in factory automation is camera-based defect inspection. Vision camera inspection shows high performance and efficiency in factory automation, but it is difficult to overcome the instability of lighting and environmental conditions. Although camera inspection using deep learning can solve the problem of vision camera inspection with much higher performance, it is difficult to apply to actual industrial fields because it requires a huge amount of normal and abnormal data for learning. Therefore, in this study, we propose a network that overcomes the problem of collecting abnormal data with 72 geometric transformation deep learning methods using only normal data and adds an outlier exposure method for performance improvement. By applying and verifying this to the MVTec data set, which is a database for auto-mobile parts data and outlier detection, it is shown that it can be applied in actual industrial sites.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.5
• /
• pp.29-36
• /
• 2009

Sequential pattern mining, which discovers frequent subsequences as patterns in a sequence database, is an important data mining problem with broad applications. Since a sequential pattern in DNA sequences can be a motif, we studied to find sequential patterns in DNA sequences. Most previously proposed mining algorithms follow the exact matching with a sequential pattern definition. They are not able to work in noisy environments and inaccurate data in practice. Theses problems occurs frequently in DNA sequences which is a biological data. We investigated approximate matching method to deal with those cases. Our idea is based on the observation that all occurrences of a frequent pattern can be classified into groups, which we call approximated pattern. The existing PrefixSpan algorithm can successfully find sequential patterns in a long sequence. We improved the PrefixSpan algorithm to find approximate sequential patterns. The experimental results showed that the number of repeats from the proposed method was 5 times more than that of PrefixSpan when the pattern length is 4.

• Composites Research
• /
• v.36 no.5
• /
• pp.329-334
• /
• 2023

The classical multiscale finite element (FE² ) method involves iterative calculations of micro-boundary value problems for representative volume elements at every integration point in macro scale, making it a computationally time and data storage space. To overcome this, we developed the data-driven multiscale analysis method based on the mean-field homogenization (MFH). Data-driven computational mechanics (DDCM) analysis is a model-free approach that directly utilizes strain-stress datasets. For performing multiscale analysis, we efficiently construct a strain-stress database for the microstructure of composite materials using mean-field homogenization and conduct data-driven computational mechanics simulations based on this database. In this paper, we apply the developed multiscale analysis framework to an example, confirming the results of data-driven computational mechanics simulations considering the microstructure of a hyperelastic composite material. Therefore, the application of data-driven computational mechanics approach in multiscale analysis can be applied to various materials and structures, opening up new possibilities for multiscale analysis research and applications.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.1
• /
• pp.551-557
• /
• 2024

Hydrogen is a renewable energy source with various characteristics such as clean, carbon-free and high-energy, and is internationally recognized as a "future energy". With the rapid development of the hydrogen energy industry, more hydrogen infrastructure is needed to meet the demand for hydrogen. However, hydrogen infrastructure accidents have been occurring frequently, hindering the development of the hydrogen industry. HyRAM+, developed by Sandia National Laboratories, is a software toolkit that integrates data and methods related to hydrogen safety assessments for various storage applications, including hydrogen refueling stations. HyRAM+'s physics mode simulates hydrogen leak results depending on the hydrogen refueling station components, graphing gas plume dispersion, jet frame temperature and trajectory, and radiative heat flux. In this paper, hydrogen leakage data was extracted from a hydrogen refueling station in Samcheok, Gangwon-do, using HyRAM+ software. A hydrogen leakage simulator was developed using data extracted from HyRAM+. It was implemented as a dashboard that shows the data generated by the simulator using a database and Grafana.

• Journal of the Korea Organic Resources Recycling Association
• /
• v.17 no.4
• /
• pp.48-58
• /
• 2009

Glycosyl hydrolases (GH, EC 3.2.1.-) are key enzymes which can hydrolyze the glycosidic bonds between two or more carbohydrates, or between a carbohydrate and a non-carbohydrate moiety. The new enzyme nomenclature of glycoside hydrolases is based on their amino acid sequence similarity and structural features. Here, we examined the glycosyl hydrolase family(GHF) genes reported from earthworm species. Among overall 115 GHFs, 12 GHFs could be identified from earthworm species through CAZy database. Of 12 GHF group genes, five genes including GHF2, 5, 17, 18, 20 are thought to be potent for industrial applications. The alignment of these genes with same genes from other animal species exhibited high sequence homology and some important amino acid residues necessary for enzyme activity appeared to be conserved. These genes can be utilized as a pest control agent or applicable to the food industry, clinical therapeutics and organic wastes disposition.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.9
• /
• pp.429-436
• /
• 2024

As Large Language Models (LLMs) continue to advance, effectively harnessing their potential has become increasingly important. LLMs, trained on vast datasets, are capable of generating text across a wide range of topics, making them useful in applications such as content creation, machine translation, and chatbots. However, they often face challenges in generalization due to gaps in specific or specialized knowledge, and updating these models with the latest information post-training remains a significant hurdle. To address these issues, Retrieval-Augmented Generation (RAG) models have been introduced. These models enhance response generation by retrieving information from continuously updated external databases, thereby reducing the hallucination phenomenon often seen in LLMs while improving efficiency and accuracy. This paper presents the foundational architecture of RAG, reviews recent research trends aimed at enhancing the retrieval capabilities of LLMs through RAG, and discusses evaluation techniques. Additionally, it explores performance optimization and real-world applications of RAG in various industries. Through this analysis, the paper aims to propose future research directions for the continued development of RAG models.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.4B
• /
• pp.646-680
• /
• 2000

In this paper, we discuss the design and implementation of a concurrency control manager for a main memory DBMS(MMDBMS). Since an MMDBMS, unlike a disk-based DBMS, performs all of data update or retrieval operations by accessing main memory only, the portion of the cost for concurrency control in the total cost for a data update or retrieval is fairly high. Thus, the development of an efficient concurrency control manager highly accelerates the performance of the entire system. Our concurrency control manager employs the 2-phase locking protocol, and has the following characteristics. First, it adapts the partition, an allocation unit of main memory, as a locking granule, and thus, effectively adjusts the trade-off between the system concurrency and locking cost through the analysis of applications. Second, it enjoys low locking costs by maintaining the lock information directly in the partition itself. Third, it provides the latch as a mechanism for physical consistency of system data. Our latch supports both of the shared and exclusive modes, and maximizes the CPU utilization by combining the Bakery algorithm and Unix semaphore facility. Fourth, for solving the deadlock problem, it periodically examines whether a system is in a deadlock state using lock waiting information. In addition, we discuss various issues arising in development such as mutual exclusion of a transaction table, mutual exclusion of indexes and system catalogs, and realtime application supports.

• Korean Journal of Cognitive Science
• /
• v.27 no.1
• /
• pp.27-41
• /
• 2016

Named entity recognition is to classify elements in text into predefined categories and used for various departments which receives natural language inputs. In this paper, we propose a method which can generate named entity training corpus automatically using knowledge bases. We apply two different methods to generate corpus depending on the knowledge bases. One of the methods attaches named entity labels to text data using Wikipedia. The other method crawls data from web and labels named entities to web text data using Freebase. We conduct two experiments to evaluate corpus quality and our proposed method for generating Named entity recognition corpus automatically. We extract sentences randomly from two corpus which called Wikipedia corpus and Web corpus then label them to validate both automatic labeled corpus. We also show the performance of named entity recognizer trained by corpus generated in our proposed method. The result shows that our proposed method adapts well with new corpus which reflects diverse sentence structures and the newest entities.

• Journal of the Korean Applied Science and Technology
• /
• v.36 no.4
• /
• pp.1119-1127
• /
• 2019

This study analyzes the content of niacin, B1, and B2, which are among the water-soluble vitamin B group, in cultivars of the commonly consumed agricultural products of apples, peaches, nectarines and strawberries to compare content differences and to use results as base material for the Korean Food Composition Table. While the vitamin B1 content of apples according to different cultivars was found to be within the ranges of 0.063-0.208 mg/100g, and the content of vitamin B2 was found to be within the value ranges of 0.006-0.031 mg/100g, no niacin was found. The vitamin B1 content of peaches and nectarines according to different cultivars was found to be within the value ranges of 0.014-0.276 mg/100g, the content of vitamin B2 was found to be within he value ranges of 0.019-0.042 mg/100g, and niacin content was found to be within the value ranges of 0.298-1.096 mg/100g. The vitamin B1 content of strawberries according to cultivars was found to be within the value ranges of 0.112-0.394 mg/100g, the content of vitamin B2 was found to within the value ranges of 0.001-0.027 mg/100g, and niacin content was found to be within the value ranges of 0.388-0.809 mg/100g. Therefore, when nutrient composition analysis databases for the fruits of apples, peaches, and strawberries are constructed, cultivar factors must be put into consideration. In addition, differences can be found according to fruit harvest times, cultivation methods, and environmental factors, so related additional is needed.

• KIISE Transactions on Computing Practices
• /
• v.21 no.2
• /
• pp.101-108
• /
• 2015

This paper presents the implementation of a system that recommends empirical best routes based on classification of large trajectory data. As many location-based services are used, we expect the amount of location and trajectory data to become big data. Then, we believe we can extract the best empirical routes from the large trajectory repositories. Large trajectory data is clustered into similar route groups using Hadoop MapReduce framework. Clustered route groups are stored and managed by a DBMS, and thus it supports rapid response to the end-users' request. We aim to find the best routes based on collected real data, not the ideal shortest path on maps. We have implemented 1) an Android application that collects trajectories from users, 2) Apache Hadoop MapReduce program that can cluster large trajectory data, 3) a service application to query start-destination from a web server and to display the recommended routes on mobile phones. We validated our approach using real data we collected for five days and have compared the results with commercial navigation systems. Experimental results show that the empirical best route is better than routes recommended by commercial navigation systems.