• Investigative Magnetic Resonance Imaging
• /
• v.14 no.2
• /
• pp.103-114
• /
• 2010

Purpose : To generate a Korean specific brain template, especially in patients with Alzheimer's disease (AD) by optimizing the voxel-based analysis. Materials and Methods : Three-dimensional T1-weighted images were obtained from 123 subjects who were 43 cognitively normal subjects and patients with 44 mild cognitive impairment (MCI) and 36 AD. The template and the corresponding aprior maps were created by using the matched pairs approach with considering differences of age, gender and differential diagnosis (DDX). We measured several characteristics in both our and the MNI templates, including in the ventricle size. Also, the fractions of gray matter and white matter voxels normalized by the total intracranial were evaluated. Results : The high resolution template and the corresponding aprior maps of gray matter, white matter (WM) and CSF were created with the voxel-size of $1{\times}1{\times}1\;mm$. Mean distance measures and the ventricle sizes differed between two templates. Our brain template had less gray matter and white matter areas than the MNI template. There were volume differences more in gray matter than in white matter. Conclusion : Gray matter and/or white matter integrity studies in populations of Korean elderly and patients with AD are needed to investigate with this template.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.43-61
• /
• 2019

Development of technologies in artificial intelligence has been rapidly increasing with the Fourth Industrial Revolution, and researches related to AI have been actively conducted in a variety of fields such as autonomous vehicles, natural language processing, and robotics. These researches have been focused on solving cognitive problems such as learning and problem solving related to human intelligence from the 1950s. The field of artificial intelligence has achieved more technological advance than ever, due to recent interest in technology and research on various algorithms. The knowledge-based system is a sub-domain of artificial intelligence, and it aims to enable artificial intelligence agents to make decisions by using machine-readable and processible knowledge constructed from complex and informal human knowledge and rules in various fields. A knowledge base is used to optimize information collection, organization, and retrieval, and recently it is used with statistical artificial intelligence such as machine learning. Recently, the purpose of the knowledge base is to express, publish, and share knowledge on the web by describing and connecting web resources such as pages and data. These knowledge bases are used for intelligent processing in various fields of artificial intelligence such as question answering system of the smart speaker. However, building a useful knowledge base is a time-consuming task and still requires a lot of effort of the experts. In recent years, many kinds of research and technologies of knowledge based artificial intelligence use DBpedia that is one of the biggest knowledge base aiming to extract structured content from the various information of Wikipedia. DBpedia contains various information extracted from Wikipedia such as a title, categories, and links, but the most useful knowledge is from infobox of Wikipedia that presents a summary of some unifying aspect created by users. These knowledge are created by the mapping rule between infobox structures and DBpedia ontology schema defined in DBpedia Extraction Framework. In this way, DBpedia can expect high reliability in terms of accuracy of knowledge by using the method of generating knowledge from semi-structured infobox data created by users. However, since only about 50% of all wiki pages contain infobox in Korean Wikipedia, DBpedia has limitations in term of knowledge scalability. This paper proposes a method to extract knowledge from text documents according to the ontology schema using machine learning. In order to demonstrate the appropriateness of this method, we explain a knowledge extraction model according to the DBpedia ontology schema by learning Wikipedia infoboxes. Our knowledge extraction model consists of three steps, document classification as ontology classes, proper sentence classification to extract triples, and value selection and transformation into RDF triple structure. The structure of Wikipedia infobox are defined as infobox templates that provide standardized information across related articles, and DBpedia ontology schema can be mapped these infobox templates. Based on these mapping relations, we classify the input document according to infobox categories which means ontology classes. After determining the classification of the input document, we classify the appropriate sentence according to attributes belonging to the classification. Finally, we extract knowledge from sentences that are classified as appropriate, and we convert knowledge into a form of triples. In order to train models, we generated training data set from Wikipedia dump using a method to add BIO tags to sentences, so we trained about 200 classes and about 2,500 relations for extracting knowledge. Furthermore, we evaluated comparative experiments of CRF and Bi-LSTM-CRF for the knowledge extraction process. Through this proposed process, it is possible to utilize structured knowledge by extracting knowledge according to the ontology schema from text documents. In addition, this methodology can significantly reduce the effort of the experts to construct instances according to the ontology schema.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.189-198
• /
• 2024

Lipreading is one of the important parts of speech recognition, and several studies have been conducted to improve the performance of lipreading in lipreading systems for speech recognition. Recent studies have used method to modify the model architecture of lipreading system to improve recognition performance. Unlike previous research that improve recognition performance by modifying model architecture, we aim to improve recognition performance without any change in model architecture. In order to improve the recognition performance without modifying the model architecture, we refer to the cues used in human lipreading and set other regions such as chin and cheeks as regions of interest along with the lip region, which is the existing region of interest of lipreading systems, and compare the recognition rate of each region of interest to propose the highest performing region of interest In addition, assuming that the difference in normalization results caused by the difference in interpolation method during the process of normalizing the size of the region of interest affects the recognition performance, we interpolate the same region of interest using nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation, and compare the recognition rate of each interpolation method to propose the best performing interpolation method. Each region of interest was detected by training an object detection neural network, and dynamic time warping templates were generated by normalizing each region of interest, extracting and combining features, and mapping the dimensionality reduction of the combined features into a low-dimensional space. The recognition rate was evaluated by comparing the distance between the generated dynamic time warping templates and the data mapped to the low-dimensional space. In the comparison of regions of interest, the result of the region of interest containing only the lip region showed an average recognition rate of 97.36%, which is 3.44% higher than the average recognition rate of 93.92% in the previous study, and in the comparison of interpolation methods, the bilinear interpolation method performed 97.36%, which is 14.65% higher than the nearest neighbor interpolation method and 5.55% higher than the bicubic interpolation method. The code used in this study can be found a https://github.com/haraisi2/Lipreading-Systems.

• Clean Technology
• /
• v.25 no.4
• /
• pp.283-288
• /
• 2019

Monodisperse microparticles has been particularly enabling for various applications in the encapsulation and delivery of pharmaceutical agents. The microfluidic devices are attractive candidates to produce highly uniform droplets that serve as templates to form monodisperse microparticles. The microfluidic devices that have micro-scale channel allow precise control of the balance between surface tension and viscous forces in two-phase flows. One of its essential abilities is to generate highly monodisperse droplets. In this paper, a microfluidic approach for preparing monodisperse polycaprolactone (PCL) microparticles is presented. The microfluidic devices that have a flow-focusing generator are manufactured by soft-lithography using polydimethylsiloxane (PDMS). The crucial factors in the droplet generation are the controllability of size and monodispersity of the microdroplets. For this, the volumetric flow rates of the dispersed phase of oil solution and the continuous phase of water to generate monodisperse droplets are optimized. As a result, the optimal flow condition for droplet dripping region that is able to generate uniform droplet is found. Furthermore, the droplets containing PCL polymer by solvent evaporation after collection of droplet from device is solidified to generate the microparticle. The particle size can be controlled by tuning the flow rate and the size of the microchannel. The monodispersity of the PCL particles is measured by a coefficient of variation (CV) below 5%.

• Journal of Distribution Science
• /
• v.14 no.11
• /
• pp.103-115
• /
• 2016

Purpose - Our research started out with an assumption that a further study about the competency of B2B and B2C salespeople is needed, and has thus focused on enhancing their abilities throughout various fields. It seems that the ability of sales representatives has a positive influence on the image and revenue of the company. Since businesses find it apparent that the role of salespeople will become significant in the future, our research aims to provide a foundation for future research on exploring the necessary competencies for salespeople by conducting an in-depth interview as well as survey. Research design, data, and methodology - The methods of our research can be broken down into four steps - an interview on salespeople, eliciting salespeople's potential competency, surveys, and analysis. First, our research team conducted interviews on forty subjects. Second, we strived to elicit potential competency of salespeople based on data gained from previous studies and in-depth interviews. And Third, we came up with our own survey templates. Last but not least, our research team analyzed the results from the surveys to elicit necessary capabilities for the salespeople. Results - The results of our research show a clear distinction between B2B and B2C salespeople on all categories that we measured such as the character fitness, competence of emotion and sales marketing. As for B2B salespeople, the results indicated openness(M: 3.8265) in character fitness, and self-motivation(M: 4.1887), group cognition(M: 3.8735), teamwork(M: 3.9956) in competence of emotion, and previous research(M: 3.8735), proposal of values(M: 4.3873), cooperation with other team(M: 4.0441) in competence of sales marketing. The difference in capability required between B2B salespeople and B2C salespeople was very pronounced. Conclusions - For future studies, enhanced pool of subjects with various backgrounds is needed in order for our research to reach a wide range of population. The results from our research are advised to be used for eliciting the competency required by salespeople and for a practical application to further enhance their competency. Companies need continuous efforts to develop the skills of salespeople based on competence analysis and research of sales representatives.

• Journal of Korean Society on Water Environment
• /
• v.34 no.1
• /
• pp.46-56
• /
• 2018

A mcyB-specific Ultra-Rapid quantitative PCR was developed for the quantitative detection of Microcystis aeruginosa, which is often a dominant species in green tide. McyB-specific UR-qPCR was optimized under extremely short times of each step in thermal cycles, based on the specific primers deduced from the mcyB in microcystin synthetase of M. aeruginosa. The M. aeruginosa strain KG07 was used as a standard for quantification, after the microscopic counting and calculation by mcyB-specific UR-qPCR. The water samples from the river water with the Microcystis outbreak were also measured by using both methods. The $1.0{\times}10^8$ molecules of mcyB-specific DNA was recognized inner 4 minutes after beginning of UR-qPCR, while $1.0{\times}10^4$ molecules of mcyB-specific templates was detected inner 7 minutes with quantitative manner. From the range of $1.0{\times}10^2$ to $1.0{\times}10^8$ initial molecules, quantification was well established based on $C_T$ using mcyB-specific UR-qPCR (Regression coefficiency, $R^2=0.9977$). Between the numbers of M. aeruginosa cell counting under microscope and calculated numbers using mcyB-specific UR-qPCR, some differences were often found. The reasons for these differences were discussed; therefore, easy compensation method was proposed that was dependent on the numbers of the cell counting. Additionally, to easily extract the genomic DNA (gDNA) from the samples, a freeze-fracturing of water-sample using liquid nitrogen was tested, by excluding the conventional gDNA extraction method. It was also verified that there were no significant differences using the UR-qPCR with both gDNAs. In conclusion, the mcyB-specific UR-qPCR that we proposed would be expected to be a useful tool for rapid quantification and easy monitoring of M. aeruginosa in environmental water.

• Journal of Internet Computing and Services
• /
• v.4 no.3
• /
• pp.57-67
• /
• 2003

Software product lines are to attain the rapid development of qualify applications by concretizing the general components populated in software assets and assembling them according to the predefined architectures. For supporting the construction of the software product lines, this paper proposes a component code generation techniques based on GenVoca architecture and XML/XSLT technologies, In addition, CompGenX(Component Generator using XML), a component code generation system, is proposed on the basis of this techniques. By providing reconfigurability of component at the time of code generation, CompGenX allows the reusers to create the component source code that is appropriate to their purpose, In this system, the process of the component development is divided into two tasks which are the component family construction task and the component reuse task, For the component family construction, CompGenX provides the feature modeling tool for domain analysis and the domain architecture definition tool. Also, it provides the tool for building the component configuration know1edge specification and the code templates, For the component reuse task, it offers the component family search tool. the component customizing tool and the component code generator. Component code generation techniques and system in this paper should be applicable as basic technology to build the component-based software product lines.

• Journal of Dental Rehabilitation and Applied Science
• /
• v.22 no.4
• /
• pp.329-340
• /
• 2006

Statement of problem: Damping of the peak force transmitted to implants has been reported by in vitro studies using impact forces on resin-veneered superstructures. Theoretical assumptions suggest that use of acrylic resin for the occlusal surfaces of a prosthesis would protect the connection between implant and bone. Therefore, the relationship between prosthesis materials and the force transmitted through the implant system also needs to be investigated under conditions that resemble the intraoral mechanical environment. Purpose: The purpose of this study was to analyze the fracture strength and modes of temporary prosthesis when a flange or occlusally extended structure were connected on the top of the abutment. Material and method: Modified abutments of winged and bulk design were made by casting the desired wax pattern which is made on the UCLA type plastic cylinder. Temporary crowns were made using templates on the modified abutments, and its fracture toughness and strain were compared to the traditional temporary prosthesis. To evaluate the effect of aging, 5.000 times of thermocycling were performed, and their result was compared to the 24hours specimen result. Results: The following conclusions were drawn from this study: 1. In the fracture toughness test, temporary crown's fracture line located next to the screw hole while modified designs with metal support showed fracture line on the metal and its propagation along the metal-resin interface. 2. Wing and bulk structure didn't show significant difference in the fracture toughness (p>0.05), but wing structure showed stress concentration on the screw hole area compared to bulk structure which showed even stress distribution. 3. In the fracture toughness test after thermocycling, wing and bulk structure showed increased or similar results in metal supported area while off-metal area and temporary crown showed decreased results. 4. In the strain measurement after thermocycling, its value increased in the temporary and bulk structure. However, wing structure showed decreased value in the loading point while increased value in the screw hole area. Conclusion: Wing type design showed compatible result to the bulk type that its application with composite resin prosthesis to the implant dentistry is considered promising.

• Journal of Life Science
• /
• v.24 no.2
• /
• pp.181-185
• /
• 2014

We identified SNPs (single nucleotide polymorphisms) for endothelial nitric oxide synthase (eNOS) genes in the Korean genome. eNOS is present in the vascular endothelium, platelets, and several other cell types that continuously produce modest amounts of NO. Endothelium-derived NO plays a key role in the regulation of vascular tone, and the impaired effects of NO on the cardiovascular system appear to be responsible for coronary atherosclerosis and thrombosis. In recent studies, a missense variant within exon 7 of the eNOS gene in patients with coronary spastic angina-GAG to GAT substitution, which results in the replacement of glutamic acid by aspartic acid (Glu298Asp [G894T])-has been identified and is known to be significantly associated with coronary spasm. We prepared PCR primers based on sequences in Genbank. Primers were prepared for normal and SNPs separately, as reported for other Asian countries, such as G894T. Their sequences were different only at the 3' ends so that primer extension could only by possible when base pairs between templates and primers matched. We also employed ARMS (Amplification Refractory Mutation System) technology to improve the specificity of the PCR reaction. In conclusion, we were able to demonstrate the eNOS G894A polymorphism in Korean gemone. This study should facilitate research on the cause of myocardial infarction and development on further therapy at the genetic level.

• Korean Journal of Microbiology
• /
• v.41 no.3
• /
• pp.168-176
• /
• 2005

Vibrio vulnificus, one of the marine bacterial pathogens causing septicemia, was detected using molecular methods, namely, PCR and/or Southern hybridization, and real-time PCR. Extracted and purified total DNAs by using commercial kits were used as templates for PCR. Multiplex-PCR was conducted by employing three sets of primers for the genes, hemolysin (vvhA), phosphomannomutase (pmm), and metalloprotease (vvpE), for V vulnificus virulence. The presence of DMSO ($5\%$) and BSA ($0.1\%$) in PCR reaction mixture improved a detection efficiency by higher PCR band intensities. TaqMan real-time PCR was carried out by using gene segment of vvhA as a target. Detection limit of PCR/Southern hybridization without enrichments was to be around $10^2\;cells\;g^{-1}$ of sample. However, those three methods using the enrichment at $35^{\circ}C$ in APW showed high sensitivity ($2\~10\;cells\;g^{-1}$ of sediments). Highly sensitive detection of V vulnificus by real-time PCR was achieved within $5\~6$ hr, whereas the detection by PCR/Southern hybridization required about 36 hr. Thus, it was evident that real-time PCR is the most rapid and efficient method for detecting V vulnificus in tidal flat sediments.