• Smart Media Journal
• /
• v.8 no.2
• /
• pp.58-65
• /
• 2019

Automatic text summarization is a process of shortening a text document by either extraction or abstraction. The abstraction approach inspired by deep learning methods scaling to a large amount of document is applied in recent work. Abstractive text summarization involves utilizing pre-generated word embedding information. Low-frequent but salient words such as terminologies are seldom included to dictionaries, that are so called, out-of-vocabulary(OOV) problems. OOV deteriorates the performance of Encoder-Decoder model in neural network. In order to address OOV words in abstractive text summarization, we propose a copy mechanism to facilitate copying new words in the target document and generating summary sentences. Different from the previous studies, the proposed approach combines accurate pointing information and selective copy mechanism based on bidirectional RNN and bidirectional LSTM. In addition, neural network gate model to estimate the generation probability and the loss function to optimize the entire abstraction model has been applied. The dataset has been constructed from the collection of abstractions and titles of journal articles. Experimental results demonstrate that both ROUGE-1 (based on word recall) and ROUGE-L (employed longest common subsequence) of the proposed Encoding-Decoding model have been improved to 47.01 and 29.55, respectively.

• Journal of Industrial Convergence
• /
• v.21 no.1
• /
• pp.159-165
• /
• 2023

While the frequency of seismic occurrence has been increasing recently, the domestic seismic response system is weak, the objective of this research is to compare and analyze the seismic vulnerability of buildings using statistical analysis and machine learning techniques. As the result of using statistical technique, the prediction accuracy of the developed model through the optimal scaling method showed about 87%. As the result of using machine learning technique, because the accuracy of Random Forest method is 94% in case of Train Set, 76.7% in case of Test Set, which is the highest accuracy among the 4 analyzed methods, Random Forest method was finally chosen. Therefore, Random Forest method was derived as the final machine learning technique. Accordingly, the statistical analysis technique showed higher accuracy of about 87%, whereas the machine learning technique showed the accuracy of about 76.7%. As the final result, among the 22,296 analyzed building data, the seismic vulnerabilities of 1,627(0.1%) buildings are expected as more dangerous when the statistical analysis technique is used, 10,146(49%) buildings showed the same rate, and the remaining 10,523(50%) buildings are expected as more dangerous when the machine learning technique is used. As the comparison of the results of using advanced machine learning techniques in addition to the existing statistical analysis techniques, in spatial analysis decisions, it is hoped that this research results help to prepare more reliable seismic countermeasures.

• 한국체육학회지인문사회과학편
• /
• v.54 no.3
• /
• pp.439-452
• /
• 2015

The purpose of this study is to encourage continuous participation in sports and to provide basic data for the promotion of participation in leisure sports. To achieve the purpose, this study conducted factor scaling analysis on participation motives for leisure sports and subdivided them to analyze psychological reactions of participants. As for study methods, this study collected master and doctor's degree theses and academic journals on motives for sports participation that were conducted from 1997 to 2012 from Korean major search engines. On the search engines, a keyword 'motive' was searched first and then studies on participation motive for leisure sports were collected. Key words that appeared when searching 'motive' were combined with other key words and word spacing between them were checked before conducting a literature analysis. The study results showed that participation motives for leisure sports were divided into a participation motive, an internal motive, an external motive, a leisure motive and other motives. It was identified that there were 23 factors for the participation motive, 17 factors each for the internal motive and the external motive, 8 factors for the leisure motive and 57 factors for other motives. It was found out that 76 factors were used to study a participation motive for leisure sports, excluding the factors that have similar or overlapping meaning based on each factor.

• Journal of Periodontal and Implant Science
• /
• v.54 no.1
• /
• pp.13-24
• /
• 2024

Purpose: This study investigated the adjunctive effect of light-emitting diodes (LEDs) in the treatment of experimental periodontitis. Methods: Experimental periodontitis was induced by placing ligatures around the mandibular second, third, and fourth premolars of 6 beagles for 3 months. After ligature removal, periodontitis progressed spontaneously for 2 months. The animals' hemimandibles were allocated among the following 3 groups: 1) no treatment (control), 2) scaling and root planing (SRP), and 3) SRP with LED irradiation at 470-nm and 630-nm wavelengths (SRP/LED). The probing pocket depth (PPD) and gingival recession (GR) were measured at baseline, 6 weeks, and 12 weeks. The clinical attachment level (CAL) was calculated. After 12 weeks, histological and histomorphometric assessments were performed. The distances from the gingival margin to the apical extent of the junctional epithelium (E) and to the connective tissue (CT) attachment were measured, as was the total length of soft tissue (ST). Results: PPD and CAL increased at 12 weeks compared with baseline in the control group (6.31±0.43 mm to 6.93±0.50 mm, and 6.46±0.60 mm to 7.61±0.78 mm, respectively). PPD and CAL decreased at 12 weeks compared with baseline in the SRP group (6.01±0.59 to 4.81±0.65 mm, and 6.51±0.98 to 5.39±0.93 mm, respectively). PPD and CAL decreased at 12 weeks compared with baseline in the SRP/LED group (6.03±0.39 to 4.46±0.47 mm, and 6.11±0.47 to 4.78±0.57 mm, respectively). The E/ST and CT/ST ratios significantly differed among the 3 groups (P<0.05). The clinical parameters and histologic findings demonstrated that 470-nm and 630-nm wavelength LED irradiation accompanying SRP could improve treatment results. Conclusions: Within the study limitations, 470 nm and 630 nm wavelength LED irradiation might provide additional benefits for periodontitis treatment.

• Journal of Ecology and Environment
• /
• v.48 no.1
• /
• pp.32-48
• /
• 2024

Background: Longitudinal connectivity in river systems strongly affects biological components related to ecosystem functioning, thereby playing an important role in shaping local biodiversity and ecosystem health. Environmental DNA (eDNA)-based metabarcoding has an advantage of enabling to sensitively diagnose the presence/absence of species, becoming an efficient/effective approach for studying the community structure of ecosystems. However, little attention has been paid to eDNA-based biomonitoring for river systems, particularly for assessing the river longitudinal connectivity. In this study, by using eDNA we analyzed and compared species diversity and composition among artificial barriers to assess the longitudinal connectivity of the fish community along down-, mid- and upstream in the Hotancheon from the Geum River basin. Moreover, we investigated temporal variation in eDNA fish community structure and species diversity according to season. Results: The results of species detected between eDNA and conventional surveys revealed higher sensitivity for eDNA and 61% of species (23/38) detected in both methods. The results showed that eDNA-based fish community structure differs from down-, mid- and upstream, and species diversity decreased from down to upstream regardless of season. We found that there was generally higher species diversity at the study sites in spring (a total number of species across the sites [n] = 29) than in autumn (n = 27). Nonmetric multidimensional scaling and heatmap analyses further suggest that there was a tendency for community clusters to form in the down-, mid- and upstream, and seasonal variation in the community structure also existed for the sites. Dominant species in the Hotancheon was Rhynchocypris oxycephalus (26.07%) regardless of season, and subdominant species was Nipponocypris koreanus (16.50%) in spring and Odontobutis platycephala (15.73%) in autumn. Artificial barriers appeared to negatively affect the connectivity of some fish species of high mobility. Conclusions: This study attempts to establish a biological monitoring system by highlighting the versatility and power of eDNA metabarcoding in monitoring native fish community and further evaluating the longitudinal connectivity of river ecosystems. The results of this study suggest that eDNA can be applied to identify fish community structure and species diversity in river systems, although some shortcomings remain still need to be resolved.

• Korean Journal of Remote Sensing
• /
• v.31 no.5
• /
• pp.449-459
• /
• 2015

The conventional National Forest Inventory(NFI)-based forest carbon stock estimation method is suitable for national-scale estimation, but is not for regional-scale estimation due to the lack of NFI plots. In this study, for the purpose of regional-scale carbon stock estimation, we created grid-based forest carbon stock maps using spatial ancillary data and two types of up-scaling methods. Chungnam province was chosen to represent the study area and for which the $5^{th}$ NFI (2006~2009) data was collected. The first method (method 1) selects forest type map as ancillary data and uses regression model for forest carbon stock estimation, whereas the second method (method 2) uses satellite imagery and k-Nearest Neighbor(k-NN) algorithm. Additionally, in order to consider uncertainty effects, the final AGB carbon stock maps were generated by performing 200 iterative processes with Monte Carlo simulation. As a result, compared to the NFI-based estimation(21,136,911 tonC), the total carbon stock was over-estimated by method 1(22,948,151 tonC), but was under-estimated by method 2(19,750,315 tonC). In the paired T-test with 186 independent data, the average carbon stock estimation by the NFI-based method was statistically different from method2(p<0.01), but was not different from method1(p>0.01). In particular, by means of Monte Carlo simulation, it was found that the smoothing effect of k-NN algorithm and mis-registration error between NFI plots and satellite image can lead to large uncertainty in carbon stock estimation. Although method 1 was found suitable for carbon stock estimation of forest stands that feature heterogeneous trees in Korea, satellite-based method is still in demand to provide periodic estimates of un-investigated, large forest area. In these respects, future work will focus on spatial and temporal extent of study area and robust carbon stock estimation with various satellite images and estimation methods.

• Korean Journal of Ecology and Environment
• /
• v.54 no.3
• /
• pp.199-208
• /
• 2021

Environmental DNA (eDNA) metabarcoding is effective method with high detection sensitivity for evaluating fish biodiversity and detecting endangered fish from natural water samples. We compared the richness of operational taxonomic units(OTUs) and composition of freshwater fishes according to filters(cellulose nitrate filter vs. glass fiber filter), extraction kits(DNeasy2^® Blood & Tissue Kit vs. DNeasy2^® PowerWater Kit), primer sets (12S rDNA vs. 16S rDNA), and PCR methods (conventional PCR vs. touchdown PCR) to determine the optimal conditions for metabarcoding analysis of Korean freshwater fish. The glass fiber filter and DNeasy2^® Blood & Tissue Kit combination showed the highest number of freshwater fish OTUs in both 12S and 16S rDNA. Among the four types, the primer sets only showed statistically significant difference in the average number of OTUs in class Actinopterygii (non-parametric Wilcoxon signed ranks test, p=0.005). However, there was no difference in the average number of OTUs in freshwater fish. The species composition also showed significant difference according to primer sets (PERMANOVA, Pseudo-F=6.9489, p=0.006), but no differences were observed in the other three types. The non-metric multidimensional scaling (NMDS) results revealed that species composition clustered together according to primer sets based on similarity of 65%; 16S rDNA primer set was mainly attributed to endangered species such as Microphysogobio koreensis and Pseudogobio brevicorpus. In contrast, the 12S rDNA primer set was mainly attributed to common species such as Zacco platypus and Coreoperca herzi. This study provides essential information on species diversity analysis using metabarcoding for environmental water samples obtained from rivers in Korea.

• Journal of the Korean Recycled Construction Resources Institute
• /
• v.6 no.4
• /
• pp.297-303
• /
• 2018

With the changing trend of the building construction to high rising and large scaling, the underground structure has been increased, and its usage also increased and variety. Hence, to protect the underground structure against underground water, various water proofing methods has been developed. Among the many water proofing methods, the combined water proofing method using both asphalt seal and sheet has been widely used to secure the sufficient performance and decrease the construction failure. However, during the summer period of extremely high temperature conditions, the asphalt sealing materials were separated and leaked into the structure. Therefore, the aim of the research is to provide the quality standard of asphalt sealing material based on the various temperature changes depending on seasons. According to the experimental results, the temperature of the sealing materials applied on the slab was increased up to $54^{\circ}C$ which was $3^{\circ}C$ higher than the structure temperature of $51^{\circ}C$. Based on the melting test for asphalt sealing materials applied on the outside wall of the structure, in the case of water-dispersing typed materials showed significant melting down due to its slow evaporation and low viscosity. Furthermore, from the accelerated test conducted indoor conditions, a solvent-type and water-dispersing typed materials showed significant melting down due to their low viscosity and melting point in ambient conditions. Based on these results, viscosity and melting point are found as the important factors on asphalt sealing materials' quality, and it is necessary to designate the quantitative level of the viscosity and melting point for quality control.

• Journal of dental hygiene science
• /
• v.10 no.1
• /
• pp.11-16
• /
• 2010

The subjects of this study were patients at scaling practice, the total number of the subjects was 249 of 138 men and 111 women. The results were analyzed using SPSS 12.0. 1. In oral health care, among tooth brushing methods, scrub method was found to be 65.9%(164 patients), and 83.9%(209 patients) did not use oral care aids. 2. Mean calculus index was $0.57{\pm}0.22$. 3. For calculus index per tooth brushing method, there was statistically significant difference in all of the upper jaw, the lower jaw, the anterior portion, and the posterior portion(p < 0.05). 4. For calculus index per frequency of tooth brushing, there was statistically significant difference in the upper jaw, the lower jaw, and the posterior portion(p < 0.05). 5. For calculus index per time of tooth brushing and per use of oral care aids, there was statistically significant difference in all of the upper jaw, the lower jaw, the anterior portion, and the posterior portion(p < 0.05). 6. For factors influencing calculus index, as a result of applying stepwise method based on 0.05 of significance level, age, use of oral care aids, sex, and time of tooth brushing were found to influence calculus index. Final regression model was calculus index = $0.362^*age+0.216^*$use of oral care aids - $0.161*sex-0.127^*$time of tooth brushing, and explanatory power of the model was 23.4%. $0.362^*age+0.216^*$use of oral care aids - $0.161^*sex-0.127^*$time of tooth brushing, and explanatory power of the model was 23.4%.

• Journal of dental hygiene science
• /
• v.16 no.6
• /
• pp.442-448
• /
• 2016

Dental caries and periodontal disease are considered to be chronic, but can be prevented through an incremental oral health program covering all ages. The National Oral Health Program for adults provides oral health exam and scaling, and is covered by national health insurance for those over 20 years of age in Korea. The aim of this study was to collect basic data for developing an oral health program for adults by identifying factors related to awareness and need. The data were obtained by convenience sampling of 303 subjects. The use of dental plaque disclosing agents affected tooth brushing frequency, toothbrushing time and use of oral auxiliary devices. Education on toothbrushing methods affected toothbrushing time and use of oral auxiliary devices. Of those surveyed, 93.1% replied that an incremental oral health program for adults was needed, and 68.0% intended to participate. In a regression model, the factors that had an effect on the perceived need for an oral health program were education level, use of oral hygiene auxiliary devices, and toothbrushing time, and the factors affecting intent to participate were education for prevention of periodontal disease and the use of oral hygiene auxiliary devices. The subjects stated that the following oral health programs were needed: an oral bacteria exam (74.3%), toothbrushing education (71.6%), a bad breath exam (69.3%), education on use of oral hygiene auxiliary devices (46.9%), a dental plaque exam (42.9%) and a saliva exam (37.6%). Oral health education appears to be an important factor for participation in an incremental oral health program.