• Proceedings of the Korea Water Resources Association Conference
• /
• 한국수자원학회 2011년도 학술발표회
• /
• pp.28-28
• /
• 2011

치수는 고대국가로부터 현재의 국가에 이르기까지 중차대한 국가사업이다. 근대를 지나면서 산업화가 가속화되었고 수자원의 사용과 관리는 더욱 세분화 되었다. 또한 수자원의 사용과 관리에 앞서 기후와 밀접한 관계가 있는 수자원의 물리적 특성을 분석하는 연구가 활발히 이루어지고 있다. 최근 우리나라도 급격한 산업화로 인한 환경파괴와 지구온난화로 인한 강수의 편중이 더욱 심해지고 있다. 즉, 한반도도 전 세계가 직면한 기후변화로 인한 자연재해로 부터 안전하지 않다는 것을 의미한다. 특히 수자원을 관리함에 있어 가뭄의 경우에는 장기적으로 진행되는 경우가 많고 피해 규모와 복구 등도 가뭄의 진행 기간과 밀접한 관계를 가지므로 적극적인 대비가 필요하다. 따라서 가뭄을 연구함에 앞서 과거 한반도의 가뭄의 경향성 및 주기성 같은 특성을 분석할 수 있는 연구가 수반되어야 할 것으로 사료된다. 이에 따라 본 논문에서는 과거 한반도 가뭄 사상의 특성분석을 위해 대표 가뭄지수로 SPI(Standardized Precipitation Index)와 PDSI(Palmer Drought Severity Index)를 선정하여 우리나라 전역에 위치한 총 59개 기상관측소의 1980년~2009년까지의 기상자료를 수집하여 유역별 월평균 가뭄지수를 산정하였다. 또한 이를 이용하여 가뭄발생의 경향성 및 주기성을 중심으로 과거한반도 가뭄의 통계학적 특성분석을 실시하였다. 각 지수의 경향성을 분석한 결과, SPI3와 SPI6는 봄과 겨울에는 가뭄이 심화되는 경향을 보였고 여름철에는 가뭄이 완화되는 경향을 보였다. 반면 SPI12의 경우는 섬진강과 영산강은 유의하지는 않으나 전 기간에 걸쳐 가뭄이 심화되는 경향을 보였고 한강, 낙동강, 금강유역은 가뭄이 완화되는 것으로 나타났다. PDSI의 경우에는 SPI와는 반대로 섬진강과 영산강은 전 기간에 걸쳐 가뭄이 완화되는 경향을 보였고 한강, 낙동강, 금강유역은 유의한 수준에서 가뭄이 심화되는 경향을 보임으로써, SPI와 PDSI에 의한 가뭄평가 경향성이 유역별로 다소 다르게 나타나는 것을 확인할 수 있었다. 한편, 각 유역의 주기성을 분석한 결과로는 모든 유역에서 1년~3년 또는 6년 이하의 주기성을 나타냈으며, SPI의 경우 지속기간이 길어질수록 6년 이상의 저빈도 주기성을 나타냈고, SPI3는 짧게는 1년 미만의 고빈도 주기를 보이는 경우도 있었으며, SPI6와 SPI12는 4년~6년 주기를 나타냈다. 또한 PDSI도 마찬가지로 6년 내외의 장주기를 보였다. 특히 유역별로 분석할 경우 남부지역의 가뭄발생 주기가 중부지역보다 길게 나타나는 성향을 보였다. 그리고 SPI와 PDSI, 두 가뭄지수의 공통주기를 분석한 결과에서는 상관관계가 적은 것으로 확인 되었다.

• KIPS Transactions on Software and Data Engineering
• /
• 제8권5호
• /
• pp.213-222
• /
• 2019

Determining the meaning of a keyword in a speech dialogue system is an important technology for the future implementation of an intelligent speech dialogue interface. After extracting keywords to grasp intention from user's utterance, the intention of utterance is determined by using the semantic mark of keyword. One keyword can have several semantic marks, and we regard the task of attaching the correct semantic mark to the user's intentions on these keyword as a problem of word sense disambiguation. In this study, about 23% of all keywords in the corpus is manually tagged to build a semantic mark dictionary, a synonym dictionary, and a context vector dictionary, and then the remaining 77% of all keywords is automatically tagged. The semantic mark of a keyword is determined by calculating the context vector similarity from the context vector dictionary. For an unregistered keyword, the semantic mark of the most similar keyword is attached using a synonym dictionary. We compare the performance of the system with manually constructed training set and semi-automatically expanded training set by selecting 3 high-frequency keywords and 3 low-frequency keywords in the corpus. In experiments, we obtained accuracy of 54.4% with manually constructed training set and 50.0% with semi-automatically expanded training set.

• Journal of the Korea Convergence Society
• /
• 제12권11호
• /
• pp.457-466
• /
• 2021

This study sought to identify the effectiveness of training programs by developing mobile app-based training programs to enhance memory, attention, and language function, which is known to be vulnerable to menopause women. It was conducted on 40 Climacteric woman between 40 to 60 years complaining about cognitive function decline. The mobile app-based cognitive training was an 8 week program. There were a total of 24 sessions and each session took 20-30 minutes, three times per week. The survey was carried out including a baseline study pre and post intervention study. The research variables were objective cognitive function (overall cognitive function, memory, attention, and language function), subjective cognitive function and quality of life. The cognitive training program showed a significant increase in overall cognitive function(t=-8.688, p<.001), memory(t=-4.765, p<.001), attention : number of correct answers(t=-7.293, p<.001), language high frequency response speed(Z=-2.179, p=.036), language low frequency response speed(Z=-2.737, p=.009) and quality of life (t=-3.358, p=.002). However, there was no significant difference in the scores for subjective cognitive function. The cognitive training program was found to be an effective intervention for improving the cognitive function of Climacteric women. It could be used as a cognitive intervention tool that is accessible at home without expert help.

• Journal of Korea Water Resources Association
• /
• 제45권9호
• /
• pp.887-900
• /
• 2012

Drought is a non-negligible disaster of nature and it is mainly caused by rainfall shortage for a long time though there are many definitions of drought. 'Standard Precipitation Index' (SPI) that is widely used to express the level of meteorological drought intensity has a limit of not being able to consider the hydrological changes such as rainfall and evapotranspiration caused by climate change, because it does not consider the temperature-related variables other than the precipitation. Recently, however, 'Standardized Precipitation Evapotranspiration Index' (SPEI), a drought index of new concept which is similar to SPI but can reflect the effect of temperature variability as well as the rainfall change caused by climate variation, was developed. In this study, the changes of drought occurrence in South Korea were analyzed by applying SPEI for meteorological data (1973~2011) of 60 climate observatories under Korea Meteorological Administration (KMA). As the result of application, both of SPI and SPEI showed the trend of deepening drought in spring and winter and mitigating drought in summer for the entire nation, with SPI showing greater drought intensity than SPI. Also, SPI and SPEI with 12 months of duration showed that severe droughts with low frequency of around 6 years are generally being repeated.

• Journal of the Korean Society of Hazard Mitigation
• /
• 제10권3호
• /
• pp.119-131
• /
• 2010

Drought is one of disaster causing factors to produce severe damage in the World because drought is destroyed to the ecosystem as well as to make difficult the economy of the drought area. This study, using Palmer Drought Severity Index carries out correlation analysis with sea surface temperatures. Comparative analysis carries out by calculated Palmer Drought Severity Index and past drought occurrence year. Result of comparative analysis, PDSI indexes were in accord with the past drought. Cluster analysis for correlation analysis carries out using precipitation and temperature that is input datas palmer drought severity index, and the result of cluster analysis was classified as 6. Also, principal component carries out using result of cluster analysis. 14 principal component analyze out through principal component analysis. Using analyzed 14 principal component carries out correlation analysis with sea surface temperature that is delay time from 0month until 11month. Correlation analysis carries out sea surface temperatures and calculated cycle component of the low frequency through Wavelet Transform analysis form principal component. Result of correlation analysis, yang(+) correlation is bigger than yin(-) correlation. It is possible to check similar correlation statistically the area of sea surface temperature with sea surface temperature in the Pacific. Forecasting possibility of the future drought make propose using sea surface temperature.

• Journal of Acupuncture Research
• /
• 제25권4호
• /
• pp.1-9
• /
• 2008

Objectives : This study was to investigate the effect of low frequency electroacupuncture according to acupoints combination on NOS positive neurons in the PAG and Hippocampus of Rat with adjuvant induced Rheumatoid arthritis. Methods : Arthritis was induced by intradermal injection of FCA into base of tail. Experimental group were divided into 6 group ; Normal, Control, $ST_{36}$, $SP_9$, $ST_{36}+SP_9$, and NA. Normal group was injected with normal saline and not treated electroacupuncture. Control group was injected with FCA and not treated electroacupuncture. ST36 group was injected with FCA and treated by 2 Hz electroacupuncture at $ST_{36}$. $SP_9$ group was injected with FCA and treated by 2 Hz electroacupuncture at $SP_9$. $ST_{36}+SP_9$ group was injected with FCA and treated by 2 Hz electroacupuncture at $ST_{36}+SP_9$. NA group was injected with FCA and treated by 2 Hz electroacupuncture at non acupoint. Each groups were evaluated by the number of NOS-positive neurons in PAG and hippocampus by using an image analyzer and a microscope. Results : 1. The number of NOS-positive neurons of the NA group were significantly decreased in the CA2-3. 2. The number of NOS-positive neurons of the ST36 group were significantly decreased in the PAG and CA2-3. 3. The number of NOS-positive neurons of the SP9 group were significantly decreased in the PAG and DG 4. The number of NOS-positive neurons of the ST36+SP9 group were significantly decreased in the PAG, CA1 and DG. Conclusions : It is expected that low frequency electroacupuncture can be used a treatment of rheumatoid arthritic pain and that treatment of electroacupuncture at combined acupoints is more efficient than treatment of low frequency electroacupuncture at one acupoint.

• Journal of Korea Water Resources Association
• /
• 제47권1호
• /
• pp.83-93
• /
• 2014

This research was carried out for suggesting design criteria and procedure for maximizing flood control capacity by building flood control facilities like flood retention basin built in connection with existing facilities in order to cope with increased uncertainty due to factors such as urbanization and climate change. We suggested the procedure for the analysis under the various scenarios applicable for the cases of determining retention basin capacity as provision for the flood water level increase in main river channel or estimating flood water level reduction effect when retention basin capacity is given. Procedure for estimating design flood hydrograph at any duration using Intensity-Duration-Quantity (IDQ) originated from the existing IDF, and its application example were provided. Based on rainfall estimated by the IDQ analysis, it is possible to calculate an equivalent peak hydrographs under various scenarios, e.g. lower frequency hydrograph under same rainfall duration with water level higher than existing hydrograph, hydrograph with same peak and higher volume due to increased rainfall duration, hydrograph with higher peak and volume than existing hydrograph, etc.

• Smart Media Journal
• /
• 제8권2호
• /
• 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.

• The Journal of the Korea Contents Association
• /
• 제21권10호
• /
• pp.702-711
• /
• 2021

In order to examine the type of PAI profile for criminally charged juvenile offenders, the study sought to find out personality types and behavioral characteristics of 251 juvenile offenders and 173 ordinary adolescents from 2018-2020. Looking at the overall profiles of juvenile offenders and ordinary adolescents, we found differences in Infrequency(INF), Mania(MAN), Antisocial Features(ANT), Alcohol problems(ALC), Drug problems(DRG), Aggression (AGG), and Dominance(DOM) and Warmth(WRM). Based on these results, we perform clustering as factors Antisocial features(ANT), Aggression(AGG), and Dominance (DOM) with an average difference of more than five points. It was classified as Cluster 1 with a high percentage of adolescents and Cluster 2 with a high percentage of juvenile offenders, and Cluster 1 was named as a defense group because it showed a similar model to the profile of ordinary adolescents. The profile type of cluster 2 was named externalization, which can be represented as an externalization group. The results were similar to previous studies, and the profile type of juvenile offenders has higher overall clinical scale than that of ordinary adolescents, indicating behavioral problems. Continued research on juvenile offenders could lead to understanding of youth as well as juvenile offenders.

• Journal of the Korean Society for Library and Information Science
• /
• 제57권2호
• /
• pp.435-452
• /
• 2023

This study experimented with automatic classification of subject headings using BERT-based transfer learning model, and analyzed its performance. This study analyzed the classification performance according to the main class of KDC classification and the category type of subject headings. Six datasets were constructed from Korean national bibliographies based on the frequency of the assignments of subject headings, and titles were used as classification features. As a result, classification performance showed values of 0.6059 and 0.5626 on the micro F1 and macro F1 score, respectively, in the dataset (1,539,076 records) containing 3,506 subject headings. In addition, classification performance by the main class of KDC classification showed good performance in the class General works, Natural science, Technology and Language, and low performance in Religion and Arts. As for the performance by the category type of the subject headings, the categories of plant, legal name and product name showed high performance, whereas national treasure/treasure category showed low performance. In a large dataset, the ratio of subject headings that cannot be assigned increases, resulting in a decrease in final performance, and improvement is needed to increase classification performance for low-frequency subject headings.