• Proceedings of the Korea Information Processing Society Conference
• /
• 2010.04a
• /
• pp.505-507
• /
• 2010

웹(Web)상에 전문용어를 포함한 문서가 증가함에 따라 전문용어를 자동으로 추출하는 연구가 계속해서 이루어지고 있다. 기존 연구에서는 전문용어를 추출하는 단계에서 대부분 형태소 분석기를 이용한다. 그러나 전문용어의 특성으로 인해 형태소 분석 단계에서 오분석 되는 경우가 발생한다. 이러한 문제를 해결하기 위해서 본 논문에서는 음절 바이그램과 CRFs(Conditional Random Fields)를 이용하여 의학 전문 용어를 추출하는 방법을 제안한다. 네이버 지식인의 의사 답변 문서 2000개로부터 5-fold cross validation을 이용하여 실험하였다. 실험 결과 정확률은 평균 68.91%, 재현율은 평균 71.25%로 나타났으며 F-measure는 70.06%로 나타났다.

• Journal of IKEEE
• /
• v.26 no.1
• /
• pp.133-136
• /
• 2022

This paper is about motion gesture recognition system, and proposes the following improvement to the flaws of the current system: a motion gesture recognition system and such algorithm that uses the video image of the entire hand and reading its motion gesture to advance the accuracy of recognition. The motion gesture recognition system includes, an image capturing unit that captures and obtains the images of the area applicable for gesture reading, a motion extraction unit that extracts the motion area of the image, and a hand gesture recognition unit that read the motion gestures of the extracted area. The proposed application of the motion gesture algorithm achieves 20% improvement compared to that of the current system.

• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.78-80
• /
• 2019

개체명 인식 연구에서 우수한 성능을 보이고 있는 bi-LSTM-CRFs 모델은 처리 속도가 느린 단점이 있고, CNN-CRFs 모델은 한국어 말뭉치를 사용하여 제대로 분석되지 않았다. 본 논문에서는 한국어 개체명 인식 말뭉치를 이용한 CNN-CRFs 모델의 음절 단위 한국어 개체명 인식 방법을 제안한다. 실험 결과 bi-LSTM-CRFs 모델보다 CNN-CRFs 모델의 F1 score가 0.4% 높았고, 27.5% 빠른 처리 속도를 보였다.

• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.528-531
• /
• 2021

띄어쓰기에 대한 오류는 한국어 처리 전반에 영향을 주므로 자동 띄어쓰기는 필수적인 요소이다. 글쓴이의 대부분은 띄어쓰기 오류를 범하지 않으므로 글쓴이의 의도가 띄어쓰기 시스템에 반영되어야 한다. 그러나 대부분의 자동 띄어쓰기 시스템은 모든 띄어쓰기 정보를 제거하고 새로이 공백문자를 추가하는 방법으로 띄어쓰기를 수행한다. 이런 문제를 완화하기 위해서 본 논문에서는 기계학습에서 글쓴이의 의도가 반영된 자질을 추가하는 방법을 제안한다. 실험을 위해서 CRFs(Conditional Random Fields)를 사용하여 기존 시스템과 사용자의 의도를 반영한 띄어쓰기 시스템과의 성능을 비교하고 분석한다.

• Childhood Kidney Diseases
• /
• v.4 no.2
• /
• pp.144-153
• /
• 2000

Purpose : Effects to predict tile progression of chronic renal failure (CRF) in children, using mathematical models based on transformations of serum creatinine (Scr) concentration, have failed. Error may be introduced by age-related variations in creatinine production rate. Height (Ht) is a reliable reference for creatinine production in children. Thus, Scr, factored for Ht, could provide a more accurate predictive model. We examined this hypothesis. Methods : The progression of of was detected in 63 children who proceeded to end-stage renal disease. Derivatives of Scr, including 1/Scr, log Scr & Ht/Scr, were defined fir the period Scr was between 2 and 5 mg/dl. Regression equation were used to predict the time, in months, to Scr > 10 mg/dl. The prediction error (PE) was defined as the predicted time minus actual time for each Scr transformation. Result : The PE for Ht/Scr was lower than the PE for either 1/Scr or log Scr (median: -0.01, -2.0 & +10.6 mos respectively; P<0.0001). For children with congenital renal diseases, the PE for Ht/Scr was also lower than for the other two transformations (median: -1.2, -3.2 & +8.2 mos respectively; P<0.0001). However, the PEs for children with glomerular diseases was not as clearly different (median: +0.9, +0.5 & +9.9 respectively). In children < 13 yrs, PE for Ht/Scr was tile lowest, while in older children, 1/Scr provided the lowest PE but not significantly different from that for Ht/Scr. The logarithmic transformation tended to predict a slower progression of CRF than actually occurred. Conclusion : Scr, floored for Ht, appears to be a useful model to predict the rate of progression of CRF, particularly in the prepubertal child with congenital renal disease.

• Journal of the Korea Convergence Society
• /
• v.8 no.12
• /
• pp.55-62
• /
• 2017

The Named Entity Recognition system is a system that recognizes words or phrases with object names such as personal name (PS), place name (LC), and group name (OG) in the document as corresponding object names. Traditional approaches to named entity recognition include statistical-based models that learn models based on hand-crafted features. Recently, it has been proposed to construct the qualities expressing the sentence using models such as deep-learning based Recurrent Neural Networks (RNN) and long-short term memory (LSTM) to solve the problem of sequence labeling. In this research, to improve the performance of the Korean named entity recognition system, we used a hand-crafted feature, part-of-speech tagging information, and pre-built lexicon information to augment features for representing sentence. Experimental results show that the proposed method improves the performance of Korean named entity recognition system. The results of this study are presented through github for future collaborative research with researchers studying Korean Natural Language Processing (NLP) and named entity recognition system.

• The Journal of Engineering Geology
• /
• v.19 no.3
• /
• pp.295-306
• /
• 2009

The risk assessment of groundwater and soil in Sasang industrial complex in Busan Metropolitan City was carried out in order to estimate risks to human health and the environment. The carcinogenic risk (CR) of receptors to soil and air was not identified. However, the CRs for TCE and PCE were 6.7E-6 and 1.0E-5, respectively. Hazard quotient (HQ) and hazard index (HI) did not appear through air exposure pathways. Yet the HQ and HI of soil were 3.4E-5 and 5E-5, respectively, and lower than the critical value (1.0). On the contrary, HQ and HI with respect to groundwater were calculated as 0.7 (not hazardous) and 1.4 (hazardous). The constituent reduction factor (CRF) for TCE in the study area was determined as 2.5, and thus remediation work is demanded. As a result of sensitivity analysis for 18 exposure factors, eight exposure factors (life time of carcinogens, age, body weight, exposure duration, exposure frequency, dermal exposure frequency, water ingestion rate, and soil ingestion rate) varied with the variation of risk.

• Korean Journal of Health Education and Promotion
• /
• v.16 no.1
• /
• pp.149-166
• /
• 1999

Self-care and the performance of their own role might be important for the prevention of complications and improvement of quality of life in hemodialysis patients with chronic renal failure(CRF). To improve well-being and quality of life for the patients, the author estimated the level of self-care and associating factors through a questionnaire. The information was composed of the knowledge for hemodialysis and renal disease, the level of self-care, health belief, supports from the family, disease-related stresses, personal characteristics, medical history, relationships with medical personnel, etc. The data was gathered from 126 hemodialysis patients who were undergoing hemodialysis in one university hospital and five hospitals in Kyungsangnam-Do area from December 1997 to January 1998, and was analyzed by PC SAS program(version 6.12) with the level of significance($\alpha$=0.05). The mean age of subjects was 47.0$\pm$13.5years with no significant difference in gender distribution. The mean duration of hemodialysis was 39.0 months, and their frequencies of hemodialysis were more than three times per week(77.0%). Only 21.4% had the specific education on hemodialysis and CRF. In the level which was expressed as the score out of 100, the mean of knowledge was 90.7$\pm$9.1 and the mean of self-care was 73.9$\pm$12.7, that means, they only partially carried their knowledge into practice. They showed a significant correlation between knowledge and health belief($\gamma$=0.282); self-care and health belief($\gamma$=0.357), family supports and knowledge($\gamma$=0.221), self-care($\gamma$=0.402), health belief($\gamma$=0.431); and health belief and stress($\gamma$=-0.361). Age, religion, marrital status, education, and relationships with medical personnel showed positive correlations, and smoking showed negative correlation with self-care. In the multiple regression with the level of self-care as dependent variable, and each of the characeristics as independent variables, supports from the family($\beta$=6.615=0.158), the experience of disease specific education($\beta$=4.959), relationships with medical personnel($\beta$=6.615), current smoking($\beta$=-6.986), and current drinking ($\beta$=-7.095) were detected as significant factors. The value of R-square was 34%. In summary, to promote the level self-care and to improve the well beings and Quality of life for the hemodialysis patients, it would be emphasized that they terminate smoking and drinking, and it would be recommended that the education programs and supports from the family be strengthened. And, because there was a considerable difference between the level of knowledge and self-care, it would also be emphasized to propose the education programs which focused on execution. In addition to that, there is a need to improve relationships between the patients and medical personnel through positive changes in the attitudes of the medical personnel.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.9
• /
• pp.5585-5593
• /
• 2014

This study tested the hypothesis that the significance of the heterogeneous dispersion parameter in safety performance function (SPF) used to estimate the expected crashes is affected by the endogenous heterogeneous prior distributions, and analyzed the impacts of the mis-specified dispersion parameter on the evaluation results for traffic safety countermeasures. In particular, this study simulated the Poisson means based on the heterogeneous dispersion parameters and estimated the SPFs using both the negative binomial (NB) model and the heterogeneous negative binomial (HNB) model for analyzing the impacts of the model mis-specification on the mean and dispersion functions in SPF. In addition, this study analyzed the characteristics of errors in the crash reduction factors (CRFs) obtained when the two models are used to estimate the posterior means and variances, which are essentially estimated through the estimated hyper-parameters in the heterogeneous prior distributions. The simulation study results showed that a mis-estimation on the heterogeneous dispersion parameters through the NB model does not affect the coefficient of the mean functions, but the variances of the prior distribution are seriously mis-estimated when the NB model is used to develop SPFs without considering the heterogeneity in dispersion. Consequently, when the NB model is used erroneously to estimate the prior distributions with heterogeneous dispersion parameters, the mis-estimated posterior mean can produce large errors in CRFs up to 120%.

• Journal of the Korean Applied Science and Technology
• /
• v.37 no.2
• /
• pp.162-172
• /
• 2020

The purpose of this study was to investigate the improvement effect of 5% mung bean (phaseolus aureus L.) on the blood glucose and lipid metabolism function of streptozotocin (STZ, 45 mg/kg body weight)-induced diabetic rats. Seven-week-old male rats were divided into four groups (n=6), and fed experimental diets containing mung bean meal [basal diet+5% mung bean (BM), basal diet+STZ+5% mung bean (SM)], and control (BD), BS groups (basal diet+STZ). The results of this study, mung bean diet groups (BM, SM) in lipid composition evidenced the significantly reduction of serum total cholesterol, low density lipoprotein-cholesterol (LDL-cholesterol), atherosclerotic index (AI), cardiac risk factor (CRF), triglyceride (TG), phospholipid (PL), free cholesterol, cholesteryl ester, uric acid, blood glucose, non esterified fatty acid (NEFA), and elevation of high density lipoprotein-cholesterol (HDL-cholesterol). The serum albumin/globulin ratio (A/G ratio) was increased in mung bean supplementation diet than STZ-induced diabetic rats (p<0.05). Concentrations of sodium (Na) and chlorine (Cl) in sera were lower in the mung bean diet than diabetic group. Total calcium (T-Ca), phosphorus (Pi) and potassium (K) concentrations in sera were higher in the BM, SM and BD groups than BS group. In vivo experiments with Sprague-Dawley rats showed that ingestion of mung bean (phaseolus aureus L.) were effective in blood glucose and lipid composition.