• Journal of Intelligence and Information Systems
• /
• v.27 no.4
• /
• pp.1-22
• /
• 2021

Aspect Based Sentiment Analysis (ABSA), which analyzes sentiment based on aspects that appear in the text, is drawing attention because it can be used in various business industries. ABSA is a study that analyzes sentiment by aspects for multiple aspects that a text has. It is being studied in various forms depending on the purpose, such as analyzing all targets or just aspects and sentiments. Here, the aspect refers to the property of a target, and the target refers to the text that causes the sentiment. For example, for restaurant reviews, you could set the aspect into food taste, food price, quality of service, mood of the restaurant, etc. Also, if there is a review that says, "The pasta was delicious, but the salad was not," the words "steak" and "salad," which are directly mentioned in the sentence, become the "target." So far, in ABSA, most studies have analyzed sentiment only based on aspects or targets. However, even with the same aspects or targets, sentiment analysis may be inaccurate. Instances would be when aspects or sentiment are divided or when sentiment exists without a target. For example, sentences like, "Pizza and the salad were good, but the steak was disappointing." Although the aspect of this sentence is limited to "food," conflicting sentiments coexist. In addition, in the case of sentences such as "Shrimp was delicious, but the price was extravagant," although the target here is "shrimp," there are opposite sentiments coexisting that are dependent on the aspect. Finally, in sentences like "The food arrived too late and is cold now." there is no target (NULL), but it transmits a negative sentiment toward the aspect "service." Like this, failure to consider both aspects and targets - when sentiment or aspect is divided or when sentiment exists without a target - creates a dual dependency problem. To address this problem, this research analyzes sentiment by considering both aspects and targets (Target-Aspect-Sentiment Detection, hereby TASD). This study detected the limitations of existing research in the field of TASD: local contexts are not fully captured, and the number of epochs and batch size dramatically lowers the F1-score. The current model excels in spotting overall context and relations between each word. However, it struggles with phrases in the local context and is relatively slow when learning. Therefore, this study tries to improve the model's performance. To achieve the objective of this research, we additionally used auxiliary loss in aspect-sentiment classification by constructing CNN(Convolutional Neural Network) layers parallel to existing models. If existing models have analyzed aspect-sentiment through BERT encoding, Pooler, and Linear layers, this research added CNN layer-adaptive average pooling to existing models, and learning was progressed by adding additional loss values for aspect-sentiment to existing loss. In other words, when learning, the auxiliary loss, computed through CNN layers, allowed the local context to be captured more fitted. After learning, the model is designed to do aspect-sentiment analysis through the existing method. To evaluate the performance of this model, two datasets, SemEval-2015 task 12 and SemEval-2016 task 5, were used and the f1-score increased compared to the existing models. When the batch was 8 and epoch was 5, the difference was largest between the F1-score of existing models and this study with 29 and 45, respectively. Even when batch and epoch were adjusted, the F1-scores were higher than the existing models. It can be said that even when the batch and epoch numbers were small, they can be learned effectively compared to the existing models. Therefore, it can be useful in situations where resources are limited. Through this study, aspect-based sentiments can be more accurately analyzed. Through various uses in business, such as development or establishing marketing strategies, both consumers and sellers will be able to make efficient decisions. In addition, it is believed that the model can be fully learned and utilized by small businesses, those that do not have much data, given that they use a pre-training model and recorded a relatively high F1-score even with limited resources.

• Korean Journal of Psychosomatic Medicine
• /
• v.27 no.1
• /
• pp.50-59
• /
• 2019

Objectives : Many of the patients with type 2 diabetes are associated with sleep problems, and the rate of insomnia is known to be higher in the general population. The aims of this study were to know the frequency and clnical characteristics of insomnia, and related variables to insomnia in patients diagnosed with type 2 diabetes. Methods : For 99 patients from 18 to 80 years of age (65 males and 34 females) with type 2 diabetes, interviews were performed. Total sleep time and sleep latency was evaluated. Insomnia was evaluated using the Korean Version of the Insomnia Severity Index (ISI-K). Severity of depressive symptoms were evaluted using the Korean version of the Hamilton Depression Scale (K-HDRM). According to the cutoff score of 15.5 on the ISI-K, subjects were divided into the group of type 2 diabetics with insomnia (N=34) and those without insomnia (N=65) at first, and then statistically analyzed. Results : TInsomnia could be found in 34.34% of type 2 diabetics. Type 2 diabetics with insomnia had significantly more single or divorced (respectively 11.8%, p<0.05), higher total scores of the K-HDRS ($11.76{\pm}5.52$, p<0.001), shorter total sleep time ($5.35{\pm}2.00hours$, p<0.001), and longer sleep latency ($50.29{\pm}33.80minutes$, p<0.001). The all item scores of the ISI-K in type 2 diabetics with insomnia were significantly higher than those in type 2 diabetics without insomnia, that is, total ($18.38{\pm}2.69$), A1 (Initial insomnia) ($2.97{\pm}0.76$), A2 (Middle insomnia) ($3.06{\pm}0.69$), A3 (Terminal insomnia) ($2.76{\pm}0.61$), B (Satisfaction) ($3.18{\pm}0.72$), C (Interference) ($2.09{\pm}0.97$), D (Noticeability) ($2.12{\pm}1.09$) and E (Distress) ($2.21{\pm}0.81$) (respectively p<0.001). Variables associated with insomnia in type 2 diabetics were as following. Age had significant negative correlation with A3 items of the ISI-K (${\beta}=-0.241$, p<0.05). Total scores of the K-HDRS had significant positive correlation, while total sleep time had significant negative correlation with all items of the ISI-K (respectively p<0.05). Sleep latency had significant positive correlation with total,, A1, B and E item scores of the ISI-K (respectively p<0.05). Conclusions : Insomnia was found in about 1/3 of type 2 diabetics. According to the presence of insomnia, clinical characteristics including sleep quality as well as quantity seemed to be different. Because depression seemed to be correlated with insomnia, clinicians should pay attention to early detection and intervention of depression among type 2 diabetics.

• Journal of Animal Science and Technology
• /
• v.46 no.3
• /
• pp.335-346
• /
• 2004

Factors associated with milk urea nitrogen on milk yield and milk composition were investigated in this study regarding feeding management and Physiological status of lactating cows. The data for 3 years between 1999 and 2002 for this study were collected from 129,645 cows by Korean Agricultural Cooperatives Federation. The objectives of this study were to describe the relationships between milk urea concentrations and seasonal factors, cow factors and production of milk, milk fat, protein and somatic cell score(SCS). Milk urea was highest in summer and it also showed a nonlinear association with milk yield. Milk yield was higher at milk urea concentrations of 21 ${\sim}$ 24mg/dl, however it decreased at higher level of milk urea concentrations than 24mg/dl. Milk urea was higher with increased parities of cows and in particular, at 3 to 4 parities. There was a negative association between milk urea and SCS in milk. SCS in milk was lowest at milk urea concentrations of 2l${\sim}$24mg/dl. Milk fat and milk protein were greatly affected by days in milk(DIM), year of birth, season and milk urea concentrations, respectively. While milk urea increased in summer, milk fat and protein were decreased. Milk protein decreased according to longer days in milk. With regard to the influences of parity, milk protein in overall was negatively correlated to milk urea in all lactations, however the extent of decrease of milk protein at high concentrations of milk urea was bigger at third lactation. The balanced supply of energy and protein to the animals might have greatly affected the urea concentrations and protein content of milk.

• Food Science of Animal Resources
• /
• v.27 no.2
• /
• pp.142-149
• /
• 2007

This study was carried out to investigate the effects of washing time and the addition of sodium chloride(2%) on the quality characteristics of surimi made with chicken breast. The control(C) prepared from Alaska pollack was washed 2 times without sodium chloride. For the test treatments, ground chicken breast was washed 2 times only(T1), washed 2 times followed by the addition of sodium chloride(T2), washed 3 times(T3), washed 4 times with added sodium chloride(T4), washed 6 times(T5), and washed 6 times with added sodium chloride(T6) to produce chicken breast surimi. The $L^*,\;a^*$, W, shear force, and juiciness values were significantly higher, but the hardness, cohesiveness, gumminess, chewiness, aroma, flavor, and overall acceptability of Tl were significantly lower than those of the control(p<0.05). The $L^*$ value decreased as the washing time increased, and the $a^*$ and W values were significantly higher, however the hardness, breaking force, gel strength, shear force, and overall sensory scores of the samples washed 2 times were lower than those washed 4 and 6 times (p<0.05). The $L^*,\;b^*$, and shear force values were significantly lower but the $a^*$, W, hardness, cohesiveness, gumminess, chewiness, folding test results and overall sensory scores were significantly higher due to the addition of sodium chloride (p<0.05). The correlation coefficients(r>0.6) for the overall sensory scores and other items were positive for the folding test, cohesiveness, gumminess, chewiness, and flavor, but negative for shear force(p<0.05). Overall, T4 had the highest qualities and economic value among all treatments.

• Sleep Medicine and Psychophysiology
• /
• v.4 no.1
• /
• pp.107-119
• /
• 1997

Objectives: The psychophysiologic response pattern between healthy subjects and patients with generalized anxiety disorder, and the relationship among anxiety rating scales and those patterns in patients were examined. Methods: Twenty-three patients with generalized anxiety disorder(AD) and 23 healthy subjects were evaluated by Hamilton Rating Scale for Anxiety(HRSA) and State-Trait Anxiety Inventory before baseline stressful tasks. Subjective Units of Distress were evaluated just before baseline period, immediately after stressful tasks, at the end of the entire procedure, and psychophysiologic measures, i.e., skin temperature(ST), electromyographic activity(EMG), heart rate(HR), electrodermal response(EDR) during baseline & rest and during two psychologically stressful tasks (mental arithmetic, TM; talk about a stressful event, TT) were also evaluated. Results: 1) AD group showed significantly higher EMG level during rest after stressful tasks and higher HR level during all period except TM compared to control group. 2) AD group showed lower change in the startle response(SR) of ST, in the SR & the recovery response(RR) of EMG during TM, and in the RR of EDR immediately after TM than control group. AD group showed that the RR of EDR was significantly lower than the SR during stressful tasks. 3) We found that there was significantly negative correlation between state anxiety and the RR of EDR after TT in AD group. We also found that there were significantly positive correlations between HRSA score and the SRs of EDR during stressful tasks, and between state anxiety and the SR of EDR during TT. Conclusion: Our results suggest that patients with generalized anxiety disorder show higher autonomic arousal than healthy subjects and decreased physiologic flexibility or reduced autonomic flexibility.

• Journal of Nutrition and Health
• /
• v.52 no.3
• /
• pp.285-296
• /
• 2019

Purpose: This study was conducted to identify the factors affecting vegetable preferences of children based on the social cognitive theory to reduce imbalances in vegetable consumption. Methods: The survey investigated 177 elementary school students in Yangsan, Gyeongsangnam-do, in June of 2018. The subjects consisted of 44 fifth graders (27.7%) and 128 (72.3%) sixth graders. Results: Among personal factors of the Social Cognitive Theory, positive outcome expectation and self-efficacy of the vegetable preference group were significantly higher than those of the non-preference group. Negative barrier scores of the non-preference group were significantly higher than those of the preference group, and the biggest barrier was that vegetables were tasteless. Among behavioral factors, the nutritional knowledge of vegetables was high, but the degree of practice was low. Practice score of the vegetable preference group was significantly higher than that of the non-preference group. Among environmental factors, the vegetable preference group was more likely to accept advice from people around them than the non-preference group and the most influential people were doctors and parents. In the vegetable intake environment, children in the vegetable preference group had high accessibility to vegetables. Correlation analysis and regression analysis of the social cognitive factors and vegetable preferences revealed all factors except nutritional knowledge showed significant correlation with vegetable preference. And surrounding people (p < 0.01), practice (p < 0.01), and self-efficacy (p < 0.05) had positive effects on vegetable preference. Conclusion: These results suggest that providing the health benefits from eating vegetables and educating children for improving their self-confidence are necessary for increasing the preference for vegetables and their intake by children.

• Journal of Life Science
• /
• v.30 no.3
• /
• pp.250-259
• /
• 2020

Diabetes is a well-known risk factor for dementia and cognitive impairment. Diabetic polyneuropathy (DPN) is the most prevalent microvascular complication in type 2 diabetes mellitus (T2DM) patients. The purpose of this study was to evaluate the relation between diabetic peripheral polyneuropathy and cognitive factors in T2DM patients. Retrospective chart review of type 2 diabetic patients with results of a nerve conduction study (NCS) and a neurocognitive study. A total of 19 patients were included. DPN was defined using data from a nerve conduction study: a score of less than 24 in the Korean version of the Mini-Mental State Examination (K-MMSE) was considered as an indicator of cognitive impairment (CI). The mean age of the 19 patients was 71.6±5.0 years. The mean duration of diabetes was 8.4±9.1 years, and the mean HbA1c level was 8.1±1.8%. DPN was present in 7 of the 19 patients. Based on the K-MMSE score, CI was diagnosed in eight patients. The mean K-MMSE scores and the prevalence of CI was not different between the groups with and without DPN. There was no significant difference in DPN prevalence between the groups with and without CI. Education was significantly correlated with cognitive factors. Only the digit span-forward among the cognitive factors showed a significant negative correlation with nerve conduction velocity. In conclusion, the longer education period was associated with higher cognitive function and no significant correlation was observed between diabetic peripheral neuropathy and cognitive dysfunction in type 2 diabetic patients. Further prospective research is needed in the future.

• Tuberculosis and Respiratory Diseases
• /
• v.44 no.3
• /
• pp.493-504
• /
• 1997

Background : Non-small cell lung cancer is one of the most frequent cause of death due to cancer in men, and its incidence among women is rapidly increasing. Although there has been a recent surge of interest in combined modality therapy for stageIII non-small cell lung cancer(NSCLC), the optimal treatment is still not well established. Thoracic irradiation has long been the gold standard for locally advanced unresectable NSCLC. However, although conventional radiotherapy(XRT) can palliate symptom and improve local control of disease, it has at most only a modest effect on survival. Recently, cisplatin(cis-diamminedichloroplatinum) has been reported to enhance the cell-killing effect of radiation For patients with unresectable NSCLC, cisplatin-based concurrent chemoradiotherapy(CCRT) had the advantage of therapeutic response over XRT alone and therapeutic side effect more commonly occurred in CCRT group in EORTC(European Organization for Research and Treatment of Cancer) and other trials. Objectives : We compared therapeutic response, compliance, and side effects between CCRT and XRT in patients with advanced NSCLC. Patients and Method : Thirty patients with biopsy-proven inoperable NSCLC were randomized to one of two treatment arms. Arm A consisted of XRT, radiotherapy for 4~6 weeks(1.8 Gy given 20~33 times, in five fractions a week), and arm B consisted of CCRT, radiotherapy for 2 weeks(3 Gy given to times, in five fractions a week), followed by 3 week rest period and then radiotherapy 2 more weeks(2.5 Gy given 10 times, in five fractions a week), combined with 6mg cisplatin per square meter, given daily before radiotherapy. We evaluate therapeutic response, compliance, change of performance status, side effects, and radiation pneumonitis by using the author's made scoring system. Results : There was no significant difference in therapeutic response and compliance. But there was a significantly lower laboratory complication and radiation pneumonitis in CCRT group (p < 0.05). There's significant negative correlation between stage and therapeutic response score in both groups(R=0.353, p < 0.05) In both groups, patients with squamous cell carcinoma had a tendency to higher therapeutic response score than those with adenocarcinoma. Conclusion : There was no difference between CCRT and XRT in respect to therapeutic response and compliance. But CCRT had a advantage of decreased side effects.

• Journal of Intelligence and Information Systems
• /
• v.25 no.4
• /
• pp.105-122
• /
• 2019

Dimensionality reduction is one of the methods to handle big data in text mining. For dimensionality reduction, we should consider the density of data, which has a significant influence on the performance of sentence classification. It requires lots of computations for data of higher dimensions. Eventually, it can cause lots of computational cost and overfitting in the model. Thus, the dimension reduction process is necessary to improve the performance of the model. Diverse methods have been proposed from only lessening the noise of data like misspelling or informal text to including semantic and syntactic information. On top of it, the expression and selection of the text features have impacts on the performance of the classifier for sentence classification, which is one of the fields of Natural Language Processing. The common goal of dimension reduction is to find latent space that is representative of raw data from observation space. Existing methods utilize various algorithms for dimensionality reduction, such as feature extraction and feature selection. In addition to these algorithms, word embeddings, learning low-dimensional vector space representations of words, that can capture semantic and syntactic information from data are also utilized. For improving performance, recent studies have suggested methods that the word dictionary is modified according to the positive and negative score of pre-defined words. The basic idea of this study is that similar words have similar vector representations. Once the feature selection algorithm selects the words that are not important, we thought the words that are similar to the selected words also have no impacts on sentence classification. This study proposes two ways to achieve more accurate classification that conduct selective word elimination under specific regulations and construct word embedding based on Word2Vec embedding. To select words having low importance from the text, we use information gain algorithm to measure the importance and cosine similarity to search for similar words. First, we eliminate words that have comparatively low information gain values from the raw text and form word embedding. Second, we select words additionally that are similar to the words that have a low level of information gain values and make word embedding. In the end, these filtered text and word embedding apply to the deep learning models; Convolutional Neural Network and Attention-Based Bidirectional LSTM. This study uses customer reviews on Kindle in Amazon.com, IMDB, and Yelp as datasets, and classify each data using the deep learning models. The reviews got more than five helpful votes, and the ratio of helpful votes was over 70% classified as helpful reviews. Also, Yelp only shows the number of helpful votes. We extracted 100,000 reviews which got more than five helpful votes using a random sampling method among 750,000 reviews. The minimal preprocessing was executed to each dataset, such as removing numbers and special characters from text data. To evaluate the proposed methods, we compared the performances of Word2Vec and GloVe word embeddings, which used all the words. We showed that one of the proposed methods is better than the embeddings with all the words. By removing unimportant words, we can get better performance. However, if we removed too many words, it showed that the performance was lowered. For future research, it is required to consider diverse ways of preprocessing and the in-depth analysis for the co-occurrence of words to measure similarity values among words. Also, we only applied the proposed method with Word2Vec. Other embedding methods such as GloVe, fastText, ELMo can be applied with the proposed methods, and it is possible to identify the possible combinations between word embedding methods and elimination methods.

• Korean Journal of Health Education and Promotion
• /
• v.17 no.1
• /
• pp.95-113
• /
• 2000

The purpose of this study is to investigate the state of the dental health of adults, used self-reporting qestionnaire as objects of 923 residents living in nine districts at random among the sites of eleven town located in a County. The level of knowledge on dental health of adults in a County was 65.6 points out of 100 points, comparatively low. Accodingly, it is necessary for entire adults to have a dental health-related education and get high standard of knowledge. Especially, educational approach should be performed for groups of 40-49 years old, 50-59 years old, above 60 years old, male, no educational background, having only elementary and middle school education, not having any jobs, engaging in agriculture, doing business on their own and so on. When planning the contents of health education, one actually has to include the habit of amalgam, the factor in influencing on dental health as well as show an example such as how to brush teeth, checking point of proper brushing, how to grip toothbrush. The attitude score related to dental heath was 71.2 points out of 100 points, relatively low. Consequently, the change of attitude related the dental health among entire adults is necessary, particularily, the strategical approach is essential to change dental health connected to attitude positively for groups of male, having high school education background, office workers and the civil service. Besides, among dental health related symtoms, more that 30-40％ of objects showed negative attitude toward as the following cases; in case that plaque or food debris are attached to the teeth (40.8％), upper and lower teeth do not fit together(40.3％), you cannot sleep well because of toothache(31.0％), more than one tooth fall out(31.0％), you have loosing teeth(30.6％), the approach should be conducted to form attitude that dental care is necessry. The state of dental health through dental health related symtoms was 33.3 points out of 100 points, which was fairly satisfactory. However, dental treatment for the state of dental health should be executed in case of comparison of the dental health state according to general characteristics, the group who are above 60 years old, have elementary school education background, engage in aggriculture who are not good in dental health state as opposed to other groups. Furthermore, there should be dental care needs according to dental health related symtoms, particularly, more than 60-70％ of objects have experienced symtoms that plaque or food debris attached to the teeth, tartar is on the teeth so dental treatment should be peformed for a large number of adults. In addition, for the people who have indications that there was a cavity, more than one tooth loss, chilled teeth, toothache when chewing, loose teeth, upper and lower teeth do not fit together, you cannnt sleep well due to the toothach, etc, there should be care through dental treatment. The actual conditions of the hygine of the mouth was relatively good and the difference of the actual state of dental health care in terms of general characteristics showed in only gender; female was more careful in dental health. Comparing the state of oral health synthetically, when they have symtoms, only 34.8％ of objects go to a dentist, 60.7％ are using passive or negative care such as gargling, tolerating or ignoring. There was many symtoms to care through dental therapy such as plaque or food debris get in between the teeth, tartar on the teeth, teeth are very cold, more than one tooth fallout, loose teeth, there is wrong amalgam, and so on, among symtoms to care passively or negatively. Therefore the education for proper treatment program should be performed. As a result of dental health-related knowlege, attitude, health state, verification of correlation between the actual condition of care, the higher the dental heath-related knowldege becomes, the more positive the dental health-related attitude is, and the state of dental health, that is, the standard of the symtom of dental health diminishes. the care for dental health executed through more active method and the more positive dental health-related attitude is, the more active means they performed. Consequently, the high level of dental health-related knowledge should be necessary, the more positive the dental health-related attitude was, the more active method they adopted, therefore, the program is needed to form attitude related to the dental health positively. As the knowledge on dental health is increasing, the attitude is also positive, after all, the plan to increase the standard of knowledge on dental health should be contrived through education program related to dental health.