• Journal of Intelligence and Information Systems
• /
• v.22 no.3
• /
• pp.45-69
• /
• 2016

Recently, sentiment analysis using open Internet data is actively performed for various purposes. As online Internet communication channels become popular, companies try to capture public sentiment of them from online open information sources. This research is conducted for the purpose of analyzing pulbic sentiment of Korean Top-10 companies using a multi-categorical sentiment lexicon. Whereas existing researches related to public sentiment measurement based on big data approach classify sentiment into dimensions, this research classifies public sentiment into multiple categories. Dimensional sentiment structure has been commonly applied in sentiment analysis of various applications, because it is academically proven, and has a clear advantage of capturing degree of sentiment and interrelation of each dimension. However, the dimensional structure is not effective when measuring public sentiment because human sentiment is too complex to be divided into few dimensions. In addition, special training is needed for ordinary people to express their feeling into dimensional structure. People do not divide their sentiment into dimensions, nor do they need psychological training when they feel. People would not express their feeling in the way of dimensional structure like positive/negative or active/passive; rather they express theirs in the way of categorical sentiment like sadness, rage, happiness and so on. That is, categorial approach of sentiment analysis is more natural than dimensional approach. Accordingly, this research suggests multi-categorical sentiment structure as an alternative way to measure social sentiment from the point of the public. Multi-categorical sentiment structure classifies sentiments following the way that ordinary people do although there are possibility to contain some subjectiveness. In this research, nine categories: 'Sadness', 'Anger', 'Happiness', 'Disgust', 'Surprise', 'Fear', 'Interest', 'Boredom' and 'Pain' are used as multi-categorical sentiment structure. To capture public sentiment of Korean Top-10 companies, Internet news data of the companies are collected over the past 25 months from a representative Korean portal site. Based on the sentiment words extracted from previous researches, we have created a sentiment lexicon, and analyzed the frequency of the words coming up within the news data. The frequency of each sentiment category was calculated as a ratio out of the total sentiment words to make ranks of distributions. Sentiment comparison among top-4 companies, which are 'Samsung', 'Hyundai', 'SK', and 'LG', were separately visualized. As a next step, the research tested hypothesis to prove the usefulness of the multi-categorical sentiment lexicon. It tested how effective categorial sentiment can be used as relative comparison index in cross sectional and time series analysis. To test the effectiveness of the sentiment lexicon as cross sectional comparison index, pair-wise t-test and Duncan test were conducted. Two pairs of companies, 'Samsung' and 'Hanjin', 'SK' and 'Hanjin' were chosen to compare whether each categorical sentiment is significantly different in pair-wise t-test. Since category 'Sadness' has the largest vocabularies, it is chosen to figure out whether the subgroups of the companies are significantly different in Duncan test. It is proved that five sentiment categories of Samsung and Hanjin and four sentiment categories of SK and Hanjin are different significantly. In category 'Sadness', it has been figured out that there were six subgroups that are significantly different. To test the effectiveness of the sentiment lexicon as time series comparison index, 'nut rage' incident of Hanjin is selected as an example case. Term frequency of sentiment words of the month when the incident happened and term frequency of the one month before the event are compared. Sentiment categories was redivided into positive/negative sentiment, and it is tried to figure out whether the event actually has some negative impact on public sentiment of the company. The difference in each category was visualized, moreover the variation of word list of sentiment 'Rage' was shown to be more concrete. As a result, there was huge before-and-after difference of sentiment that ordinary people feel to the company. Both hypotheses have turned out to be statistically significant, and therefore sentiment analysis in business area using multi-categorical sentiment lexicons has persuasive power. This research implies that categorical sentiment analysis can be used as an alternative method to supplement dimensional sentiment analysis when figuring out public sentiment in business environment.

• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.71-88
• /
• 2017

Language models were originally developed for speech recognition and language processing. Using a set of example sentences, a language model predicts the next word or character based on sequential input data. N-gram models have been widely used but this model cannot model the correlation between the input units efficiently since it is a probabilistic model which are based on the frequency of each unit in the training set. Recently, as the deep learning algorithm has been developed, a recurrent neural network (RNN) model and a long short-term memory (LSTM) model have been widely used for the neural language model (Ahn, 2016; Kim et al., 2016; Lee et al., 2016). These models can reflect dependency between the objects that are entered sequentially into the model (Gers and Schmidhuber, 2001; Mikolov et al., 2010; Sundermeyer et al., 2012). In order to learning the neural language model, texts need to be decomposed into words or morphemes. Since, however, a training set of sentences includes a huge number of words or morphemes in general, the size of dictionary is very large and so it increases model complexity. In addition, word-level or morpheme-level models are able to generate vocabularies only which are contained in the training set. Furthermore, with highly morphological languages such as Turkish, Hungarian, Russian, Finnish or Korean, morpheme analyzers have more chance to cause errors in decomposition process (Lankinen et al., 2016). Therefore, this paper proposes a phoneme-level language model for Korean language based on LSTM models. A phoneme such as a vowel or a consonant is the smallest unit that comprises Korean texts. We construct the language model using three or four LSTM layers. Each model was trained using Stochastic Gradient Algorithm and more advanced optimization algorithms such as Adagrad, RMSprop, Adadelta, Adam, Adamax, and Nadam. Simulation study was done with Old Testament texts using a deep learning package Keras based the Theano. After pre-processing the texts, the dataset included 74 of unique characters including vowels, consonants, and punctuation marks. Then we constructed an input vector with 20 consecutive characters and an output with a following 21st character. Finally, total 1,023,411 sets of input-output vectors were included in the dataset and we divided them into training, validation, testsets with proportion 70:15:15. All the simulation were conducted on a system equipped with an Intel Xeon CPU (16 cores) and a NVIDIA GeForce GTX 1080 GPU. We compared the loss function evaluated for the validation set, the perplexity evaluated for the test set, and the time to be taken for training each model. As a result, all the optimization algorithms but the stochastic gradient algorithm showed similar validation loss and perplexity, which are clearly superior to those of the stochastic gradient algorithm. The stochastic gradient algorithm took the longest time to be trained for both 3- and 4-LSTM models. On average, the 4-LSTM layer model took 69% longer training time than the 3-LSTM layer model. However, the validation loss and perplexity were not improved significantly or became even worse for specific conditions. On the other hand, when comparing the automatically generated sentences, the 4-LSTM layer model tended to generate the sentences which are closer to the natural language than the 3-LSTM model. Although there were slight differences in the completeness of the generated sentences between the models, the sentence generation performance was quite satisfactory in any simulation conditions: they generated only legitimate Korean letters and the use of postposition and the conjugation of verbs were almost perfect in the sense of grammar. The results of this study are expected to be widely used for the processing of Korean language in the field of language processing and speech recognition, which are the basis of artificial intelligence systems.

• Journal of Intelligence and Information Systems
• /
• v.23 no.1
• /
• pp.47-67
• /
• 2017

Steel plate faults is one of important factors to affect the quality and price of the steel plates. So far many steelmakers generally have used visual inspection method that could be based on an inspector's intuition or experience. Specifically, the inspector checks the steel plate faults by looking the surface of the steel plates. However, the accuracy of this method is critically low that it can cause errors above 30% in judgment. Therefore, accurate steel plate faults diagnosis system has been continuously required in the industry. In order to meet the needs, this study proposed a new steel plate faults diagnosis system using Simultaneous MTS (S-MTS), which is an advanced Mahalanobis Taguchi System (MTS) algorithm, to classify various surface defects of the steel plates. MTS has generally been used to solve binary classification problems in various fields, but MTS was not used for multiclass classification due to its low accuracy. The reason is that only one mahalanobis space is established in the MTS. In contrast, S-MTS is suitable for multi-class classification. That is, S-MTS establishes individual mahalanobis space for each class. 'Simultaneous' implies comparing mahalanobis distances at the same time. The proposed steel plate faults diagnosis system was developed in four main stages. In the first stage, after various reference groups and related variables are defined, data of the steel plate faults is collected and used to establish the individual mahalanobis space per the reference groups and construct the full measurement scale. In the second stage, the mahalanobis distances of test groups is calculated based on the established mahalanobis spaces of the reference groups. Then, appropriateness of the spaces is verified by examining the separability of the mahalanobis diatances. In the third stage, orthogonal arrays and Signal-to-Noise (SN) ratio of dynamic type are applied for variable optimization. Also, Overall SN ratio gain is derived from the SN ratio and SN ratio gain. If the derived overall SN ratio gain is negative, it means that the variable should be removed. However, the variable with the positive gain may be considered as worth keeping. Finally, in the fourth stage, the measurement scale that is composed of selected useful variables is reconstructed. Next, an experimental test should be implemented to verify the ability of multi-class classification and thus the accuracy of the classification is acquired. If the accuracy is acceptable, this diagnosis system can be used for future applications. Also, this study compared the accuracy of the proposed steel plate faults diagnosis system with that of other popular classification algorithms including Decision Tree, Multi Perception Neural Network (MLPNN), Logistic Regression (LR), Support Vector Machine (SVM), Tree Bagger Random Forest, Grid Search (GS), Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). The steel plates faults dataset used in the study is taken from the University of California at Irvine (UCI) machine learning repository. As a result, the proposed steel plate faults diagnosis system based on S-MTS shows 90.79% of classification accuracy. The accuracy of the proposed diagnosis system is 6-27% higher than MLPNN, LR, GS, GA and PSO. Based on the fact that the accuracy of commercial systems is only about 75-80%, it means that the proposed system has enough classification performance to be applied in the industry. In addition, the proposed system can reduce the number of measurement sensors that are installed in the fields because of variable optimization process. These results show that the proposed system not only can have a good ability on the steel plate faults diagnosis but also reduce operation and maintenance cost. For our future work, it will be applied in the fields to validate actual effectiveness of the proposed system and plan to improve the accuracy based on the results.

• Journal of Intelligence and Information Systems
• /
• v.21 no.4
• /
• pp.17-35
• /
• 2015

Demand forecasting is the activity of estimating the quantity of a product or service that consumers will purchase for a certain period of time. Developing precise forecasting models are considered important since corporates can make strategic decisions on new markets based on future demand estimated by the models. Many studies have developed market growth curve models, such as Bass, Logistic, Gompertz models, which estimate future demand when a market is in its early stage. Among the models, Bass model, which explains the demand from two types of adopters, innovators and imitators, has been widely used in forecasting. Such models require sufficient demand observations to ensure qualified results. In the beginning of a new market, however, observations are not sufficient for the models to precisely estimate the market's future demand. For this reason, as an alternative, demands guessed from those of most adjacent markets are often used as references in such cases. Reference markets can be those whose products are developed with the same categorical technologies. A market's demand may be expected to have the similar pattern with that of a reference market in case the adoption pattern of a product in the market is determined mainly by the technology related to the product. However, such processes may not always ensure pleasing results because the similarity between markets depends on intuition and/or experience. There are two major drawbacks that human experts cannot effectively handle in this approach. One is the abundance of candidate reference markets to consider, and the other is the difficulty in calculating the similarity between markets. First, there can be too many markets to consider in selecting reference markets. Mostly, markets in the same category in an industrial hierarchy can be reference markets because they are usually based on the similar technologies. However, markets can be classified into different categories even if they are based on the same generic technologies. Therefore, markets in other categories also need to be considered as potential candidates. Next, even domain experts cannot consistently calculate the similarity between markets with their own qualitative standards. The inconsistency implies missing adjacent reference markets, which may lead to the imprecise estimation of future demand. Even though there are no missing reference markets, the new market's parameters can be hardly estimated from the reference markets without quantitative standards. For this reason, this study proposes a case-based expert system that helps experts overcome the drawbacks in discovering referential markets. First, this study proposes the use of Euclidean distance measure to calculate the similarity between markets. Based on their similarities, markets are grouped into clusters. Then, missing markets with the characteristics of the cluster are searched for. Potential candidate reference markets are extracted and recommended to users. After the iteration of these steps, definite reference markets are determined according to the user's selection among those candidates. Then, finally, the new market's parameters are estimated from the reference markets. For this procedure, two techniques are used in the model. One is clustering data mining technique, and the other content-based filtering of recommender systems. The proposed system implemented with those techniques can determine the most adjacent markets based on whether a user accepts candidate markets. Experiments were conducted to validate the usefulness of the system with five ICT experts involved. In the experiments, the experts were given the list of 16 ICT markets whose parameters to be estimated. For each of the markets, the experts estimated its parameters of growth curve models with intuition at first, and then with the system. The comparison of the experiments results show that the estimated parameters are closer when they use the system in comparison with the results when they guessed them without the system.

• Tuberculosis and Respiratory Diseases
• /
• v.57 no.5
• /
• pp.449-460
• /
• 2004

Background : PS-341 is a novel, highly selective and potent proteasome inhibitor, which showed cytotoxicity against some tumor cells. Its anti-tumor activity has been suggested to be associated with modulation of the expression of apoptosis-associated proteins, such as p53, $p21^{WAF/CIP1}$, $p27^{KIP1}$, NF-${\kappa}B$, Bax and Bcl-2. c-Jun N-terminal kinase (JNK) and glycogen synthase kinase-$3{\beta}$ (GSK-$3{\beta}$) are important modulators of apoptosis. However, their role in PS-341-induced apoptosis is unclear. This study was undertaken to elucidate the role of JNK and GSK-$3{\beta}$ in the PS-341-induced apoptosis in lung cancer cells. Method : NCI-H157 and A549 cells were used in the experiments. The cell viability was assayed using the MTT assay and apoptosis was evaluated by proteolysis of PARP. The JNK activity was measured by an in vitro immuno complex kinase assay and by phosphorylation of endogenous c-Jun. The protein expression was evaluated by Western blot analysis. Dominant negative JNK1 (DN-JNK1) and GSK-$3{\beta}$ were overexpressed using plasmid and adenovirus vectors, respectively. Result : PS-341 reduced the cell viability via apoptosis, activated JNK and increased the c-Jun expression. Blocking of the JNK activation by overexpression of DN-JNK1, or pretreatment with SP600125, suppressed the apoptosis induced by PS-341. The activation of caspase 3 was mediated by JNK activation. Blocking of the caspase 3 activation suppressed PS-341-induced apoptosis. PS-341 activated the phosphatidylinositol 3-kinase (PI3K)/Akt pathway, but its blockade enhanced the PS-341-induced cell death via apoptosis. GSK-$3{\beta}$ was inactivated by PS-341 via the PI3K/Akt pathway. Overexpression of constitutively active GSK-$3{\beta}$ enhanced PS-341-induced apoptosis; in contrast, this was suppressed by dominant negative GSK-$3{\beta}$ (DN-GSK-$3{\beta}$). Inactivation of GSK-$3{\beta}$ by pretreatment with lithium chloride or the overexpression of DN-GSK-$3{\beta}$ suppressed both the JNK activation and c-Jun up-regulation induced by PS-341. Conclusion : The JNK/caspase pathway is involved in PS-341-induced apoptosis, which is negatively regulated by the PI3K/Akt-mediated inactivation of GSK-$3{\beta}$ in lung cancer cells.

• Journal of Korean Society of Forest Science
• /
• v.30 no.1
• /
• pp.8-18
• /
• 1976

This study was carried out to exploit and utilize the exotic American Sycamore(Platanus occidentalis) grown in Korea as a veneer species for plywood manufacture. At present most parts of veneer Legs used in Korea were depended entirely upon the gonus Shorea woods(lauan logs) imported from Southeast Asia region. To decrease manufacturing cost and save imported lauan veneer logs, the effects on properties affecting to the six types of plywood made from proper veneer constructions composed by domestic American Sycamore and imported lauan veneers were compared and discussed. The study has important meanings for the promotion of plywood manufacture by domestic materials. Important items dealt with this study were dry and wet shear strength, moisture contents, specific gravties, and bending strength. By the results and discussions it may be summarized as follows: 1) In dry shear strength platanus(sycamore) core lauan plywood was shown most excellent strong results, and next orders were all lauan plywood, platanus faced lauan plywood, lauan core platanus plywood, lauan faced platanus plywood, and all platanus plywood. There was no difference between platanus core lauan plywood and all lauan plywood, but the differences between those plywoods and the other types of plywood were recognized. 2) In wet shear strength platanus core lauan plywood was shown excellent result the same as dry strength. The difference between platanus core lauan plywood and the other types of plywood was shown, but among the other types of plywood except platanus core lauan plywood were not recognized. 3) The differences among moisture contents according to the veneer construction were not recognized. 4) The plywood constructed by two or more sheets of lauan veneer was shown lower specific gravities than the plywood constructed by two or more sheets of platanus veneer. It is believed that this tendency due to the original specific gravities of veneer before construction. 5) The differences among specific gravities of lauan core platanus plywood, all platanus plywood and lauan faced platanus plywood were not recognized, and like this analyzed result among platanus core lauan plywood, all lauan plywood and platanus faced lauan plywood were not recognized. Accordingly it is believed that the differences are not shown among the plywood constructed by two or more veneers of same species. 6) In bending strength platanus core lauan plywood was shown most excellent values. Next orders were all lauan plywood, platanus faced lauan plywood and the other types of plywood. The differences among the plywood constructed by two or veneers of lauan were shown, but not shown among the plywood constructed by tow or more veneers of platanus.

• The Korean Journal of Nuclear Medicine
• /
• v.39 no.6
• /
• pp.481-488
• /
• 2005

Purpose: Functional imaging of dopamine transporter (DAT) defines integrity of the dopaminergic system, and DAT is the target site of drugs of abuse such as cocaine and methamphetamine. Functional imaging the DAT may be a sensitive and selective indicator of neurotoxic change by the drug. The aim of the present study is to evaluate the clinical implications of qualitative/quantitative analyses of dopamine transporter imaging in methamphetamine abusers. Materials and Methods: Six detoxified methamphetamine abusers (abuser group) and 4 volunteers (control group) were enrolled in this study. Brain MRI was performed in all of abuser group. Abuser group underwent psychiatric and depression assessment using brief psychiatric rating scale (BPRS) and Hamilton depression rating scale (HAMD), respectively. All of the subjects underwent I-123 IPT SPECT (IPT SPECT). IPT SPECT image was analysed with visual qualitative method and quantitative method using basal ganglia dopamine transporter (DAT) specific/non-specific binding ratio (SBR). Comparison of DAT SBR between abuser and control groups was performed. We also performed correlation tests between psychiatric and depression assessment results and DAT SBR in abuser group. Results: All of abuser group showed normal MRI finding, but had residual psychiatric and depressive symptoms, and psychiatric and depressive symptom scores were exactly correlated (r=1.0, p=0.005) each other. Five of them showed abnormal finding on qualitative visual I-123 IPT SPECT Abuser group had lower basal ganglia DAT SBR than that of control ($2.38{\pm}0.20\;vs\;3.04{\pm}0.27$, p=0.000). Psychiatric and depressive symptoms were negatively well correlated with basal ganglia DAT SBR (r=-0.908, p=0.012, r=-0.924, p=0.009). Conclusion: These results suggest that dopamine transporter imaging using I-123 IPT SPECT may be used to evaluate dopaminergic system of the basal ganglia and the clinical status in methamphetamine abusers.

• Journal of Food Hygiene and Safety
• /
• v.26 no.3
• /
• pp.260-265
• /
• 2011

Plantago asiatica L. (P. asiatica) has been used as one of the popular folk medicines in Asia for human health care practices. Various activities of P. asiatica have been reported, such as anti-oxidant, anti-glycation, anti-inflammatory and hepatoprotective activity. Therefore, the potential of P. asiatica to reduce oxidative stress has been studied in several ways for over 20 years, especially at liver and kidney. However no investigation has been reported revealing its protective effect on prostate. Method: Treatment of P. asiatica leaf ethanolic extract (PLE) (1 g/kg body weight (b.w.), 2 g/kg b.w., or 4 g/kg b.w.) were given separately to animals for pretreatment once per day for 7 days, and on the seventh day ferric nitrilotriacetate (Fe-NTA; 0.24 mmol Fe/kg b.w.), which is known as an oxidative stress-inducer at prostate, was administrated by i.p to negative control group. At the end of the study period, dissection was carried out for detecting the prostate protective effect of PLE. Result: Fe-NTA-treated animals produced reactive oxygen species (ROS) resulting in depletion of antioxidant biomaker, such as glutathione (GSH), glutathione reductase (GR), and glutathione s-transferase (GST) and increase of lipid peroxidation in prostate. However, PLE pretreatment resulted in an increase in the GSH, GST and GR levels concentration dependent manner and in an significant decrease in the levels of lipid peroxidation. Conclusion: Our data suggest that PLE may be effective in protecting oxidative stress-induced damage of prostate, and PLE may be an chemopreventive agent against Fe-NTA-mediated prostate oxidative damage.

• Korean Journal of Food Science and Technology
• /
• v.8 no.1
• /
• pp.19-32
• /
• 1976

The laboratory Mejus as well as home-made Meju and improved Meju received from Korea were ripened in the brine for up to 8 months and the changes is the chemical composition during the process were determined and the differences between the types of Meju were compared. On the basis of the amino acid pattern, the changes in the protein quality of soybean during the process was evaluated. No significant changes in the general chemical composition of Meju were noticed during the ripening for 8 months. However, the nitrogen solubility of Meju increased for $13{\sim}29%$ to $66{\sim}78%$ during 8 month ripening of the Meju-brine mixture. The concentration of free amino-N to the total-N increased from $4{\sim}7%$ in Meju to $29{\sim}35%$ in the 8month ripened mixture. The concentration of amino-N to the total-N increased from $1{\sim}4%$ in Meju to $5{\sim}14%$ in the 8month ripened mixture and the changes varied with the type of Meju used. Remarkable changes in the amino acid pattern of soybean were occured during the ripening process. The concentration of methionine decreased to the half of original Meju during the first month of ripening. Arginine and histidine were destroyed rapidly by the ripening longer than 1 month. A considerable amount of ornithine was synthesized during the ripening. The amino acid pattern of Meju did change drastically during the ripening longer than 3 months and the changes varied with the type of Meju. The retention of the nutrients in soybean during 8 month ripening of the laboratory 3 month Meju in the brine was 49% for carbohydrates, 107% for crude fat, 93% for crude protein and 74% for the total amino acid. Histidine, arginine and methionine and 74% for the total amino acid. Histidine, arginine and methionine were the most damaged during the process, retaining only 25%, 27% and 49% of the contents in raw soybean, respectively, whereas lysine retained 79%. By the separation of the 8 month ripened mixture, approximately 60% of crude protein, all of crude fat and 80% of carbohydrates in the mixture were retained in soypaste. Soypaste contained higher concentrations of amino acids per 16gN compared to soysauce, except for lysine. The most limiting amino acid of the protein was the S-containing amino acids in all cases studied, whereas the second limiting amino acid varied from valine in soybean to threonine in most of Mejus and the brine mixtures, lysine in most of soypastes and tryptophan in some of soysauces. According to the protein quality evaluation made by the reference of the FAO provisional pattern of amino acid, the chemical score of raw soybean was 82, which was reduced to 77 by cooking and further reduced to $71{\sim}74$ by Meju fermentation. At the eighth month of ripening the chemical score of the Meju-brine mixtures were reduced to $51{\sim}66. After the separation, the chemical score of soypaste ranged from 60 to 71, whereas that of soysauce varied from 45 to 57. Generally, the products made from improved Meju recorded the highest score, whereas those made from homemade Meju showed the poorest protein quality. The essential amino acid index(EAAI) of the samples was similar to the chemical score, but it appeared to fit the overall changes in the amino acid pattern during the process better than the chemical score.

• Childhood Kidney Diseases
• /
• v.3 no.2
• /
• pp.130-144
• /
• 1999

Purpose : Long-term use of Cyclosporine(CsA) reduce renal blood flow by afferent arteriolar vasoconstriction and lead to chronic pathologic changes of CsA nephrotoxicity - 1) interstitial nephritis(IN); tubular atrophy (TA) and/or interstitial fibrosis(IF),2) arteriolopathy(AP). The Object of this study is to estimate the incidence of chronic pathologic CsA nephrotoxicity by duration of treatment and type of renal disease, relationship between histologic and clinical nephrotoxicity, and optimal duration of CsA therapy. Methods : 102 children with steroid resistant or dependent nephrotic syndrome confirmed by renal biopsy and treated with CsA from 1986 to 1997 were enrolled in this study(58 MCNS, 10 FSGS, 10 MGN, 15 $Henoch-Sch\"{o}nlein$ purpura nephritis with nephrotic syndrome (HSPN) and 9 IgA nephropathy with nephrotic syndrome(IgAN)). CsA was administered for 1yr, 1.5yr, 2yr in 24, 12, 22 MCNS patients and 2, 2, 6 FSGS patients respectively, 1yr, 2yr in MGN and 1yr in HSPN and IgAN. Sequential biopsies were done in all 102 patients after CsA treatment for evaluation of pathologic nephrotoxicity. Results : Complete remission rate was 92.2% (100% in MCNS and MGN, 80% in FSGS, 86.6% in HSPN and 55.5% in IgAN). Incidence of relapse during 6months after CsA treatment was significantly decreased compaed with relapsing spisodes during 6months before CsA treatment in MCNS(P<0.0001) and FSGS(P<0.0001). According to pathologic changes, 71 patients(69.6%) showed no pathological change, 24 patients(23.5%) showed IN and 7 patients(6.8%) showed AP. IN was 16.6%, 33.3%, 27.2% in 1, 1.5, 2 year of CsA treatment group in MCNS. AP was 0%, 16.6%, 9% in 1, 1.5, 2 year of CsA treatment group in MCNS. 14 out of 58 MCNS(24.1%) showed IN and 4 out of 58 MCNS(6.8%) showed AP. Incidence of pathologic change was significantly lower in CsA therapy of <1yr than >1yr(P=0.03). There were no significant difference of incidence of pathologic change in original renal disease, age and sex. Conclusion : Duration of CsA treatment was significant risk factor for nephrotoxicity and optimal duration seemed to be 1 year. Pathologic change due to nephrotoxicity did not correlate with deterioration of renal function and only detectable by renal biopsy.