• Journal of Chest Surgery
• /
• v.39 no.2 s.259
• /
• pp.106-110
• /
• 2006

Background: Bronchioloalveolar carcinoma (BAC) is an uncommon primary malignancy of the lung, and it accounts for $2{\~}14\%$ of all pulmonary malignancies. According to World Health Organization (WHO) categorisation, BAC is a subtype of adenocarcinoma. The current definition of BAC includes the following: malignant neoplasms of the lung that have no evidence of extrathoracic primary adenocarcinoma, an absence of a central bronchogenic source, a peripheral parenchymal location, and neoplastic cells growing along the alveolar septa. Previous reports had demonstrated a better prognosis following surgery for patients affected by BAC than those affected by other type of non-small cell lung cancer (NSCLC). We aim to analyse Asan Medical Center experiences of BAC. Material and Method: Between 1990 and 2002, 31 patients were received operations for BAC. We analyse retrosepectively sex, age, disease location, preoperative clinical stage, postoperative pathologic stage & complications, survival according to medical record. Result: There were 12 men and 19 women, the average age was 61.09$\pm$10.63 ($31{\~}79$) years. Tumor locations were 7 in RUL, 1 in RML, 4 in RLL, 8 in LUL, 11 in LLL. Operations were 28 lobectomies, 2 pneumonectomies. Postoperative pathologic stage were 12 T1N0M0, 15 T2N0M0, 1 T1N1M0, 1 T1N2M0, 1 T2N2M0, 1 T1N0M1. Mortality were 4 cases ($12.9\%$) and there were no early mortality. Cancer free death was 1 cases, other 3 were cancer related deaths. All of them were affected by distal metastasis and received chemotherapy and each metastatic locations were right rib, brain, and both lung field. The average follow up periods were 50.87$\pm$24.77 months. The overall 3, 5-year survival rate among all patients was $97.1\%,\;83.7\%$, stage I patients overall 2, 5year survival rate was $96.3\%$. The overall disease free 1, 2, 5-year survival rate among all patients was $100\%,\;90\%,\;76\%$ and 2, 5-year survival rate in cases of stage I was $96.4\%,\;90.6\%$. 7 cases ($22.58\%$) were chemotherapies, 1 case ($3.22\%$) was radiation therapy, and 2 cases ($6.45\%$) were chemoradiation therapies. Metastatic locations were 3 cases in lung, 1 case in bone, 1 cases in brain. Conclusion: BAC has a favourable survival and low recurrence rate compare with reported other NSCLC after operative resections.

• Sijohaknonchong
• /
• v.30
• /
• pp.263-300
• /
• 2009

Adoration for nature constitutes one of the primary subjects that literature has tackled since the origin of human history. Nature expressed through a poet's subjective imagination is the internalized and reorganized nature. This study examines the view on nature enacted in Ch'o-Jung's three-verse poems (sijo) in light of the traditional views on nature implicated in the ancient three-verse poems (koshijo), which is in line with the long-established Oriental view on nature. To dignitaris(sadaebu) in the Chosun Dynasty, nature appeared as the idealistic subject for moral culture ($shims{\breve{o}}ngsuyang$), which also becomes the literary space where the purity and justice of the world view of Neo-Confucianism(Sungrihak) is contained in the form of the three-verse poem, the lyrical poetic space where the "I" is united with nature by way of "enjoying of wind and moon"($umpungnongw{\breve{o}}i$) and "living in quiet retiremen"($yuyuchaj{\breve{o}}k$), and the object for the poetical perception of the surrounding world. Ch'o-Jung' s three-verse odes are found in Reed pipe ($Ch'oj{\breve{o}}k$), Sixty Five Pieces of Three-Verse Odes (Samhaengshi-$yukshipopy{\breve{o}}n$), Autumn Fragrance ($Hyangginam{\check{u}}n-ga{\check{u}}l$), and The Words of Zelko va Tree ($N{\check{u}}tinamu{\check{u}}i-mal$). This study analyzes 212 pieces of Ch'o-Jung' s three-verse poems chosen from theses books. In Ch'o-Jung's poems, the traditional view on nature expressed in the ancient three-verse poems is rendered in such a way that metaphysical understanding of nature is indirectly transmitted through the objective correlatives found nature. Nature is no longer the object of straightforward utterance, but transformed, displaced, and removed: that way, nature gets objectified to form a complicated and multi-layered structure. In conclusion, the view on nature manifested in Ch'o-Jung's three-verse poems is based on traditional metaphysics. Second, nature is the object of lyrical nostalgia and adoration. Third, nature is imbued with the fundamental affection for parents. Fourth, nature is associated with organic life. Fifth, the nature in Ch'o-Jung's poems reveals the beauty of stillness endorsed in Lao-tse's and Chung-tze's philosophy. And last, nature is the agent for self-realization and meditation.

• Korean Journal of Agricultural Science
• /
• v.17 no.1
• /
• pp.1-8
• /
• 1990

The paper board sludge(PBS) itself and compost sludge manure(CSM) mixed with sawdust, fowl droppings and urea to the PBS were treated to soybean plants to find the effects of growing characters, yield components, and nutritional compounds in the plant tissues. 1. Percentages of missing plants were 5-9% and 3.6-4.1% in the treatments of PBS and CSM, respectively. After that, the plants were restored to normal conditions. Anyway it is not desirable to use the paper board sludge and immature compost sludge manure in seeding time or to young seedlings. 2. Growth of the plant height was retarded in early growing phase, but it was normal in later stages. And the width and length of the largest leaf, numbers of main stem nodes and pods were not significant among the treatments. 3. The yield intends to increase through the treatments of 1,200, 1,600, and 2,000Kg PBS per 10a. In the CSM treatments with high rate of sawdust, fowl droppings and urea, the numbers of pods and grains were higher than the treatment of high content of PBS. 100-grain and one liter weights were opposite intention. It was suggested that the excess nitrogen amounts from the compost sludge manure than conventional fertilization affected to the yield components. 4. The contents of the main chemical compounds such as N, $P_2O_5$, $K_2O$, Ca, Mg were determined. The concentrations of nitrogen were higher in the treatment of PBS and CSM than none and control. 5. In conclusion, the nutritional effects of PBS were in evidence. To use the sludges, it should be fermented with adequate additives to improve the aeration, C/N ratio, activity of microbial, and other conditions. The sludges could be used to crops as a fertilizers effectively.

• Journal of Intelligence and Information Systems
• /
• v.20 no.3
• /
• pp.151-171
• /
• 2014

Firms today have sought management effectiveness and efficiency utilizing information technologies (IT). Numerous firms are outsourcing specific information systems functions to cope with their short of information resources or IT experts, or to reduce their capital cost. Recently, Software-as-a-Service (SaaS) as a new type of information system has become one of the powerful outsourcing alternatives. SaaS is software deployed as a hosted and accessed over the internet. It is regarded as the idea of on-demand, pay-per-use, and utility computing and is now being applied to support the core competencies of clients in areas ranging from the individual productivity area to the vertical industry and e-commerce area. In this study, therefore, we seek to quantify the value that SaaS has on business performance by examining the relationships among firm strategies, SaaS technology maturity, and business performance of SaaS providers. We begin by drawing from prior literature on SaaS, technology maturity and firm strategy. SaaS technology maturity is classified into three different phases such as application service providing (ASP), Web-native application, and Web-service application. Firm strategies are manipulated by the low-cost strategy and differentiation strategy. Finally, we considered customer acquisition as a business performance. In this sense, specific objectives of this study are as follows. First, we examine the relationships between customer acquisition performance and both low-cost strategy and differentiation strategy of SaaS providers. Secondly, we investigate the mediating and moderating effects of SaaS technology maturity on those relationships. For this purpose, study collects data from the SaaS providers, and their line of applications registered in the database in CNK (Commerce net Korea) in Korea using a questionnaire method by the professional research institution. The unit of analysis in this study is the SBUs (strategic business unit) in the software provider. A total of 199 SBUs is used for analyzing and testing our hypotheses. With regards to the measurement of firm strategy, we take three measurement items for differentiation strategy such as the application uniqueness (referring an application aims to differentiate within just one or a small number of target industry), supply channel diversification (regarding whether SaaS vendor had diversified supply chain) as well as the number of specialized expertise and take two items for low cost strategy like subscription fee and initial set-up fee. We employ a hierarchical regression analysis technique for testing moderation effects of SaaS technology maturity and follow the Baron and Kenny's procedure for determining if firm strategies affect customer acquisition through technology maturity. Empirical results revealed that, firstly, when differentiation strategy is applied to attain business performance like customer acquisition, the effects of the strategy is moderated by the technology maturity level of SaaS providers. In other words, securing higher level of SaaS technology maturity is essential for higher business performance. For instance, given that firms implement application uniqueness or a distribution channel diversification as a differentiation strategy, they can acquire more customers when their level of SaaS technology maturity is higher rather than lower. Secondly, results indicate that pursuing differentiation strategy or low cost strategy effectively works for SaaS providers' obtaining customer, which means that continuously differentiating their service from others or making their service fee (subscription fee or initial set-up fee) lower are helpful for their business success in terms of acquiring their customers. Lastly, results show that the level of SaaS technology maturity mediates the relationships between low cost strategy and customer acquisition. That is, based on our research design, customers usually perceive the real value of the low subscription fee or initial set-up fee only through the SaaS service provide by vender and, in turn, this will affect their decision making whether subscribe or not.

• Journal of Intelligence and Information Systems
• /
• v.23 no.3
• /
• pp.95-118
• /
• 2017

Recent centrally the downtown area, the transaction between the row housing and multiplex housing is activated and platform services such as Zigbang and Dabang are growing. The row housing and multiplex housing is a blind spot for real estate information. Because there is a social problem, due to the change in market size and information asymmetry due to changes in demand. Also, the 5 or 25 districts used by the Seoul Metropolitan Government or the Korean Appraisal Board(hereafter, KAB) were established within the administrative boundaries and used in existing real estate studies. This is not a district classification for real estate researches because it is zoned urban planning. Based on the existing study, this study found that the city needs to reset the Seoul Metropolitan Government's spatial structure in estimating future housing prices. So, This study attempted to classify the area without spatial heterogeneity by the reflected the property price characteristics of row housing and Multiplex housing. In other words, There has been a problem that an inefficient side has arisen due to the simple division by the existing administrative district. Therefore, this study aims to cluster Seoul as a new area for more efficient real estate analysis. This study was applied to the hedonic model based on the real transactions price data of row housing and multiplex housing. And the K-Means Clustering algorithm was used to cluster the spatial structure of Seoul. In this study, data onto real transactions price of the Seoul Row housing and Multiplex Housing from January 2014 to December 2016, and the official land value of 2016 was used and it provided by Ministry of Land, Infrastructure and Transport(hereafter, MOLIT). Data preprocessing was followed by the following processing procedures: Removal of underground transaction, Price standardization per area, Removal of Real transaction case(above 5 and below -5). In this study, we analyzed data from 132,707 cases to 126,759 data through data preprocessing. The data analysis tool used the R program. After data preprocessing, data model was constructed. Priority, the K-means Clustering was performed. In addition, a regression analysis was conducted using Hedonic model and it was conducted a cosine similarity analysis. Based on the constructed data model, we clustered on the basis of the longitude and latitude of Seoul and conducted comparative analysis of existing area. The results of this study indicated that the goodness of fit of the model was above 75 % and the variables used for the Hedonic model were significant. In other words, 5 or 25 districts that is the area of the existing administrative area are divided into 16 districts. So, this study derived a clustering method of row housing and multiplex housing in Seoul using K-Means Clustering algorithm and hedonic model by the reflected the property price characteristics. Moreover, they presented academic and practical implications and presented the limitations of this study and the direction of future research. Academic implication has clustered by reflecting the property price characteristics in order to improve the problems of the areas used in the Seoul Metropolitan Government, KAB, and Existing Real Estate Research. Another academic implications are that apartments were the main study of existing real estate research, and has proposed a method of classifying area in Seoul using public information(i.e., real-data of MOLIT) of government 3.0. Practical implication is that it can be used as a basic data for real estate related research on row housing and multiplex housing. Another practical implications are that is expected the activation of row housing and multiplex housing research and, that is expected to increase the accuracy of the model of the actual transaction. The future research direction of this study involves conducting various analyses to overcome the limitations of the threshold and indicates the need for deeper research.

• Cartoon and Animation Studies
• /
• s.39
• /
• pp.215-231
• /
• 2015

The Cultural Revolution in the history of Chinese animation hinder the development of the initial animation, and after a negative assessment instrument provided the cause is to become sluggish growth of the Chinese animation. So this time animation are things that are the subject of academic research studies or analysis has been depreciating almost uniformly without evaluation. However, of all the cultural and artistic creation it is developing in its own specific historical conditions and has the aesthetic results. This paper puts the primary purpose is to hold in consideration the aesthetic principles that led to cultural and artistic creativity and objective perspective the achievements the Chinese animation of the time period of the Cultural Revolution. Cultural Revolution is avoided to the previous period in accordance with the socialist ideology of Mao Ze-dong(毛澤東) sikindaneun highlight the culture of the proletariat and placed our goal to create a new class culture. Therefore, cultural and artistic creation of this period is often inconsistent with this part of our aesthetic principles generally accepted character has a non- elitist and anti properties. Best drama is a creative one hand as a model to implement the principles of aesthetics, art and culture Cultural Revolution period kkophimyeo reference for understanding the aesthetic principles that animated the Chinese Cultural Revolution period of orientation. This paper has San Tu Chu(三突出), Hong Guang Liang(紅光亮), and Gao Da Quan(高大全) at the time of the Cultural Revolution aesthetic principles are reflected in how the concrete work, the Cultural Revolution when the animation is how to accommodate these aesthetic principles and placed emphasis on comparative studies on best pole and correlation of the Cultural Revolution when the Chinese animation to ensure that adaptation in own way. First, after analyzing whether the aesthetic principles of focusing on the similarities of the best pole time of the Cultural Revolution and China, and how to implement animation in the works, these aesthetic principles according to the analysis of positive and negative influence on the creation of Chinese animation It was described as neutral. The detailed analysis and comparative study courses were trying to access in two significant aspects of the characters and scenes directing. In terms of character animation of the Cultural Revolution in China when a young boy or girl, emphasis should emphasize the health tinged with red lips and cheek blush to highlight the desired Gong Nong Bing(工農兵) shape as the main character and smooth texture and sophisticated highlights the glittering feeling to the touch, it was confirmed focused hayeoteum to implement the principle of 'Hong Guang Liang', highlighting the brilliant colors with a clean, bright colors. Highlighting a number of protagoniste compared to the antagonist in the animated scene of the Cultural Revolution a few times in terms of production and, among a number of protagoniste also emphasizes the outstanding hero figure, "yet three outstanding heroes heroic figures also emphasize the leading figures among the the director of the extrusion step-by-step approach "('San Tu Chu')was used. In addition, the hero figure is generally high and low angle by directing a large and perfect aesthetic appearance was to faithfully implement the principle of 'high-charged'('Gao Da Quan').

• Journal of Intelligence and Information Systems
• /
• v.25 no.4
• /
• pp.141-154
• /
• 2019

Rapid growth of internet technology and social media is progressing. Data mining technology has evolved to enable unstructured document representations in a variety of applications. Sentiment analysis is an important technology that can distinguish poor or high-quality content through text data of products, and it has proliferated during text mining. Sentiment analysis mainly analyzes people's opinions in text data by assigning predefined data categories as positive and negative. This has been studied in various directions in terms of accuracy from simple rule-based to dictionary-based approaches using predefined labels. In fact, sentiment analysis is one of the most active researches in natural language processing and is widely studied in text mining. When real online reviews aren't available for others, it's not only easy to openly collect information, but it also affects your business. In marketing, real-world information from customers is gathered on websites, not surveys. Depending on whether the website's posts are positive or negative, the customer response is reflected in the sales and tries to identify the information. However, many reviews on a website are not always good, and difficult to identify. The earlier studies in this research area used the reviews data of the Amazon.com shopping mal, but the research data used in the recent studies uses the data for stock market trends, blogs, news articles, weather forecasts, IMDB, and facebook etc. However, the lack of accuracy is recognized because sentiment calculations are changed according to the subject, paragraph, sentiment lexicon direction, and sentence strength. This study aims to classify the polarity analysis of sentiment analysis into positive and negative categories and increase the prediction accuracy of the polarity analysis using the pretrained IMDB review data set. First, the text classification algorithm related to sentiment analysis adopts the popular machine learning algorithms such as NB (naive bayes), SVM (support vector machines), XGboost, RF (random forests), and Gradient Boost as comparative models. Second, deep learning has demonstrated discriminative features that can extract complex features of data. Representative algorithms are CNN (convolution neural networks), RNN (recurrent neural networks), LSTM (long-short term memory). CNN can be used similarly to BoW when processing a sentence in vector format, but does not consider sequential data attributes. RNN can handle well in order because it takes into account the time information of the data, but there is a long-term dependency on memory. To solve the problem of long-term dependence, LSTM is used. For the comparison, CNN and LSTM were chosen as simple deep learning models. In addition to classical machine learning algorithms, CNN, LSTM, and the integrated models were analyzed. Although there are many parameters for the algorithms, we examined the relationship between numerical value and precision to find the optimal combination. And, we tried to figure out how the models work well for sentiment analysis and how these models work. This study proposes integrated CNN and LSTM algorithms to extract the positive and negative features of text analysis. The reasons for mixing these two algorithms are as follows. CNN can extract features for the classification automatically by applying convolution layer and massively parallel processing. LSTM is not capable of highly parallel processing. Like faucets, the LSTM has input, output, and forget gates that can be moved and controlled at a desired time. These gates have the advantage of placing memory blocks on hidden nodes. The memory block of the LSTM may not store all the data, but it can solve the CNN's long-term dependency problem. Furthermore, when LSTM is used in CNN's pooling layer, it has an end-to-end structure, so that spatial and temporal features can be designed simultaneously. In combination with CNN-LSTM, 90.33% accuracy was measured. This is slower than CNN, but faster than LSTM. The presented model was more accurate than other models. In addition, each word embedding layer can be improved when training the kernel step by step. CNN-LSTM can improve the weakness of each model, and there is an advantage of improving the learning by layer using the end-to-end structure of LSTM. Based on these reasons, this study tries to enhance the classification accuracy of movie reviews using the integrated CNN-LSTM model.

• Korean Journal of Veterinary Research
• /
• v.35 no.2
• /
• pp.337-345
• /
• 1995

To evaluate the pathogenicity and immunogenicity of Eimeria tenella to the chicken treated with dexamethasone(DEX) and testosterone propionate (TES), we administered 0.1ml/chicken of dexamethasone and 40mg/chicken of testosterone propionate at 1-, 2-, and 7-days old, respectively. We also immunized with ND oil-emulsion vaccine at 2 weeks old. After that, we immunized and challenged with 100 and $1{\times}10^5$ oocysts/chicken of E tenella at 2 and 4 weeks old, respectively. And then we investigated the HI titers for ND virus, survival rate, body weight gain, lesion score and the weight of the bursa of Fabricius and thymus. The titers for ND virus in the groups treated with TES were higher than those in the groups treated with DEX and CON during 3 to 6 weeks. After challenge, the survival rate of testosterone propionate treated-challenged(TES-CHA) and TES-immunized and challenged(TES-V&C) groups were 61.5 and 83.3% and those of the other groups were all 100%. At 1 week after challenge, the lesion scores of TES-CHA group(4.0) was the highest of all experimental groups. Those of DEX and controlchallenged( CON-CHA) groups were 2.8, and those of all V&C groups were 2.4. During 1 and 2 weeks after immunization, the body weight gains of TES groups were severe low(61.6-82.2g and 189.6-260.4g). During 1 and 2 weeks after challenge, the body weight gains of all CHA groups were lower than those of not challenged groups. But, those of all V AC groups were not different from those of not immunized groups. At 4- and 6-weeks old, the weight of the bursa of Fabricius and thymus in the chicken of all TES groups were lower than those of all control (CON) and DEX groups. Therefore, testosterone propionate acted as immunosuppressive drug. Also, it was thought that the chicken affected a little humoral immunity to E tenella.

• Journal of Nutrition and Health
• /
• v.48 no.6
• /
• pp.476-487
• /
• 2015

Purpose: This study was conducted to examine the concentration of nutrients in transitional breast milk from Korean lactating mothers and to evaluate daily intakes of their infants based on the Dietary Reference Intakes for Koreans 2010 (KDRIs 2010). Methods: Breast milk samples were collected at 5~15 days postpartum from 100 healthy lactating Korean mothers. Macro- and micro-nutrients, and immunoglobulin (Igs) concentrations in breast milk were analyzed. Results: The mean energy, protein, fat, and carbohydrate concentrations in breast milk were $59.99{\pm}8.01kcal/dL$, $1.47{\pm}0.27g/dL$, $2.88{\pm}0.89g/dL$, and $6.72{\pm}0.22g/dL$. The mean linoleic acid (LA), a-linolenic acid (ALA), arachidonic acid (AA), and docosahexaenoic acid (DHA) concentrations were $181.44{\pm}96.41mg/dL$, $28.15{\pm}8.89mg/dL$, $5.67{\pm}1.86mg/dL$, and $5.74{\pm}2.57mg/dL$. The mean vitamin A, vitamin D, vitamin E, vitamin $B_1$, vitamin $B_2$, vitamin $B_{12}$, and folate concentrations were $2.75{\pm}1.75{\mu}g/dL$, $2.31{\pm}1.12ng/dL$, $0.74{\pm}1.54mg/dL$, $3.02{\pm}1.84mg/dL$, $7.51{\pm}20.96{\mu}g/dL$, $61.78{\pm}26.78{\mu}g/dL$, $63.71{\pm}27.19ng/dL$, and $0.52{\pm}0.26{\mu}g/dL$. The mean concentrations of calcium, iron, potassium, sodium, zinc, and copper were $20.71{\pm}3.34mg/dL$, $0.59{\pm}0.86mg/dL$, $66.71{\pm}10.35mg/dL$, $27.72{\pm}10.16mg/dL$, $0.44{\pm}0.41mg/dL$, and $70.48{\pm}30.41{\mu}g/dL$. The mean IgA and total IgE concentrations were $61.85{\pm}31.97mg/dL$ and $235.00{\pm}93.00IU/dL$. The estimated daily intakes of infants for protein, vitamin D, vitamin E, vitamin $B_2$, vitamin $B_{12}$, iron, potassium, sodium, zinc, and copper were sufficient compared to KDRIs 2010 adjusted by transitory milk intakes. The estimated infants' intakes of energy, fat, carbohydrate, vitamin A, vitamin C, vitamin $B_1$, folate, and calcium did not meet KDRIs 2010 adjusted by transitory milk intakes. Conclusion: In general most estimated nutrient intakes of Korean breast-fed infants in transitory breast milk were sufficient, however some nutrient intakes were not sufficient based on KDRIs 2010. These results warrant conduct of future studies for investigation of important dietary factors associated with nutrients in breast milk to improve the quality of breast milk, which may contribute to understanding nutrition in early life and promoting growth and development of breast-fed infants.

• Journal of Intelligence and Information Systems
• /
• v.26 no.4
• /
• pp.127-148
• /
• 2020

The data center is a physical environment facility for accommodating computer systems and related components, and is an essential foundation technology for next-generation core industries such as big data, smart factories, wearables, and smart homes. In particular, with the growth of cloud computing, the proportional expansion of the data center infrastructure is inevitable. Monitoring the health of these data center facilities is a way to maintain and manage the system and prevent failure. If a failure occurs in some elements of the facility, it may affect not only the relevant equipment but also other connected equipment, and may cause enormous damage. In particular, IT facilities are irregular due to interdependence and it is difficult to know the cause. In the previous study predicting failure in data center, failure was predicted by looking at a single server as a single state without assuming that the devices were mixed. Therefore, in this study, data center failures were classified into failures occurring inside the server (Outage A) and failures occurring outside the server (Outage B), and focused on analyzing complex failures occurring within the server. Server external failures include power, cooling, user errors, etc. Since such failures can be prevented in the early stages of data center facility construction, various solutions are being developed. On the other hand, the cause of the failure occurring in the server is difficult to determine, and adequate prevention has not yet been achieved. In particular, this is the reason why server failures do not occur singularly, cause other server failures, or receive something that causes failures from other servers. In other words, while the existing studies assumed that it was a single server that did not affect the servers and analyzed the failure, in this study, the failure occurred on the assumption that it had an effect between servers. In order to define the complex failure situation in the data center, failure history data for each equipment existing in the data center was used. There are four major failures considered in this study: Network Node Down, Server Down, Windows Activation Services Down, and Database Management System Service Down. The failures that occur for each device are sorted in chronological order, and when a failure occurs in a specific equipment, if a failure occurs in a specific equipment within 5 minutes from the time of occurrence, it is defined that the failure occurs simultaneously. After configuring the sequence for the devices that have failed at the same time, 5 devices that frequently occur simultaneously within the configured sequence were selected, and the case where the selected devices failed at the same time was confirmed through visualization. Since the server resource information collected for failure analysis is in units of time series and has flow, we used Long Short-term Memory (LSTM), a deep learning algorithm that can predict the next state through the previous state. In addition, unlike a single server, the Hierarchical Attention Network deep learning model structure was used in consideration of the fact that the level of multiple failures for each server is different. This algorithm is a method of increasing the prediction accuracy by giving weight to the server as the impact on the failure increases. The study began with defining the type of failure and selecting the analysis target. In the first experiment, the same collected data was assumed as a single server state and a multiple server state, and compared and analyzed. The second experiment improved the prediction accuracy in the case of a complex server by optimizing each server threshold. In the first experiment, which assumed each of a single server and multiple servers, in the case of a single server, it was predicted that three of the five servers did not have a failure even though the actual failure occurred. However, assuming multiple servers, all five servers were predicted to have failed. As a result of the experiment, the hypothesis that there is an effect between servers is proven. As a result of this study, it was confirmed that the prediction performance was superior when the multiple servers were assumed than when the single server was assumed. In particular, applying the Hierarchical Attention Network algorithm, assuming that the effects of each server will be different, played a role in improving the analysis effect. In addition, by applying a different threshold for each server, the prediction accuracy could be improved. This study showed that failures that are difficult to determine the cause can be predicted through historical data, and a model that can predict failures occurring in servers in data centers is presented. It is expected that the occurrence of disability can be prevented in advance using the results of this study.