• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.123-138
• /
• 2017

Since the stock market is driven by the expectation of traders, studies have been conducted to predict stock price movements through analysis of various sources of text data. In order to predict stock price movements, research has been conducted not only on the relationship between text data and fluctuations in stock prices, but also on the trading stocks based on news articles and social media responses. Studies that predict the movements of stock prices have also applied classification algorithms with constructing term-document matrix in the same way as other text mining approaches. Because the document contains a lot of words, it is better to select words that contribute more for building a term-document matrix. Based on the frequency of words, words that show too little frequency or importance are removed. It also selects words according to their contribution by measuring the degree to which a word contributes to correctly classifying a document. The basic idea of constructing a term-document matrix was to collect all the documents to be analyzed and to select and use the words that have an influence on the classification. In this study, we analyze the documents for each individual item and select the words that are irrelevant for all categories as neutral words. We extract the words around the selected neutral word and use it to generate the term-document matrix. The neutral word itself starts with the idea that the stock movement is less related to the existence of the neutral words, and that the surrounding words of the neutral word are more likely to affect the stock price movements. And apply it to the algorithm that classifies the stock price fluctuations with the generated term-document matrix. In this study, we firstly removed stop words and selected neutral words for each stock. And we used a method to exclude words that are included in news articles for other stocks among the selected words. Through the online news portal, we collected four months of news articles on the top 10 market cap stocks. We split the news articles into 3 month news data as training data and apply the remaining one month news articles to the model to predict the stock price movements of the next day. We used SVM, Boosting and Random Forest for building models and predicting the movements of stock prices. The stock market opened for four months (2016/02/01 ~ 2016/05/31) for a total of 80 days, using the initial 60 days as a training set and the remaining 20 days as a test set. The proposed word - based algorithm in this study showed better classification performance than the word selection method based on sparsity. This study predicted stock price volatility by collecting and analyzing news articles of the top 10 stocks in market cap. We used the term - document matrix based classification model to estimate the stock price fluctuations and compared the performance of the existing sparse - based word extraction method and the suggested method of removing words from the term - document matrix. The suggested method differs from the word extraction method in that it uses not only the news articles for the corresponding stock but also other news items to determine the words to extract. In other words, it removed not only the words that appeared in all the increase and decrease but also the words that appeared common in the news for other stocks. When the prediction accuracy was compared, the suggested method showed higher accuracy. The limitation of this study is that the stock price prediction was set up to classify the rise and fall, and the experiment was conducted only for the top ten stocks. The 10 stocks used in the experiment do not represent the entire stock market. In addition, it is difficult to show the investment performance because stock price fluctuation and profit rate may be different. Therefore, it is necessary to study the research using more stocks and the yield prediction through trading simulation.

• Journal of Intelligence and Information Systems
• /
• v.21 no.4
• /
• pp.111-131
• /
• 2015

There are only a handful number of research conducted on pattern analysis of corporate distress as compared with research for bankruptcy prediction. The few that exists mainly focus on audited firms because financial data collection is easier for these firms. But in reality, corporate financial distress is a far more common and critical phenomenon for non-audited firms which are mainly comprised of small and medium sized firms. The purpose of this paper is to classify non-audited firms under distress according to their financial ratio using data mining; Self-Organizing Map (SOM). SOM is a type of artificial neural network that is trained using unsupervised learning to produce a lower dimensional discretized representation of the input space of the training samples, called a map. SOM is different from other artificial neural networks as it applies competitive learning as opposed to error-correction learning such as backpropagation with gradient descent, and in the sense that it uses a neighborhood function to preserve the topological properties of the input space. It is one of the popular and successful clustering algorithm. In this study, we classify types of financial distress firms, specially, non-audited firms. In the empirical test, we collect 10 financial ratios of 100 non-audited firms under distress in 2004 for the previous two years (2002 and 2003). Using these financial ratios and the SOM algorithm, five distinct patterns were distinguished. In pattern 1, financial distress was very serious in almost all financial ratios. 12% of the firms are included in these patterns. In pattern 2, financial distress was weak in almost financial ratios. 14% of the firms are included in pattern 2. In pattern 3, growth ratio was the worst among all patterns. It is speculated that the firms of this pattern may be under distress due to severe competition in their industries. Approximately 30% of the firms fell into this group. In pattern 4, the growth ratio was higher than any other pattern but the cash ratio and profitability ratio were not at the level of the growth ratio. It is concluded that the firms of this pattern were under distress in pursuit of expanding their business. About 25% of the firms were in this pattern. Last, pattern 5 encompassed very solvent firms. Perhaps firms of this pattern were distressed due to a bad short-term strategic decision or due to problems with the enterpriser of the firms. Approximately 18% of the firms were under this pattern. This study has the academic and empirical contribution. In the perspectives of the academic contribution, non-audited companies that tend to be easily bankrupt and have the unstructured or easily manipulated financial data are classified by the data mining technology (Self-Organizing Map) rather than big sized audited firms that have the well prepared and reliable financial data. In the perspectives of the empirical one, even though the financial data of the non-audited firms are conducted to analyze, it is useful for find out the first order symptom of financial distress, which makes us to forecast the prediction of bankruptcy of the firms and to manage the early warning and alert signal. These are the academic and empirical contribution of this study. The limitation of this research is to analyze only 100 corporates due to the difficulty of collecting the financial data of the non-audited firms, which make us to be hard to proceed to the analysis by the category or size difference. Also, non-financial qualitative data is crucial for the analysis of bankruptcy. Thus, the non-financial qualitative factor is taken into account for the next study. This study sheds some light on the non-audited small and medium sized firms' distress prediction in the future.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.22 no.1
• /
• pp.67-75
• /
• 2018

Purpose The purpose of this study is to investigate and analyze the external dose rates of $^{18}F-FDG$ Whole Body PET/CT patients by distance, and to identify the main factors that contribute to the reduction of radiation dose by checking the cumulative doses of nuclear medicine technologist(NMT). Materials and Methods After completion of the $^{18}F-FDG$ Whole Body PET/CT scan($75.4{\pm}3.3min$), the external dose rates of 106 patients were measured at a distance of 0, 10, 30, 50, and 100 cm from the chest. Gender, age, BMI(Body Mass Index), fasting time, diabetes mellitus, radiopharmaceutical injection information, creatine value were collected to analyze individual factors that could affect external dose rates from a patient's perspective. From the perspective of NMT, personal pocket dosimeters were worn on the chest to record accumulated dose of NMT who performed the injection task($T_1$, $T_2$ and $T_3$) and scan task($T_4$, $T_5$ and $T_6$). In addition, patient contact time with NMT was measured and analyzed. Results External dose rates from the patient for each distance were calculated as $246.9{\pm}37.6$, $129.9{\pm}16.7$, $61.2{\pm}9.1$, $34.4{\pm}5.9$, and $13.1{\pm}2.4{\mu}Sv/hr$ respectively. On the patient's aspect, there was a significant difference in the proximity of gender, BMI, Injection dose and creatine value, but the difference decreased as the distance increased. In case of dialysis patient, external dose rates for each distance were exceptionally higher than other patients. On the NMT aspect, the doses received from patients were 0.70, 1.09, $0.55{\mu}Sv/person$ for performing the injection task($T_1$, $T_2$, and $T_3$), and were 1.25, 0.82, $1.23{\mu}Sv/person$ for performing the scan task($T_4$, $T_5$, $T_6$). Conclusion we found that maintaining proper distance with patient and reducing contact time with patient had a significant effect on accumulated doses. Considering those points, efforts such as sufficient water intake and encourage of urination, maintaining the proper distance between the NMT and the patient(at least 100 cm), and reducing the contact time should be done for reducing dose rates not only patient but also NMT.

• Korean Journal of Soil Science and Fertilizer
• /
• v.35 no.2
• /
• pp.93-104
• /
• 2002

A field survey and experiment was conducted from 1996 to 1998 to develop rational technology for turfgrass vegetation of runway side of Incheon International Airport on the reclaimed tidal land in Young-Jong Island. Backfill of the experimental site was finished on August 1995. The experimental site was 8 ha located in the middle of the construction place for the main parking lot in front of the terminal building construction. The experimental field was drained by main open ditch, and divided three main plots, no subsurface tile drain, subsurface tile drain spacing with 22.5m, and with 45 m, respectively. The 17 sub plots were designed to test the effect of soil covering with red earth loam by 5 cm and 20 cm depth, application of chemical compound fertilizers and livestock manures, dressing of artifical soils and hydrophylic soil conditioners. The tested turfgrasses were three transplanting indigenous turfgrasses, Zoysia koreana, Zoysia sinica and Zoysia japonica, and two hydroseeding mixed exotic turgrasses, cool type I(tall fescue 30%, kentucky blue grass 40%, perenial ryegrass 30%), and cool type II(tall fescue 40%, perenial ryegrass 20%, fine fescue 20%, alkaligrass 20%). The soil backfilled with dredged seasand was sand textured with high salt concentration and low fertility. The soil showed high pH, low organic matter and low available phophate contents. The percolation rate was fast with high hydraulic conductivity. Desalinization was fast after installation of the main open drainage system. No subsurface tile drainage effect was found showing little difference in turfgrass growth. The covering and visual growth of turfgrasses were the best in the 20-cm soil covering with compound fertilizer treatment. The covering and visual growth of turfgrasses were satisfactory in the 5 cm soil covering with compound fertilizer treatment and with livestock manure treatments. The hydrophillic soil conditioner treatments were effective but expensive at present. The coverage and visual quality of turfgrasses were good for Zoysia koreana and Zoysia japonica. The coverages of turfgrasses by the hydroseeding with the mixed exotic turfgrasses were less than transplanting of native turfgrasses. In conclusion, for the runway side vegetation purposes, the subsurface tile drainage might not necessary as main open ditch drainage be sufficient due to fast percolation rate of the backfilled dredged seasand. The 5 cm soil covering with red earth might be sufficient for the runway side, but the 20 cm soil covering might be necessary for the runway side where high density of turfgrass coverage was necessary to protect from the airplance air blow.

• Korean Journal of Psychosomatic Medicine
• /
• v.8 no.2
• /
• pp.212-227
• /
• 2000

This study was designed to testify the reliability and validation on the Korean version of the Social Adaptation Self-rating Scale(SASS) which was developed from Bose et al. for the evaluation of social motivation and behavior of depressed patients in 1997. Interests for the social world, those of social functioning, of patients were involved in the addition of new measure of disturbance. And those were distinct from abnormalities of thought, mood and symptoms of patients with major depression. As the previous reports there were several evidences that treatments may be less likely to be effective if the system they act on is dysfunctional. Thus, a better social situation favoured better outcome. As a matter of fact, however, those reports were developed in the course of the evaluation of interpersonal therapy(IPT) and cognitive therapy. Accordingly the conversed question -whether pharmacological therapy with antidepressants can impact on social functioning in addition to addressing the core features of illness- has been addressed. To date, anyhow, it is accepted that enhancement of social functioning may be a therapeutic principle in its own right and illness rarely divorced from social context. In terms of those concepts the introduction of an assessment of social functioning into pharmacotherapeutic studies of depression has been welcomed and might be a potent instrument for evaluating the relative pharmacoeconomic benefits of different treatments. Despite of many scales which were applied for the evaluation of symptoms in the patients with depression, however, the scale for the evaluation of social functiong has not been introduced in Korea yet. Thus, this study was designed to introduce the concepts of social functioning in the patients with depression and to testify the reliability and validation on Korean version of SASS. This Korean version of SASS was submitted to a reliability and validation procedure based on the data from healthy general population survey in 291 individuals and 40 patients with major depression. Cronbach a was 0.790 in total subjects group and the correlation of test-retest was statistically significant(y=0.653, p<0.0l). Thus, the Korean version of SASS might be shown to be valid and reliable. The results of multivariate analyses allowed the identification of 3 principle factors(factor 1 = intersts in social activities, factor 2 = active interpersonal relationship, factor 3 = selfesteem) in normal group, however, it could be counted as only one factor in the depression group because nearly total items of SASS were involved in factor 1. In the view of these results, the Korean version of SASS may be useful additional tool for the evaluation of social functioning in depression.

• Journal of Intelligence and Information Systems
• /
• v.24 no.3
• /
• pp.1-19
• /
• 2018

Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, classification, and prediction problems. Especially, fashion business adopts deep learning techniques for apparel recognition, apparel search and retrieval engine, and automatic product recommendation. The core model of these applications is the image classification using Convolutional Neural Networks (CNN). CNN is made up of neurons which learn parameters such as weights while inputs come through and reach outputs. CNN has layer structure which is best suited for image classification as it is comprised of convolutional layer for generating feature maps, pooling layer for reducing the dimensionality of feature maps, and fully-connected layer for classifying the extracted features. However, most of the classification models have been trained using online product image, which is taken under controlled situation such as apparel image itself or professional model wearing apparel. This image may not be an effective way to train the classification model considering the situation when one might want to classify street fashion image or walking image, which is taken in uncontrolled situation and involves people's movement and unexpected pose. Therefore, we propose to train the model with runway apparel image dataset which captures mobility. This will allow the classification model to be trained with far more variable data and enhance the adaptation with diverse query image. To achieve both convergence and generalization of the model, we apply Transfer Learning on our training network. As Transfer Learning in CNN is composed of pre-training and fine-tuning stages, we divide the training step into two. First, we pre-train our architecture with large-scale dataset, ImageNet dataset, which consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. We use GoogLeNet for our main architecture as it has achieved great accuracy with efficiency in ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Second, we fine-tune the network with our own runway image dataset. For the runway image dataset, we could not find any previously and publicly made dataset, so we collect the dataset from Google Image Search attaining 2426 images of 32 major fashion brands including Anna Molinari, Balenciaga, Balmain, Brioni, Burberry, Celine, Chanel, Chloe, Christian Dior, Cividini, Dolce and Gabbana, Emilio Pucci, Ermenegildo, Fendi, Giuliana Teso, Gucci, Issey Miyake, Kenzo, Leonard, Louis Vuitton, Marc Jacobs, Marni, Max Mara, Missoni, Moschino, Ralph Lauren, Roberto Cavalli, Sonia Rykiel, Stella McCartney, Valentino, Versace, and Yve Saint Laurent. We perform 10-folded experiments to consider the random generation of training data, and our proposed model has achieved accuracy of 67.2% on final test. Our research suggests several advantages over previous related studies as to our best knowledge, there haven't been any previous studies which trained the network for apparel image classification based on runway image dataset. We suggest the idea of training model with image capturing all the possible postures, which is denoted as mobility, by using our own runway apparel image dataset. Moreover, by applying Transfer Learning and using checkpoint and parameters provided by Tensorflow Slim, we could save time spent on training the classification model as taking 6 minutes per experiment to train the classifier. This model can be used in many business applications where the query image can be runway image, product image, or street fashion image. To be specific, runway query image can be used for mobile application service during fashion week to facilitate brand search, street style query image can be classified during fashion editorial task to classify and label the brand or style, and website query image can be processed by e-commerce multi-complex service providing item information or recommending similar item.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.135-149
• /
• 2020

Artificial intelligences are changing world. Financial market is also not an exception. Robo-Advisor is actively being developed, making up the weakness of traditional asset allocation methods and replacing the parts that are difficult for the traditional methods. It makes automated investment decisions with artificial intelligence algorithms and is used with various asset allocation models such as mean-variance model, Black-Litterman model and risk parity model. Risk parity model is a typical risk-based asset allocation model which is focused on the volatility of assets. It avoids investment risk structurally. So it has stability in the management of large size fund and it has been widely used in financial field. XGBoost model is a parallel tree-boosting method. It is an optimized gradient boosting model designed to be highly efficient and flexible. It not only makes billions of examples in limited memory environments but is also very fast to learn compared to traditional boosting methods. It is frequently used in various fields of data analysis and has a lot of advantages. So in this study, we propose a new asset allocation model that combines risk parity model and XGBoost machine learning model. This model uses XGBoost to predict the risk of assets and applies the predictive risk to the process of covariance estimation. There are estimated errors between the estimation period and the actual investment period because the optimized asset allocation model estimates the proportion of investments based on historical data. these estimated errors adversely affect the optimized portfolio performance. This study aims to improve the stability and portfolio performance of the model by predicting the volatility of the next investment period and reducing estimated errors of optimized asset allocation model. As a result, it narrows the gap between theory and practice and proposes a more advanced asset allocation model. In this study, we used the Korean stock market price data for a total of 17 years from 2003 to 2019 for the empirical test of the suggested model. The data sets are specifically composed of energy, finance, IT, industrial, material, telecommunication, utility, consumer, health care and staple sectors. We accumulated the value of prediction using moving-window method by 1,000 in-sample and 20 out-of-sample, so we produced a total of 154 rebalancing back-testing results. We analyzed portfolio performance in terms of cumulative rate of return and got a lot of sample data because of long period results. Comparing with traditional risk parity model, this experiment recorded improvements in both cumulative yield and reduction of estimated errors. The total cumulative return is 45.748%, about 5% higher than that of risk parity model and also the estimated errors are reduced in 9 out of 10 industry sectors. The reduction of estimated errors increases stability of the model and makes it easy to apply in practical investment. The results of the experiment showed improvement of portfolio performance by reducing the estimated errors of the optimized asset allocation model. Many financial models and asset allocation models are limited in practical investment because of the most fundamental question of whether the past characteristics of assets will continue into the future in the changing financial market. However, this study not only takes advantage of traditional asset allocation models, but also supplements the limitations of traditional methods and increases stability by predicting the risks of assets with the latest algorithm. There are various studies on parametric estimation methods to reduce the estimated errors in the portfolio optimization. We also suggested a new method to reduce estimated errors in optimized asset allocation model using machine learning. So this study is meaningful in that it proposes an advanced artificial intelligence asset allocation model for the fast-developing financial markets.

• Tuberculosis and Respiratory Diseases
• /
• v.59 no.3
• /
• pp.286-297
• /
• 2005

Background : Non-small cell lung cancer (NSCLC) is a common cause of cancer-related death in North America and Korea, with an overall 5-year survival rate of between 4 and 14%. The TNM staging system is the best prognostic index for operable NSCLC . However, epidermal growth factor receptor (EGFR), matrix metalloproteinase-9(MMP-9), and C-erbB-2 have all been implicated in the pathogenesis of NSCLC and might provide prognostic information. Methods : Immunohistochemical staining of 81 specimens from a resected primary non-small cell lung cancer was evaluated in order to determine the role of the biological markers on NSCLC . Immunohistochemical staining for EGFR, MMP-9, and C-erbB-2 was performed on paraffin-embedded tissue sections to observe the expression pattern according to the pathologic type and surgical staging. The correlations between the expression of each biological marker and the survival time was determined. Results : When positive immunohistochemical staining was defined as the extent area>20%(more than Grade 2), the positive rates for EGFR, MMP-9, and C-erbB-2 staining were 71.6%, 44.3%, and 24.1% of the 81 patients, respectively. The positive rates of EGFR and MMP-9 stain for NSCLC according to the surgical stages I, II, and IIIa were 75.0% and 41.7%, 66.7% and 47.6%, and 76.9% and 46.2%, respectively. The median survival time of the EGFR(-) group, 71.8 months, was significantly longer than that of the EGFR(+) group, 33.5 months.(p=0.018, Kaplan-Meier Method, log-rank test).. The MMP-9(+) group had a shorter median survival time than the MMP-9(-) group, 35.0 and 65.3 months, respectively (p=0.2). The co-expression of EGFR and MMP-9 was associated with a worse prognosis with a median survival time of 26.9 months, when compared with the 77 months for both negative-expression groups (p=0.0023). There were no significant differences between the C-erbB-2(+) and C-erbB-2 (-) groups. Conclusion : In NSCLC, the expression of EGFR might be a prognostic factor, and the co-expression of EGFR and MMP-9 was found to be associated with a poor prognosis. However, C-erbB-2 expression had no prognostic significance.

• Korean Journal of Agricultural Science
• /
• v.16 no.2
• /
• pp.169-178
• /
• 1989

The study was carried out to elucidate the feedback mechanism on the hypothalamo-hypophyseal system from the functional changes of ovary in female rats. One hundred and forty-four mature female rats were lloted into the three groups; ovariectoimzed group, estradiol treated group and intact control group. The varies of 48 heads of rat were completely removed. Forty eight heads of rat were administered with $200{\mu}g$ of estradiol benzoate every 48 hours. Serum FSH, LH and prolactin levels were determined with radioimmunoassay method at 3,6,12,24 ours, and 5,10, and 15days after treatment. The rats were necropsied to measure the weights of hypophysis and to examin the histological changes in the organs. The results obtained were as follows: The weights of hypophysis were increased after ovariectomy and decreased after estradiol injection. The differences in hypophysis weights were significant between the group from 5 days after treatment. The histological changes in hypophysis were appeared from 5th day after ovariectomy. Proliferation and hypertrophy began to occur in basophilic from 10th day after ovariectomy, chromophobes were slightly hypertrophied and acidophilic cells were atrophied. In estradiol injected rats the histological findings were appeard to be contrary to those of ovariectomized rats. Serum FSH levels significantly changed after ovariectomy and estradiol injection and were higher in both the treated groups than in the intact control group. Within 18 hours after treatment the level was the highest in ovariectomized group, and thereafter the highest level was found in estradiol treated gorup. In ovariectomized rats the levels were rapidly increased 3 hours after treatment and maximum levels were found 18 hours after treatment. In estradiol treated rats the levels started to increase 18 hours after treatment and reached maximum levels 24 hours treatment. 4. Serum LH levels started to increase 3 hours after ovariectomy and estradiol injection and reached maximum levels 12 hours after ovariectomy and 24 hours after estradiol injection. There were significant differences in LH levels between the groups in each observation time. Up to 18 hours after treatment levels were higher in ovariectomized rats than in estradiol treated rats. but thereafter the levels were higher in estradiol treated rats than in ovariectomized rats. The multiple range test showed that a significant difference in LH levels was not found between ovariectomized group and estradiol treated group 18 hours and 5 days after treatment. 5. Serum prolactin levels were significantly changed after ovariectomy and estradiol injection. The levels were lower in ovariectomized rats than in intact control rats.

• Journal of the Korean Society for Library and Information Science
• /
• v.25
• /
• pp.343-388
• /
• 1993

The ability of any organization to innovate itself in a rapid change of environment means the existence of the organization. Innovative activity is achieved in different ways according to the objectives of organization. the characteristics of external environmental factors. and various attributes in organization. In the present study. all the existing approaches to the innovative nature of organization were synthetically compared to each other and evaluated: then. for a more rational approach. a research model was built and suggested by establishing the inclusive variables of the innovative nature of library organization and categorizing the types of such nature. Additionally. an empirical. analytical study on such a model was done. That is. paying regard to the fact that innovation has basically a close relation with the circumstantial factors of organization. synthetic, circumstantial relations were clarified. considering the external environmental factors and internal characteristics of organization. In the study. the innovation of library organization was seen in two parts i.e .. the feasible degree of technical innovation and the feasible degree of administrative innovation. Regarding the types of innovative implementation. according to the feasible degree of innovation, four types such as a stationary type. technic-oriented type, organization-oriented type. and technical-socio systematic type were classified. There were nine independent variables-i.e., the scale of organization. available resources of the organization, formalization, differentiation, specialization. decentralization, recognizant degree of the technical attribute. degree of response to the change of technical environment, and professional activities. There were three subordinate variables - i.e., technical innovation, administrative innovation. and the performance of organization. Through establishment of such variables, the factors which might influence the innovation of library organization were understood, and with the types of the innovative implementation of library organization being classified according to the feasible degree of innovation. the characteristics of library organization were reviewed in the light of each type. Also. the performance of library organization according to the types of the innovative implementation of library organization was analyzed. and the relations between the types of innovative implementation according to circumstantial variables and the performance of library organization were clarified. In order to clarify the adequacy of the research model in the methodology of empirical study, data were collected from 72 university libraries and 38 special libraries. and for a hypothetical test of the research model. an analysis of correlations, a stepwise regression analysis. and One Way ANOVA were utilized. The following are the major results or findings from the study 1) It appeared there is a trend that the bigger the scale of organization and available resources are, the more active the professional activity of the managerial class is, and the higher the recognizant degree of technical environment (recognizant degree of technical attributes and the degree of response t9 the change of technical environment) is, the higher the feasible degree of innovation becomes. 2) It appeared that among the variables influencing the feasible degree of technical innovation, the order from the variable influencing most was first, the recognizant degree of technical innovation: second, the available resources of organization: and third, professional activity. Regarding the variables influencing the feasible degree of administrative innovation from the most influential variable, it appeared they were the available resources of organization, the differentiation of organization. and the degree of response to the change of technical environment. 3) It appeared that the higher the educational level of the managerial class is, the more active the professional activity becomes. It seemed there is a trend that the group of library managers whose experience as a librarian was at the middle level(three years to six years of experience) was more active in research activity than the group of library managers whose experience as a librarian was at a higher level(more than ten years). Also, it appeared there is a trend that the lower the age of library managers is, the higher the recognizant degree of technical attributes becomes. and the group of library managers whose experience as a librarian was at the middle level (three years to six years of experience) recognized more affirmatively the technical aspect than the group of library managers whose experience as a librarian was at a higher level(more than 10 years). Also, it appeared that, when the activity of the professional association and research activity are active, the recognizant degree of technology becomes higher, and as a result. it influences the innovative nature of organization(the feasible degree of technical innovation and the feasible degree of administrative innovation). 4) As a result of the comparison and analysis of the characteristics of library organization according to the types of innovative implementation of library organization. it was indicated there is a trend that the larger the available resources of library organization, the higher the organic nature of organization such as differentiation. decentralization, etc., and the higher the level of the operation of system development, the more the type of the innovative implementation of library organization becomes the technical-socio systematic type which is higher both in the practical degrees of technical innovation and administrative innovation. 5) As a result of the comparison and analysis of the relations between the types of innovative implementation and the performance of organization, it appeared that the order from the highest performance of organization is the technical-socio systematic type, then the technic-oriented type, the organization­oriented type, and finally the stationary type which is lowest in such performance. That is, it demonstrated that, since the performance of library organization is highest in the library of the technical-socio systematic type while it is lowest in the library whose practical degrees in both technical innovation and administrative innovation are low, the performance of library organization differs significantly according to the types of innovative implementation of library organization. The present study has extracted the factors influencing innovation, classified systematically the types of innovative implementation, and inferred the synthetical, circumstantial correlations between the types and the performance of organization, and empirically inspected those factors. However, due to the present study's restrictive matters and the limit of the research design, results from the study should be more prudently interpreted. Also, the present study, as an investigative study of the types of innovative implementation, with few preceding studies, requires more complete hypothetical inference based on the results of the present study. In other words, if more systematical studies are given to understanding the relations, it will devote the suggestion and demonstration of a more useful theory.