• Korean Journal of Ecology and Environment
• /
• v.55 no.1
• /
• pp.35-48
• /
• 2022

Fish assemblage of total 325 of Korean peninsula estuaries were surveyed to analyze the characteristics of community structure and diversity by sea areas for three years from 2016 to 2018. The scale (stream width) of Korean estuaries were various (14~3,356 m), and 68.9% of all estuaries showed salinity of less than 2 psu. Total 149 species classified into 52 families of fish were identified, and the dominant and sub-dominant species were Tribolodon hakonensis (relative abundance, RA, 12.5%) and Mugil cephalus (RA, 9.5%), respectively. The estuary of the Korean Peninsula had different physical and chemical habitat environments depending on the sea area, and accordingly, fish community structure also showed statistically significant differences (PERMANOVA, Pseudo-F=26.69, P=0.001). In addition, the NMDS (nonmetric multidimensional scaling) results showed the patterns that indicating fish community difference by sea areas, even though low community similarity within sea area (SIMPER, 21.79~26.39%). The estuaries of east sea areas were distinguished from the others in the aspects of which, the higher importance of migratory fishes and endangered species, and that of brackish species were characterized at south sea estuaries. However, the estuaries of west sea showed higher importance of species that have a relation with freshwater (primary freshwater species, exotic species), which is the result that associating with the lower salinity of west sea estuaries because of the high ratio of closed estuaries(78.2%). The SIMPER analysis, scoring the contribution rates of species to community similarity, also showed results corresponding to the tendency of different fish community structures according to each sea area. So far, In Korea, most studies on fish communities in estuaries have been conducted in a single estuary unit, which made it difficult to understand the characteristics of estuaries at the national level, which are prerequisite for policy establishment. In present study, we are providing fish community structure characteristics of Korean estuaries in a national scale, including diversity index, habitat salinity ranges of major species, distribution of migratory species. We are expecting that our results could be utilized as baseline information for establishing management policies or further study of Korean estuaries.

• Journal of the Korean Applied Science and Technology
• /
• v.37 no.3
• /
• pp.473-483
• /
• 2020

This paper proposes a tourism recommendation system in intelligent cloud environment using information of tourist behavior applied with perceived value. This proposed system applied tourist information and empirical analysis information that reflected the perceptual value of tourists in their behavior to the tourism recommendation system using wide and deep learning technology. This proposal system was applied to the tourism recommendation system by collecting and analyzing various tourist information that can be collected and analyzing the values that tourists were usually aware of and the intentions of people's behavior. It provides empirical information by analyzing and mapping the association of tourism information, perceived value and behavior to tourism platforms in various fields that have been used. In addition, the tourism recommendation system using wide and deep learning technology, which can achieve both memorization and generalization in one model by learning linear model components and neural only components together, and the method of pipeline operation was presented. As a result of applying wide and deep learning model, the recommendation system presented in this paper showed that the app subscription rate on the visiting page of the tourism-related app store increased by 3.9% compared to the control group, and the other 1% group applied a model using only the same variables and only the deep side of the neural network structure, resulting in a 1% increase in subscription rate compared to the model using only the deep side. In addition, by measuring the area (AUC) below the receiver operating characteristic curve for the dataset, offline AUC was also derived that the wide-and-deep learning model was somewhat higher, but more influential in online traffic.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.23-45
• /
• 2020

Big data is creating in a wide variety of fields such as medical care, manufacturing, logistics, sales site, SNS, and the dataset characteristics are also diverse. In order to secure the competitiveness of companies, it is necessary to improve decision-making capacity using a classification algorithm. However, most of them do not have sufficient knowledge on what kind of classification algorithm is appropriate for a specific problem area. In other words, determining which classification algorithm is appropriate depending on the characteristics of the dataset was has been a task that required expertise and effort. This is because the relationship between the characteristics of datasets (called meta-features) and the performance of classification algorithms has not been fully understood. Moreover, there has been little research on meta-features reflecting the characteristics of multi-class. Therefore, the purpose of this study is to empirically analyze whether meta-features of multi-class datasets have a significant effect on the performance of classification algorithms. In this study, meta-features of multi-class datasets were identified into two factors, (the data structure and the data complexity,) and seven representative meta-features were selected. Among those, we included the Herfindahl-Hirschman Index (HHI), originally a market concentration measurement index, in the meta-features to replace IR(Imbalanced Ratio). Also, we developed a new index called Reverse ReLU Silhouette Score into the meta-feature set. Among the UCI Machine Learning Repository data, six representative datasets (Balance Scale, PageBlocks, Car Evaluation, User Knowledge-Modeling, Wine Quality(red), Contraceptive Method Choice) were selected. The class of each dataset was classified by using the classification algorithms (KNN, Logistic Regression, Nave Bayes, Random Forest, and SVM) selected in the study. For each dataset, we applied 10-fold cross validation method. 10% to 100% oversampling method is applied for each fold and meta-features of the dataset is measured. The meta-features selected are HHI, Number of Classes, Number of Features, Entropy, Reverse ReLU Silhouette Score, Nonlinearity of Linear Classifier, Hub Score. F1-score was selected as the dependent variable. As a result, the results of this study showed that the six meta-features including Reverse ReLU Silhouette Score and HHI proposed in this study have a significant effect on the classification performance. (1) The meta-features HHI proposed in this study was significant in the classification performance. (2) The number of variables has a significant effect on the classification performance, unlike the number of classes, but it has a positive effect. (3) The number of classes has a negative effect on the performance of classification. (4) Entropy has a significant effect on the performance of classification. (5) The Reverse ReLU Silhouette Score also significantly affects the classification performance at a significant level of 0.01. (6) The nonlinearity of linear classifiers has a significant negative effect on classification performance. In addition, the results of the analysis by the classification algorithms were also consistent. In the regression analysis by classification algorithm, Naïve Bayes algorithm does not have a significant effect on the number of variables unlike other classification algorithms. This study has two theoretical contributions: (1) two new meta-features (HHI, Reverse ReLU Silhouette score) was proved to be significant. (2) The effects of data characteristics on the performance of classification were investigated using meta-features. The practical contribution points (1) can be utilized in the development of classification algorithm recommendation system according to the characteristics of datasets. (2) Many data scientists are often testing by adjusting the parameters of the algorithm to find the optimal algorithm for the situation because the characteristics of the data are different. In this process, excessive waste of resources occurs due to hardware, cost, time, and manpower. This study is expected to be useful for machine learning, data mining researchers, practitioners, and machine learning-based system developers. The composition of this study consists of introduction, related research, research model, experiment, conclusion and discussion.

• Management & Information Systems Review
• /
• v.33 no.5
• /
• pp.97-115
• /
• 2014

The purpose of this study is to illustrating how the store selection factors affect the customer's satisfaction about the store and the intention of revisit base on the analyzing the store selection factors. At the same time, the difference between the influence on the customer's satisfaction and revisit intention of that in Korea and in China is also compared. Accordingly, through the notional understanding of configuration variables and the investigation of previous research, the Research hypothesis was set and the relevance between the two was inspected. The survey was aimed at the Korean customers who visit the large supermarkets in Korea and Chinese customers who visit the large supermarkets in China. After that, the reliability and validity of the collected data was verified and the research hypothesis was validated by structure equation modeling. The result of this study can by concluded as follows: First, in Korea the customer's satisfaction is showed to be affected by store selection factors, product property, service property and physical environment. Second, in China the customer's satisfaction is showed to be affected by store selection factors, product property, service property and physical environment. Third, in Korea the revisit intention is showed to be affected by customer's satisfaction. Forth, in China the revisit intention is showed to be affected by customer's satisfaction. Fifth, it shows little difference between the store selection factors of the customers visiting large supermarkets in Korea and in China. According to the research results above, the implications can be drawn as the customer's satisfaction of those who visit the large supermarkets may be affected by store selection factors (store property, product property, service property and physical environment). In recent years, more and more overseas large supermarkets are opening in both Korea and China and the competition among each is become more intense day by day. Every larger supermarket is trying their best to refine their store property, product property, service property and physical environment, in order to enhance the customer's satisfaction. The biggest factor that affects the customer's satisfaction and revisit intention in Korea is service property, So that the services requires proper measures and improvement. In China, the factor that affects most on the customer's satisfaction and revisit intention is physical environment. In order to enhance the customer's positive consciousness of stores, the physical environment needs to be well constructed. Lastly, in the compared research between Korea and China, the distribution of survey responders was limited from certain areas. Therefore, the further study can be implemented by more research in various geographical areas and more development in store selection factors.

• Management & Information Systems Review
• /
• v.36 no.4
• /
• pp.21-32
• /
• 2017

Motivated by the recent cases of negligent social responsibility as manifested by foreign luxury fashion brands in Korea, this study investigates whether agency costs depend on the sustainability of different types of corporate governance. Agency costs refer either to vertical costs arising from the relationship between stockholders and managers, or to horizontal costs associated with the potential conflicts between majority and minority stockholders. The firms with luxury fashion brand could spend large sums of money on maintenance of magnificent brand image, thereby increasing the agency cost. On the contrary, the firms may hold down wasteful spending to report a gaudily financial achievement. This results in mitigation of the agency cost. Agency costs are measured by the value of the principal component. First, three ratios are constructed: asset turnover, operating expense to sales, and earnings before interest, tax, and depreciation. Then, the scores of each of these ratios for individual firms in the sample are differenced from the ratios for the benchmark firm of S-OIL. S-OIL was designated as the best superior governance model firm for 2013 by CGS. We perform regression analysis of each agency cost index, luxury fashion brand dummy and a set of control variables. The regression results indicate that the agency costs of the firms with luxury fashion brand exceed those of control group in the fashion industry in the part of operating expenses, but the agency cost falls short of those of control group in the part of EBITD, thus the aggregate agency costs are not differential of those of the control group. In sensitivity test, the results are same that the agency cost of the firms are higher than those of the matching control group with PSM(propensity matching method). These results are corroborated by an additional analysis comparing the group of the companies with the best brands with the control group. The results raise doubts about the effectiveness of management of the firms with luxury fashion brand. This study has a limitation that the research has performed only for 2013 and this paper suggests that there is room for improvement in the current research methodology.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.11S
• /
• pp.3651-3667
• /
• 2000

Recently, in the distributed multimedia environments based on internet, as radical growing technologies, the most of researchers focus on both streaming technology and distributed object thchnology, Specially, the studies which are tried to integrate the streaming services on the distributed object technology have been progressing. These technologies are applied to various stream service mamgements and protocols. However, the stream service management mexlels which are being proposed by the existing researches are insufficient for suporting the QoS of stream services. Besides, the existing models have the problems that cannot support the extensibility and the reusability, when the QoS-reiatedfunctions are being developed as a sub-module which is suited on the specific-purpose application services. For solving these problems, in this paper. we suggested a QoS Integrated platform which can extend and reuse using the distributed object technologies, and guarantee the QoS of the stream services. A structure of platform we suggested consists of three components such as User Control Module(UCM), QoS Management Module(QoSM) and Stream Object. Stream Object has Send/Receive operations for transmitting the RTP packets over TCP/IP. User Control ModuleI(UCM) controls Stream Objects via the COREA service objects. QoS Management Modulel(QoSM) has the functions which maintain the QoS of stream service between the UCMs in client and server. As QoS control methexlologies, procedures of resource monitoring, negotiation, and resource adaptation are executed via the interactions among these comiXments mentioned above. For constmcting this QoS integrated platform, we first implemented the modules mentioned above independently, and then, used IDL for defining interfaces among these mexlules so that can support platform independence, interoperability and portability base on COREA. This platform is constructed using OrbixWeb 3.1c following CORBA specification on Solaris 2.5/2.7, Java language, Java, Java Media Framework API 2.0, Mini-SQL1.0.16 and multimedia equipments. As results for verifying this platform functionally, we showed executing results of each module we mentioned above, and a numerical data obtained from QoS control procedures on client and server's GUI, while stream service is executing on our platform.

• Journal of Intelligence and Information Systems
• /
• v.18 no.1
• /
• pp.59-69
• /
• 2012

As exhibition industry, which is a part of 17 new growth engines of the government, is related to other industries such as tourism, transportation and financial industries. So it has a significant ripple effect on other industries. Exhibition is a knowledge-intensive, eco-friendly and high value-added Industry. Over 13,000 exhibitions are held every year around the world which contributes to getting foreign currency. Exhibition industry is closely related with culture and tourism and could be utilized as local and national development strategies and improve national brand image as well. Many countries try various efforts to invigorate exhibition industry by arranging related laws and support system. In Korea, more than 200 exhibitions are being held every year, but only 2~3 exhibitions are hosted with over 400 exhibitors and except these exhibitions most exhibitions have few foreign exhibitors. The main reason of weakness of domestic trade show is that there are no agencies managing exhibitionrelated statistics and there is no specific and reliable evaluation. This might cause impossibility of providing buyer or seller with reliable data, poor growth of exhibitions in terms of quality and thus service quality of trade shows cannot be improved. Hosting a lot of visitors (Public/Buyer/Exhibitor) is very crucial to the development of domestic exhibition industry. In order to attract many visitors, service quality of exhibition and visitor's satisfaction should be enhanced. For this purpose, a variety of real-time customized services through digital media and the services for creating new customers and retaining existing customers should be provided. In addition, by providing visitors with personalized information services they could manage their time and space efficiently avoiding the complexity of exhibition space. Exhibition industry can have competitiveness and industrial foundation through building up exhibition-related statistics, creating new information and enhancing research ability. Therefore, this paper deals with customized service with visitor's smart-phone at the exhibition space and designing mobile framework which enables exhibition devices to interact with other devices. Mobile server framework is composed of three different systems; multi-server interaction, server, client, display device. By making knowledge pool of exhibition environment, the accumulated data for each visitor can be provided as personalized service. In addition, based on the reaction of visitors each of all information is utilized as customized information and so the cyclic chain structure is designed. Multiple interaction server is designed to have functions of event handling, interaction process between exhibition device and visitor's smart-phone and data management. Client is an application processed by visitor's smart-phone and could be driven on a variety of platforms. Client functions as interface representing customized service for individual visitors and event input and output for simultaneous participation. Exhibition device consists of display system to show visitors contents and information, interaction input-output system to receive event from visitors and input toward action and finally the control system to connect above two systems. The proposed mobile framework in this paper provides individual visitors with customized and active services using their information profile and advanced Knowledge. In addition, user participation service is suggested as well by using interaction connection system between server, client, and exhibition devices. Suggested mobile framework is a technology which could be applied to culture industry such as performance, show and exhibition. Thus, this builds up the foundation to improve visitor's participation in exhibition and bring about development of exhibition industry by raising visitor's interest.

• 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.

• Information Systems Review
• /
• v.12 no.1
• /
• pp.145-164
• /
• 2010

Since the third generation (3G) mobile communication service has been launched by most mobile communication operators in Korea, the portion of data service in mobile communication service becomes one of the most important factors in mobile communication service market. In past mobile communication market, most mobile communication operators made their profit mostly from voice communication service. However, the portion of profit from data service has gradually increased based on both video phone call and mobile Internet service. In this situation, LG telecom launched the full browsing mobile Internet service. This service provides a new type of mobile Internet service platform which enables to access the World Wide Web using mobile browsers, so we generally access the Web using web browsers in the desktop computer. Under the open network structure of mobile Internet like situation, it is very important to analyze the factors which can affect the competition between mobile communication service companies. So, in this paper, we first present the current state of full browsing service, followed by the expectation of its growth potentials and barriers. Then, we analyze the advantages and disadvantage of LG telecom as a first-mover and SK telecom/KTF as followers. Finally, based on this analysis, we predict the future competition among these companies and the market.

• Information Systems Review
• /
• v.15 no.3
• /
• pp.129-139
• /
• 2013

The Internet of Things (or IoT for short) which refers to uniquely identifiable objects and their virtual representations in an Internet-like structure is to be reality today. The amount of data on IoT is expected to increase abruptly and there are several key issues like usefulness interoperability between multiple distributes systems, services and databases. In this paper a methodology is proposed to realize a recently developed cloud service model, Contents as a Service (CaaS), which is contents delivery model referred to as 'on-demand contents'. In the proposed method, the global object identification, ePosition, comprising the structured form of two sorts of text strings with a separation symbol like # is applied to identify a specific content and registered with the content at the same server. It is easy-to-realize and effective to solve the interoperability problem systematically and logically. Some APIs for the proposed CaaS service are to be converged to provide some upgraded cloud service model such as 'CaaS supported SaaS' and 'CaaS supported PaaS'.