• Korea Science and Art Forum
• /
• v.25
• /
• pp.227-240
• /
• 2016

This study aims to examine how visitors in science centers interact with gesture-based exhibits from embodied cognition perspectives. Four gesture-based exhibits in two science centers were selected for this study. In addition, we interviewed a total of 14 visitor groups to examine how they perceived the property of gesture-based exhibit. We also interviewed four experts to further examine the benefits and limitations of the current gesture-based exhibits in science centers. The research results indicate that the total amount of interaction time between visitors and gesture-based exhibits was not high overall, implying that there was little of visitors' immersive engagement. Both experts and visitors expressed that the current gesture-based exhibits tend to highlight the novelty effect but little obvious impacts linking gestures and learning. Drawing from the key findings, this study suggests the following design considerations for gesture-based exhibits. First, to increate visitor's initial engagement, the purpose and usability of gesture-based exhibits should be considered from the initial phase of design. Second, to promote meaningful interaction, it is important to sustain visitors' initial engagement. For that, gesture-based exhibits should be transformed to promote intellectual curiosity beyond simple interaction. Third, from embodied cognition perspectives, exhibits design should reflect how the mappings between specific gestures and metaphors affect learning processes. Lastly, this study suggests that future gesture-based exhibits should be designed toward promoting interaction among visitors and adaptive inquiry.

• Korean Journal of Remote Sensing
• /
• v.39 no.3
• /
• pp.311-323
• /
• 2023

Oil spill accidents can cause various environmental issues, so it is important to quickly assess the extent and changes in the area and location of the spilled oil. In the case of oil spill detection using satellite imagery, it is possible to detect a wide range of oil spill areas by utilizing the information collected from various sensors equipped on the satellite. Previous studies have analyzed the reflectance of oil at specific wavelengths and have developed an oil spill index using bands within the specific wavelength ranges. When analyzing multiple images before and after an oil spill for monitoring purposes, a significant amount of time and computing resources are consumed due to the large volume of data. By utilizing Google Earth Engine, which allows for the analysis of large volumes of satellite imagery through a web browser, it is possible to efficiently detect oil spills. In this study, we evaluated the applicability of four types of oil spill indices in the area of various land cover using Sentinel-2 MultiSpectral Instrument data and the cloud-based Google Earth Engine platform. We assessed the separability of oil spill areas by comparing the index values for different land covers. The results of this study demonstrated the efficient utilization of Google Earth Engine in oil spill detection research and indicated that the use of oil spill index B ((B3+B4)/B2) and oil spill index C (R: B3/B2, G: (B3+B4)/B2, B: (B6+B7)/B5) can contribute to effective oil spill monitoring in other regions with complex land covers.

• Journal of Internet Computing and Services
• /
• v.25 no.4
• /
• pp.121-130
• /
• 2024

As internet and communication technology (ICT) is improved exponentially, types and amount of available data also increase. Even though data analysis including statistics is significant to utilize this large amount of data, there are inevitable limits to process various and complex data in general way. Meanwhile, there are many attempts to apply machine learning (ML) in various fields to solve the problems according to the enhancement in computational performance and increase in demands for autonomous systems. Especially, data processing for the model input and designing the model to solve the objective function are critical to achieve the model performance. Data processing methods according to the type and property have been presented through many studies and the performance of ML highly varies depending on the methods. Nevertheless, there are difficulties in deciding which data processing method for data analysis since the types and characteristics of data have become more diverse. Specifically, multi-variate data processing is essential for solving non-linear problem based on ML. In this paper, we present a multi-variate tabular data processing scheme for ML-aided data analysis by using Titanic dataset from Kaggle including various kinds of data. We present the methods like input variable filtering applying statistical analysis and normalization according to the data property. In addition, we analyze the data structure using visualization. Lastly, we design an ML model and train the model by applying the proposed multi-variate data process. After that, we analyze the passenger's survival prediction performance of the trained model. We expect that the proposed multi-variate data processing and visualization can be extended to various environments for ML based analysis.

• Journal of Intelligence and Information Systems
• /
• v.20 no.2
• /
• pp.109-122
• /
• 2014

People are nowadays creating a tremendous amount of data on Social Network Service (SNS). In particular, the incorporation of SNS into mobile devices has resulted in massive amounts of data generation, thereby greatly influencing society. This is an unmatched phenomenon in history, and now we live in the Age of Big Data. SNS Data is defined as a condition of Big Data where the amount of data (volume), data input and output speeds (velocity), and the variety of data types (variety) are satisfied. If someone intends to discover the trend of an issue in SNS Big Data, this information can be used as a new important source for the creation of new values because this information covers the whole of society. In this study, a Twitter Issue Tracking System (TITS) is designed and established to meet the needs of analyzing SNS Big Data. TITS extracts issues from Twitter texts and visualizes them on the web. The proposed system provides the following four functions: (1) Provide the topic keyword set that corresponds to daily ranking; (2) Visualize the daily time series graph of a topic for the duration of a month; (3) Provide the importance of a topic through a treemap based on the score system and frequency; (4) Visualize the daily time-series graph of keywords by searching the keyword; The present study analyzes the Big Data generated by SNS in real time. SNS Big Data analysis requires various natural language processing techniques, including the removal of stop words, and noun extraction for processing various unrefined forms of unstructured data. In addition, such analysis requires the latest big data technology to process rapidly a large amount of real-time data, such as the Hadoop distributed system or NoSQL, which is an alternative to relational database. We built TITS based on Hadoop to optimize the processing of big data because Hadoop is designed to scale up from single node computing to thousands of machines. Furthermore, we use MongoDB, which is classified as a NoSQL database. In addition, MongoDB is an open source platform, document-oriented database that provides high performance, high availability, and automatic scaling. Unlike existing relational database, there are no schema or tables with MongoDB, and its most important goal is that of data accessibility and data processing performance. In the Age of Big Data, the visualization of Big Data is more attractive to the Big Data community because it helps analysts to examine such data easily and clearly. Therefore, TITS uses the d3.js library as a visualization tool. This library is designed for the purpose of creating Data Driven Documents that bind document object model (DOM) and any data; the interaction between data is easy and useful for managing real-time data stream with smooth animation. In addition, TITS uses a bootstrap made of pre-configured plug-in style sheets and JavaScript libraries to build a web system. The TITS Graphical User Interface (GUI) is designed using these libraries, and it is capable of detecting issues on Twitter in an easy and intuitive manner. The proposed work demonstrates the superiority of our issue detection techniques by matching detected issues with corresponding online news articles. The contributions of the present study are threefold. First, we suggest an alternative approach to real-time big data analysis, which has become an extremely important issue. Second, we apply a topic modeling technique that is used in various research areas, including Library and Information Science (LIS). Based on this, we can confirm the utility of storytelling and time series analysis. Third, we develop a web-based system, and make the system available for the real-time discovery of topics. The present study conducted experiments with nearly 150 million tweets in Korea during March 2013.

• Journal of the Korean Wood Science and Technology
• /
• v.2 no.4
• /
• pp.4-14
• /
• 1974

The total forest land area in Korea accounts for some 67 percent of the nation's land total. Its productivity, however, is very low. Consequently, forest production accounts for only about 2 percent of the gross national product and a minor proportion of no more than about 5 percent versus primary industry. In this case, however, only the direct income from forestry is taken into account, making no reference to the forestry output induced by other industrial sectors. The value added Or the induced forestry output in manufacturing the primary wood products into higher quality products, makes a larger contribution to the economy than direct contribution. So, this author has tried to analyze the structure of forestry and compute the repercussion effect and the induced output of primary forest products when utilized by other industries for their raw materials, Hsing the input-output table and attached tables for 1963 and 1966 issued by the Bank of Korea. 1. Analysis of forestry structure A. Changes in total output Durng the nine-year period, 1961-1969, the real gross national product in Korea increased 2.1 times, while that of primary industries went up about 1. 4 times. Forestry which was valued at 9,380 million won in 1961, was picked up about 2. 1 times to 20, 120 million won in 1969. The rate of the forestry income in the GNP, accordingly, was no more than 1.5 percent both in 1961 and 1962, whereas its rate in primary industries increased 3.5 to 5.4 percent. Such increase in forestry income is attributable to increased forest production and rise in timber prices. The rate of forestry income, nonetheless, was on the decrease on a gradual basis. B. Changes in input coefficient The input coefficient which indicates the inputs of the forest products into other sectors were up in general in 1966 over 1963. It is noted that the input coefficient indicating the amount of forest products supplied to such industries closely related with forestry as lumber and plywood, and wood products and furniture, showed a downward trend for the period 1963-1966. On the other hand, the forest input into other sectors was generally on the increase. Meanwhile, the input coefficient representing the yolume of the forest products supplied to the forestry sector itself showed an upward tendency, which meant more and more decrease in input from other sectors. Generally speaking, in direct proportion to the higher input coefficient in any industrial sector, the reinput coefficient which denotes the use of its products by the same sector becomes higher and higher. C. Changes in ratio of intermediate input The intermediate input ratio showing the dependency on raw materials went up to 15.43 percent m 1966 from 11. 37 percent in 1963. The dependency of forestry on raw materials was no more than 15.43 percent, accounting for a high 83.57 percent of value added. If the intermediate input ratio increases in any given sector, the input coefficient which represents the fe-use of its products by the same sector becomes large. D. Changes in the ratio of intermediate demand The ratio of the intermediate demand represents the characteristics of the intermediary production in each industry, the intermediate demand ratio in forestry which accunted for 69.7 percent in 1963 went up to 75.2 percent in 1966. In other words, forestry is a remarkable industry in that there is characteristics of the intermediary production. E. Changes in import coefficient The import coefficient which denotes the relation between the production activities and imports, recorded at 4.4 percent in 1963, decreased to 2.4 percent in 1966. The ratio of import to total output is not so high. F. Changes in market composition of imported goods One of the major imported goods in the forestry sector is lumber. The import value increased by 60 percent to 667 million won in 1966 from 407 million won in 1963. The sales of imported forest products to two major outlets-lumber and plywood, and wood products and furniture-increased to 343 million won and 31 million won in 1966 from 240million won and 30 million won in 1963 respectively. On the other hand, imported goods valued at 66 million won were sold to the paper products sector in 1963; however, no supply to this sector was recorded in 1963. Besides these major markets, primary industries such as the fishery, coal and agriculture sectors purchase materials from forestry. 2. Analysis of repercussion effect on production The repercussion effect of final demand in any given sector upon the expansion of the production of other sectors was analyzed, using the inverse matrix coefficient tables attached to the the I.O. Table. A. Changes in intra-sector transaction value of inverse matrix coefficient. The intra-sector transaction value of an inverse matrix coefficient represents the extent of an induced increase in the production of self-support products of the same sector, when it is generated directly and indirectly by one unit of final demand in any given sector. The intra-sector transaction value of the forestry sector rose from 1.04 in 1963 to 1, 11 in 1966. It may well be said, therefore, that forestry induces much more self-supporting products in the production of one unit of final demand for forest products. B. Changes in column total of inverse matrix coefficient It should be noted that the column total indicates the degree of effect of the output of the corresponding and related sectors generated by one unit of final demand in each sector. No changes in the column total of the forestry sector were recorded between the 1963 and 1966 figures, both being the same 1. 19. C. Changes in difference between column total and intra-sector transaction amount. The difference between the column total and intra-sector transaction amount by sector reveals the extent of effect of output of related industrial sector induced indirectly by one unit of final demand in corresponding sector. This change in forestry dropped remarkable to 0.08 in 1966 from 0.15 in 1963. Accordingly, the effect of inducement of indirect output of other forestry-related sectors has decreased; this is a really natural phenomenon, as compared with an increasing input coefficient generated by the re-use of forest products by the forestry sector. 3. Induced output of forestry A. Forest products, wood in particular, are supplied to other industries as their raw materials, increasng their value added. In this connection the primary dependency rate on forestry for 1963 and 1966 was compared, i. e., an increase or decrease in each sector, from 7.71 percent in 1963 to 11.91 percent in 1966 in agriculture, 10.32 to 6.11 in fishery, 16.24 to 19.90 in mining, 0.76 to 0.70 in the manufacturing sector and 2.79 to 4.77 percent in the construction sector. Generally speaking, on the average the dependency on forestry during the period 1963-1966 increased from 5.92 percent to 8.03 percent. Accordingly, it may easily be known that the primary forestry output induced by primary and secondary industries increased from 16, 109 million won in 1963 to 48, 842 million won in 1966. B. The forest products are supplied to other industries as their raw materials. The products are processed further into higher quality products. thus indirectly increasing the value of the forest products. The ratio of the increased value added or the secondary dependency on forestry for 1963 and 1966 showed an increase or decrease, from 5.98 percent to 7.87 percent in agriculture, 9.06 to 5.74 in fishery, 13.56 to 15.81 in mining, 0.68 to 0.61 in the manufacturing sector and 2.71 to 4.54 in the construction sector. The average ratio in this connection increased from 4.69 percent to 5.60 percent. In the meantime, the secondary forestry output induced by primary and secondary industries rose from 12,779 million Wall in 1963 to 34,084 million won in 1966. C. The dependency of tertiary industries on forestry showed very minor ratios of 0.46 percent and 0.04 percent in 1963 and 1966 respectively. The forestry output induced by tertiary industry also decreased from 685 million won to 123 million won during the same period. D. Generally speaking, the ratio of dependency on forestry increased from 17.68 percent in 1963 to 24.28 percent in 1966 in primary industries, from 4.69 percent to 5.70 percent in secondary industries, while, as mentioned above, the ratio in the case of tertiary industry decreased from 0.46 to 0.04 percent during the period 1963-66. The mining industry reveals the heaviest rate of dependency on forestry with 29.80 percent in 1963 and 35.71 percent in 1966. As it result, the direct forestry income, valued at 8,172 million won in 1963, shot up to 22,724 million won in 1966. Its composition ratio lo the national income rose from 1.9 percent in 1963 to 2.3 per cent in 1966. If the induced outcome is taken into account, the total forestry production which was estimated at 37,744 million won in 1963 picked up to 105,773 million won in 1966, about 4.5 times its direct income. It is further noted that the ratio of the gross forestry product to the gross national product. rose significantly from 8.8 percent in 1963 to 10.7 percent in 1966. E. In computing the above mentioned ratio not taken into consideration were such intangible, indirect effects as the drought and flood prevention, check of soil run-off, watershed and land conservation, improvement of the people's recreational and emotional living, and maintenance and increase in the national health and sanitation. F. In conclusion, I would like to emphasize that the forestry sector exercices an important effect upon the national economy and that the effect of induced forestry output is greater than its direct income.

• Journal of Internet Computing and Services
• /
• v.15 no.3
• /
• pp.101-107
• /
• 2014

With the development of online service, recent forms of databases have been changed from static database structures to dynamic stream database structures. Previous data mining techniques have been used as tools of decision making such as establishment of marketing strategies and DNA analyses. However, the capability to analyze real-time data more quickly is necessary in the recent interesting areas such as sensor network, robotics, and artificial intelligence. Landmark window-based frequent pattern mining, one of the stream mining approaches, performs mining operations with respect to parts of databases or each transaction of them, instead of all the data. In this paper, we analyze and evaluate the techniques of the well-known landmark window-based frequent pattern mining algorithms, called Lossy counting and hMiner. When Lossy counting mines frequent patterns from a set of new transactions, it performs union operations between the previous and current mining results. hMiner, which is a state-of-the-art algorithm based on the landmark window model, conducts mining operations whenever a new transaction occurs. Since hMiner extracts frequent patterns as soon as a new transaction is entered, we can obtain the latest mining results reflecting real-time information. For this reason, such algorithms are also called online mining approaches. We evaluate and compare the performance of the primitive algorithm, Lossy counting and the latest one, hMiner. As the criteria of our performance analysis, we first consider algorithms' total runtime and average processing time per transaction. In addition, to compare the efficiency of storage structures between them, their maximum memory usage is also evaluated. Lastly, we show how stably the two algorithms conduct their mining works with respect to the databases that feature gradually increasing items. With respect to the evaluation results of mining time and transaction processing, hMiner has higher speed than that of Lossy counting. Since hMiner stores candidate frequent patterns in a hash method, it can directly access candidate frequent patterns. Meanwhile, Lossy counting stores them in a lattice manner; thus, it has to search for multiple nodes in order to access the candidate frequent patterns. On the other hand, hMiner shows worse performance than that of Lossy counting in terms of maximum memory usage. hMiner should have all of the information for candidate frequent patterns to store them to hash's buckets, while Lossy counting stores them, reducing their information by using the lattice method. Since the storage of Lossy counting can share items concurrently included in multiple patterns, its memory usage is more efficient than that of hMiner. However, hMiner presents better efficiency than that of Lossy counting with respect to scalability evaluation due to the following reasons. If the number of items is increased, shared items are decreased in contrast; thereby, Lossy counting's memory efficiency is weakened. Furthermore, if the number of transactions becomes higher, its pruning effect becomes worse. From the experimental results, we can determine that the landmark window-based frequent pattern mining algorithms are suitable for real-time systems although they require a significant amount of memory. Hence, we need to improve their data structures more efficiently in order to utilize them additionally in resource-constrained environments such as WSN(Wireless sensor network).

• Tunnel and Underground Space
• /
• v.13 no.1
• /
• pp.52-63
• /
• 2003

Discontinuities such as faults, fractures and joints in rock mass play the dominant role in the mechanical and hydraulic properties of the rock mass. The key factors that influence on the flow of groundwater are hydraulic and geometric characteristics of discontinuities and their connectivity. In this study, a program that analyzes groundwater flow in the 3D discontinuity network was developed on the assumption that the discontinuity characteristics such as density, trace length, orientation and aperture have particular distribution functions. This program generates discontinuities in a three-dimensional space and analyzes their connectivity and groundwater flow. Due to the limited computing capacity In this study, REV was not exactly determined, but it was inferred to be greater than 25$\times$25$\times$25 ㎥. By calculating the extent of aperture that influences on the groundwater flow, it was found that the discontinuities with the aperture smaller than 30% of the mean aperture had little influence on the groundwater flow. In addition, there was little difference in the equivalent hydraulic conductivity for the the two cases when considering and not considering the boundary effect. It was because the groundwater flow was mostly influenced by the discontinuities with large aperture. Among the parameters considered in this study, the length, aperture, and orientation of discontinuities had the greatest influence on the equivalent hydraulic conductivity of rock mass in their order. In case of existence of a fault in rock mass, elements of the equivalent hydraulic conductivity tensor parallel to the fault fairly increased in their magnitude but those perpendicular to the fault were increased in a very small amount at the first stage and then converged.

• Journal of Korean Society of Transportation
• /
• v.20 no.3
• /
• pp.149-158
• /
• 2002

For the purpose of preciously describing real time traffic pattern in urban road network, dynamic network loading(DNL) models able to simulate traffic behavior are required. A number of different methods are available, including macroscopic, microscopic dynamic network models, as well as analytical model. Equivalency minimization problem and Variation inequality problem are the analytical models, which include explicit mathematical travel cost function for describing traffic behaviors on the network. While microscopic simulation models move vehicles according to behavioral car-following and cell-transmission. However, DNL models embedding such travel time function have some limitations ; analytical model has lacking of describing traffic characteristics such as relations between flow and speed, between speed and density Microscopic simulation models are the most detailed and realistic, but they are difficult to calibrate and may not be the most practical tools for large-scale networks. To cope with such problems, this paper develops a new DNL model appropriate for dynamic traffic assignment(DTA), The model is combined with vertical queue model representing vehicles as vertical queues at the end of links. In order to compare and to assess the model, we use a contrived example network. From the numerical results, we found that the DNL model presented in the paper were able to describe traffic characteristics with reasonable amount of computing time. The model also showed good relationship between travel time and traffic flow and expressed the feature of backward turn at near capacity.

• Asia pacific journal of information systems
• /
• v.24 no.1
• /
• pp.31-50
• /
• 2014

The ubiquitous accessibility of information through mobile devices has led to an increased mobility of workers from their fixed workplaces. Market researchers estimate that by 2016, 350 million workers will be using their smartphones for business purposes, and the use of smartphones will offer new business benefits. Enterprises are now adopting mobile technologies for numerous applications to increase their operational efficiency, improve their responsiveness and competitiveness, and cultivate their innovativeness. For these reasons, various organizational aspects concerning "mobile work" have received a great deal of recent attention. Moreover, many CIOs plan to allocate a considerable amount of their budgets mobile work environments. In particular, with the consumerization of information technology, enterprise mobile applications (EMA) have played a significant role in the explosive growth of mobile computing in the workplace, and even in improving sales for firms in this field. EMA can be defined as mobile technologies and role-based applications, as companies design them for specific roles and functions in organizations. Technically, EMA can be defined as business enterprise systems, including critical business functions that enable users to access enterprise systems via wireless mobile devices, such as smartphones or tablets. Specifically, EMA enables employees to have greater access to real-time information, and provides them with simple features and functionalities that are easy for them to complete specific tasks. While the impact of EMA on organizational workers' productivity has been given considerable attention in various literatures, relatively little research effort has been made to examine how EMA actually lead to users' job performance. In particular, we have a limited understanding of what the key antecedents are of such an EMA usage outcome. In this paper, we focus on employees' perceived job performance as the outcome of EMA use, which indicates the successful role of EMA with regard to employees' tasks. Thus, to develop a deeper understanding of the relationship among EMA, its environment, and employees' perceived job performance, we develop a comprehensive model that considers the perceived-fit between EMA and employees' tasks, satisfaction on EMA, and the organizational environment. With this model, we try to examine EMA to explain how job performance through EMA is revealed from both the task-technology fit for EMA and satisfaction on EMA, while also considering the antecedent factors for these constructs. The objectives of this study are to address the following research questions: (1) How can employees successfully manage EMA in order to enhance their perceived job performance? (2) What internal and/or external factors are important antecedents in increasing EMA users' satisfaction on MES and task-technology fit for EMA? (3) What are the impacts of organizational (e.g. organizational agility), and task-related antecedents (e.g., task mobility) on task-technology fit for EMA? (4) What are the impacts of internal (e.g., self-efficacy) and external antecedents (e.g., system reputation) for the habitual use of EMA? Based on a survey from 254 actual employees who use EMA in their workplace across industries, our results indicate that task-technology fit for EMA and satisfaction on EMA are positively associated with job performance. We also identify task mobility, organizational agility, and system accessibility that are found to be positively associated with task-technology fit for EMA. Further, we find that external factor, such as the reputation of EMA, and internal factor, such as self-efficacy for EMA that are found to be positively associated with the satisfaction of EMA. The present findings enable researchers and practitioners to understand the role of EMA, which facilitates organizational workers' efficient work processes, as well as the importance of task-technology fit for EMA. Our model provides a new set of antecedents and consequence variables for a TAM involving mobile applications. The research model also provides empirical evidence that EMA are important mobile services that positively influence individuals' performance. Our findings suggest that perceived organizational agility and task mobility do have a significant influence on task-technology fit for EMA usage through positive beliefs about EMA, that self-efficacy and system reputation can also influence individuals' satisfaction on EMA, and that these factors are important contingent factors for the impact of system satisfaction and perceived job performance. Our findings can help managers gauge the impact of EMA in terms of its contribution to job performance. Our results provide an explanation as to why many firms have recently adopted EMA for efficient business processes and productivity support. Our findings additionally suggest that the cognitive fit between task and technology can be an important requirement for the productivity support of EMA. Further, our study findings can help managers in formulating their strategies and building organizational culture that can affect employees perceived job performance. Managers, thus, can tailor their dependence on EMA as high or low, depending on their task's characteristics, to maximize the job performance in the workplace. Overall, this study strengthens our knowledge regarding the impact of mobile applications in organizational contexts, technology acceptance and the role of task characteristics. To conclude, we hope that our research inspires future studies exploring digital productivity in the workplace and/or taking the role of EMA into account for employee job performance.

• Journal of Internet Computing and Services
• /
• v.18 no.1
• /
• pp.77-88
• /
• 2017

Recently, with the advent of knowledge based society where information and knowledge make values, patents which are the representative form of intellectual property have become important, and the number of the patents follows growing trends. Thus, it needs to classify the patents depending on the technological topic of the invention appropriately in order to use a vast amount of the patent information effectively. IPC (International Patent Classification) is widely used for this situation. Researches about IPC automatic classification have been studied using data mining and machine learning algorithms to improve current IPC classification task which categorizes patent documents by hand. However, most of the previous researches have focused on applying various existing machine learning methods to the patent documents rather than considering on the characteristics of the data or the structure of patent documents. In this paper, therefore, we propose to use two structural fields, technical field and background, considered as having impacts on the patent classification, where the two field are selected by applying of the characteristics of patent documents and the role of the structural fields. We also construct multi-label classification model to reflect what a patent document could have multiple IPCs. Furthermore, we propose a method to classify patent documents at the IPC subclass level comprised of 630 categories so that we investigate the possibility of applying the IPC multi-label classification model into the real field. The effect of structural fields of patent documents are examined using 564,793 registered patents in Korea, and 87.2% precision is obtained in the case of using title, abstract, claims, technical field and background. From this sequence, we verify that the technical field and background have an important role in improving the precision of IPC multi-label classification in IPC subclass level.