• Journal of the Korea Convergence Society
• /
• v.11 no.11
• /
• pp.33-39
• /
• 2020

In the field of Natural Language Processing, Various research such as Translation, POS Tagging, Q&A, and Sentiment Analysis are globally being carried out. Sentiment Analysis shows high classification performance for English single-domain datasets by pretrained sentence embedding models. In this thesis, the classification performance is compared by Korean E-commerce online dataset with various domain attributes and 6 Neural-Net models are built as BOW (Bag Of Word), LSTM[1], Attention, CNN[2], ELMo[3], and BERT(KoBERT)[4]. It has been confirmed that the performance of pretrained sentence embedding models are higher than word embedding models. In addition, practical Neural-Net model composition is proposed after comparing classification performance on dataset with 17 categories. Furthermore, the way of compressing sentence embedding model is mentioned as future work, considering inference time against model capacity on real-time service.

• Journal of Korean Society of Transportation
• /
• v.29 no.2
• /
• pp.15-24
• /
• 2011

This study discusses an equity analysis performed to supplement the existing bus service evaluation system. The equity of 179 feeder bus routes in the city of Seoul(Year 2009) is analyzed by employing the modified BCC model along with the desirable inputs such as a number of bus stops and a frequency of operations. In the analysis, each DMU (Decision Making Unit) is assumed to use the desirable inputs, and produces both the index of equity and the index of accessibility. We apply an input-oriented modified BCC model to analyze the equity of the DMUs. The computational result indicates that only 9% of the feeder bus routes are equitable among the total DMUs and the average equity estimate is 0.702. It is also suggested that on average approximately 60% increases of desirable inputs (i.e., a number of bus stops and a frequency of operations) are needed in order to improve equity of DMU in the city of Seoul. The result of the tobit regression analysis on the determinant factors affecting the equity of DMUs shows that all the independent variables (i.e., a number of bus stops and a frequency of operations) are statistically significant at 5% significance level, and the number of bus stops is found to be the most influential factor.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.11B
• /
• pp.1624-1630
• /
• 2010

This paper proposes a method and its implementation for searching, converting, and delivering composite events and actions in the U-City Integrated Management Platform to process them efficiently through allocation and collaboration between the platform and u-services. The purpose of this method is to, when a composite event occurs, execute prompt and correct measures to reduce the damage caused by the event and prevent its spread. The proposed method also helps to simplify the complexity of corresponding business processes by modifying composite events and actions that are frequently added, delete, and revised, and delivering them to u-service systems.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.1
• /
• pp.68-79
• /
• 2011

This paper focuses on the opinion information extraction and analysis system through Web mining that is based on statistics collected from Web contents. That is, users' opinion information which is scattered across several websites can be automatically analyzed and extracted. The system provides the opinion information search service that enables users to search for real-time positive and negative opinions and check their statistics. Also, users can do real-time search and monitoring about other opinion information by putting keywords in the system. Proposing technique proved that the actual performance is excellent by comparison experiment with other techniques. Performance evaluation of function extracting positive/negative opinion information, the performance evaluation applying dynamic window technique and tokenizer technique for multilingual information retrieval, and the performance evaluation of technique extracting exact multilingual phonetic translation are carried out. The experiment with typical movie review sentence and Wikipedia experiment data as object as that applying example is carried out and the result is analyzed.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2001.10a
• /
• pp.97-100
• /
• 2001

For the auto-configuration of diverse hosts and automation of network management as many electronic machines and industrial components are becoming connected to Internet, it is possible to use autoconfiguration using DHCP and network management protocol such as SNMP, CMIP. But these technologies are not developed maturely and lack various standardizations yet because these are designed for restrictive purposes in the past. The new working group "Zeroconf" is made in IETF at March, 2000, it is aimed for making standards of autoconfiguration which is applied to impromptu ad-hoc networks in automobiles, trains, planes and home network or mobile networks. In this paper, we introduce the IPv6 based autoconfiguration technologies such as IP interface configuration, translation between host name and IP address, IP multicast address allocation and service discovery. Finally, trends of standardization in Zeroconf working group will be shown. be shown.

• Journal of Digital Contents Society
• /
• v.12 no.3
• /
• pp.279-289
• /
• 2011

The latest hardware and software techniques are combined to give safe walking guidance and convenient service of realtime walking assistance system for visually impaired person. This system consists of obstacle detection and perception, place recognition, and sign recognition for pedestrian can safely walking to arrive at their destination. In this paper, we exploit the sign object detection system in subway station for sign recognition that one of the important factors of walking assistance system. This paper suggest the adaptive feature map that can be robustly extract the sign object region from complexed environment with light and noise. And recognize a sign using fast zernike moment features which is invariant under translation, rotation and scale of object during walking. We considered three types of signs as arrow, restroom, and exit number and perform the training and recognizing steps through adaboost classifier. The experimental results prove that our method can be suitable and stable for real-time system through yields on the average 87.16% stable detection rate and 20 frame/sec of operation time for three types of signs in 5000 images of sign database.

• Journal of KIISE:Computing Practices and Letters
• /
• v.11 no.2
• /
• pp.180-191
• /
• 2005

The rapid growth of mobile communication technology has provided the expansion of mobile internet services, particularly mobile commerce takes much weight among them. Even though current mobile commerce service has serious problems which check its development, such as limited contents, expensive charge system and hardware restriction of mobile device, it is strongly expected as one of the next generation Internet services. In this paper, we summarize the problems like above and provide some total solution to meet them as follows: a function for automatic gathering of product information on online Internet and automatic translation it to data for mobile commerce, a middlelet application which provides functions for product search and order on the mobile device through off-line processing, and a function of user adaptive recommendation. We have actually designed and implemented the proposed system and verified the functions and effectiveness of the system.

• Journal of Korean Society of Transportation
• /
• v.28 no.5
• /
• pp.43-54
• /
• 2010

In order to improve the existing evaluation system of bus services and gain more reasonable analysis outputs, the authors evaluate the efficiency of 113 arterial bus routes in Seoul in 2009 using a modified BCC model considering not only desirable outputs but also undesirable outputs. Each Decision Making Unit (DMU) is assumed to use inputs such as possession costs, operating costs, the ratios of median bus stops overlapped route lengths to produce estimates of desirable outputs (the number of passengers and service satisfaction score) and undesirable outputs (CO2 emissions). According to the analysis, the modified BCC model considering both desirable outputs and undesirable outputs shows more appropriate results. DMUs would be more efficient on average to reduce nearly 10% of the 3 inputs (possession costs, operating costs, and overlapped route lengths) and increase by about 160% the ratios of median bus stops. Also, a Tobit regression analysis is conducted to identify the most effective variables for maximum efficiency and discover that the variable of possession costs and the ratios of median bus stops are statistically significant.

• Journal of KIISE:Computing Practices and Letters
• /
• v.5 no.6
• /
• pp.801-808
• /
• 1999

지금까지, 전기통신망과 장비들의 관리에 대해서는 많은 연구들이 진행되었지만, TMN 상위 계층의 서비스 관리와 비즈니스 관리에 대한 연구는 많이 수행되지 못했다. 최근 전세계적으로 TMN 상위 계층의 관리기능 구현을 위해서 CORBA 기술을 적용하고자하는 관심이 증가하고 있다. 본 논문에서는, 정보모델과 상호운용에 관한 TMN과 CORBA의 주요특징들을 비교하고, TMN 프레임웍와 효율적인 CORBA 기술 연동을 위한 플랫폼 구조를 제안한다. 제안된 구조에서, TMN 관리 정보를 CORBA IDL 인터페이스로 변환해주는 GDMO/ASN.1 - CORBA IDL 변환기가 고안, 구현되었고, JIDM의 상호 작용 변환의 구현을 위해서 TMN/CORBA 게이트웨이 또한 고안되었다. 그리고 이러한 기능의 확장을 위해서 모듈 형식으로 라이브러리들과 코드 생성기들이 설계, 구현되었다.Abstract Up to now, a lot of efforts have been made for the management of telecommunication networks and equipment, but less effort has been made for the realization of higher-layer service and business management. There are recently worldwide growing interests for applying CORBA technology for the realization of higher layer Telecommunication Management Network (TMN) management functions. In this paper, we compare the features of TMN and CORBA with respect to information model, and operational behavior. And then, we propose a platform architecture for the efficient integration of CORBA technology within TMN framework. Within platform architecture, GDMO/ASN.1 to IDL translator has been designed and implemented for translating TMN management information into CORBA IDL interface. The TMN/CORBA gateway is also designed for realization of the interaction translation specification of JIDM task force. The libraries and code generators have been designed and implemented in a modular fashion to provide the functional extensibility.

• Journal of Korean Academy of Dental Administration
• /
• v.10 no.1
• /
• pp.53-65
• /
• 2022

The Primary Care Patient Measure of Safety (PC-PMOS) is a measure of safety that allows patients to evaluate contributing factors related to patient safety in primary care. This study aimed to examine which questionnaire items of the PC-PMOS could be used in Korean dental institutions. A survey of the Korean translation of the PC-PMOS, a self-reported questionnaire, was conducted in August 2022 by 400 adults who had used dental-care services within the last three years. Of the 77 items, 34 were selected using principal component analysis and two experts' evaluations based on face validity. Five domains were identified from factor analysis: patient centeredness, patient information update, complaint processing, communication, and information about the complaint process. The Cronbach's alpha of the data was 0.913, indicating high reliability. As a result of the generalized multiple regression analysis, regression coefficients were not statistically significant, except for household income. This indicated that there was no bias in the patient safety scores of dental institutions evaluated by patients within the range of independent variables used in this study. The five domains with 34 items identified in this study suggested the factors that contribute to the safety of patients who used dental care services in Korea. However, validation of this study result is still important to refine questionnaires suitable for dental institutions in Korea so as to further improve the quality of dental care.