• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.504-509
• /
• 2018

Neural machine translation (NMT) has recently achieved the state-of-the-art performance. However, it is reported failing in the word sense disambiguation (WSD) for several popular language pairs. In this paper, we explore the extent to which NMT systems are able to disambiguate the Korean homographs. Homographs, words with different meanings but the same written form, cause the word choice problems for NMT systems. Consistent with the popular language pairs, we discover that NMT systems fail to translate Korean homographs correctly. We provide a Korean word sense disambiguation tool-UTagger to use for improvement of NMT's translation quality. We conducted translation experiments using Korean-English and Korean-Vietnamese language pairs. The experimental results show that UTagger can significantly improve the translation quality of NMT in terms of the BLEU, TER, and DLRATIO evaluation metrics.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.795-798
• /
• 1997

• Proceedings of the Korean Nuclear Society Conference
• /
• 1997.05a
• /
• pp.219-226
• /
• 1997

This paper presents the development of an Integrated Test Facility (ITF) that consists of a full-scope Human Machine Simulator (HMS) and some experimental measurement systems. This facility is aimed to experiment with tile design of the control room environment and the human interaction as it relates to the control of advanced nuclear power plants.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.2471-2474
• /
• 2003

This paper focuses on the stability and operability of a man-machine system considering a human operator. Some papers' main interest has been the stability only, but the operability such as fatigue is also the other main interest. In a man-machine system, feelings such as motional, visual, and kinesthetic are important since those enable operators to work easily or fatigue operators. A model of a man-machine system has been developed. Motional, visual, and kinesthetic feelings may be considered as feedbacked sensor signals. We also have quantified the degree of fatigue with respect to reference operation. This is a performance index to be optimized. Several methods are presented to optimize the degree of fatigue and the stability of the integrated system. Examples are presented to show that the usefulness of the proposed modeling method and fatigue mitigating algorithm.

• Nuclear Engineering and Technology
• /
• v.55 no.2
• /
• pp.603-622
• /
• 2023

Human error (HE) is an important concern in safety-critical systems such as nuclear power plants (NPPs). HE has played a role in many accidents and outage incidents in NPPs. Despite the increased automation in NPPs, HE remains unavoidable. Hence, the need for HE detection is as important as HE prevention efforts. In NPPs, HE is rather rare. Hence, anomaly detection, a widely used machine learning technique for detecting rare anomalous instances, can be repurposed to detect potential HE. In this study, we develop an unsupervised anomaly detection technique based on generative adversarial networks (GANs) to detect anomalies in manually collected surveillance data in NPPs. More specifically, our GAN is trained to detect mismatches between automatically recorded sensor data and manually collected surveillance data, and hence, identify anomalous instances that can be attributed to HE. We test our GAN on both a real-world dataset and an external dataset obtained from a testbed, and we benchmark our results against state-of-the-art unsupervised anomaly detection algorithms, including one-class support vector machine and isolation forest. Our results show that the proposed GAN provides improved anomaly detection performance. Our study is promising for the future development of artificial intelligence based HE detection systems.

• Annual Conference on Human and Language Technology
• /
• 1997.10a
• /
• pp.243-248
• /
• 1997

본 논문은 다국어 자동번역시스템에서 다국어의 번역지식(사전,규칙,정보)구축을 위해 언어유형론을 도입하는 방법론을 제안한다. 다국어 번역지식의 구축과 관련하여 기존 다국어 자동번역 시스템들에서 항상 문제가 되고 있는 것은 번역지식의 구축, 관리, 재활용의 문제이다. 번역지식의 구축은 다국어를 위한 번역지식의 크기, 다국어의 수용정도와 관련되며, 번역지식의 관리는 번역지식의 단순화 정도와 관련되며, 번역지식의 재활용은 기존에 구축된 번역지식을 새로운 언어들에 재사용 정도와 관련된다. 이러한 문제점들을 해결하기 위해 본 논문에서는 한국어를 포함한 다국어의 언어 친족성에 따라 번역지식을 공유하도록 하는 언어유형론에 기반한 다국어 공용 번역지식 구축 방법론을 제안하고자 한다.

• International Journal of Computer Science & Network Security
• /
• v.22 no.1
• /
• pp.129-138
• /
• 2022

With the increasing reliance of computing systems in our everyday life, there is always a constant need to improve the ways users can interact with such systems in a more natural, effective, and convenient way. In the initial computing revolution, the interaction between the humans and machines have been limited. The machines were not necessarily meant to be intelligent. This begged for the need to develop systems that could automatically identify and interpret our actions. Automatic gesture recognition is one of the popular methods users can control systems with their gestures. This includes various kinds of tracking including the whole body, hands, head, face, etc. We also touch upon a different line of work including Brain-Computer Interface (BCI), Electromyography (EMG) as potential additions to the gesture recognition regime. In this work, we present an overview of several applications of automated gesture recognition systems and a brief look at the popular methods employed.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1553-1558
• /
• 2003

The authors investigate influences of manipulator dynamics on and roles of kinesthetic sensation and visual sensation in a Human-Machine Cooperative System (HMCS). At first, the general structure and essential transfer functions of HMCSs are described based on the previous work. Then, after showing theoretical treatment of manipulator dynamics, this paper analyzes the influences on HMCSs in two cases: one is the control design focusing on tool dynamics and reaction force transfer function, and the other is that specifies maneuver transfer function and transfer function for object dynamics variation. In addition to conventional experiments only employing kinesthetic sensation, other experiments with both kinesthetic and visual sensations are performed to examine difference in the roles of these sensations and the validity of the design without the visual sensation.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.55.3-55
• /
• 2001

The implementation of human-like robot has been advanced in various parts such as mechanic arms, legs, and applications of five senses. The vision applications have been developed in several decades and especially the face recognition have become a prominent issue. In addition, the development of computer systems makes it possible to process complex algorithms in realtime. The most of human recognition systems adopt the discerning method using fingerprint, iris, and etc. These methods restrict the motion of the person to be discriminated. Recently, the researchers of human recognition systems are interested in facial recognition by using machine vision. Thus, the object of this paper is the implementation of the realtime ...

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.5 no.6
• /
• pp.33-39
• /
• 1982

In the design of some systems, the circumstance may require a determination as to whether a giver information input or processing function can best be performed by an individual or by some physical components of a system. In the consideration of human factors, also, in the design of equipment, facilities, and other physical items that people use, there are certain basic stages or processes that typically have to be carried out. The main purpose of this paper is study fundamental man-machine system assumptions in Human Factors Engineering. Therefore, this paper will deal with certain human factors that may be relevant to the design consideration that relate to human information - receiving and processing functions.