• The Journal of the Acoustical Society of Korea
• /
• v.21 no.3
• /
• pp.330-338
• /
• 2002

Spontaneous speech is ungrammatical as well as serious phonological variations, which make recognition extremely difficult, compared with read speech. In this paper, for conversational speech recognition, we analyze the transcriptions of the real conversational speech, and then classify the characteristics of conversational speech in the speech recognition aspect. Reflecting these features, we obtain the baseline system for conversational speech recognition. The classification consists of long duration of silence, disfluencies and phonological variations; each of them is classified with similar features. To deal with these characteristics, first, we update silence model and append a filled pause model, a garbage model; second, we append multiple phonetic transcriptions to lexicon for most frequent phonological variations. In our experiments, our baseline morpheme error rate (WER) is 31.65%; we obtain MER reductions such as 2.08% for silence and garbage model, 0.73% for filled pause model, and 0.73% for phonological variations. Finally, we obtain 27.92% MER for conversational speech recognition, which will be used as a baseline for further study.

• The Journal of Korean Association of Computer Education
• /
• v.15 no.3
• /
• pp.37-47
• /
• 2012

One of the major goals in high school Informatics is for students to develop creative problem-solving abilities based on knowledge on computer science. Thus, the contents of the textbooks should be accurate and appropriate. However, we discovered that the current Informatics textbooks contain the untrue and/or inappropriate descriptions of main memory and virtual memory. The textbooks describe that main memory is composed of RAM and ROM. The virtual memory is described as a technique in which a part of the secondary storage is utilized as main memory to execute an application of which size is larger than that of main memory. In this study, we attempted to uncover the root causes of the fallacies, and suggest the accurate explanations by comparing with renowned books adopted in most schools worldwide including USA. Our study reveals that it is inappropriate to include ROM in main memory from the memory hierarchy perspective. Virtual memory is a technique that provides convenience to programmers, through which an operating system loads the necessary portion of a program from secondary storage to main memory. As for the advantages of virtual memory in the current computer systems, the focus should be on providing the effective multitasking capability, rather than on executing a larger program than the size of main memory. We suggest that it is appropriate to exclude virtual memory in textbooks considering its complexity.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.10
• /
• pp.145-154
• /
• 2012

Recently, the cardiovascular disease has increased by causes such as westernization dietary life, smoking, and obesity. In particular, the acute myocardial infarction (AMI) occupies 50% death rate in cardiovascular disease. Following this trend, the AMI has been carried out a research for discovery of risk factors based on national data. However, there is a lack of diagnosis minor suitable for Korean. The objective of this paper is to develop a classifier for short-term relapse mortality prediction of cardiovascular disease patient based on prognosis data which is supported by KAMIR(Korea Acute Myocardial Infarction). Through this study, we came to a conclusion that ANN is the most suitable method for predicting the short-term relapse mortality of patients who have ST-segment elevation myocardial infarction. Also, data set obtained by logistic regression analysis performed highly efficient performance than existing data set. So, it is expect to contribute to prognosis estimation through proper classification of high-risk patients.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.31 no.1
• /
• pp.41-48
• /
• 2008

Enhanced machine reliability has dramatically reduced the rate and number of railway accidents but for further reduction human error should be considered together that accounts for about 20% of the accidents. Therefore, the objective of this study was to suggest a new taxonomy of performance shaping factors (PSFs) that could be utilized to identify the causes of a human error associated with railway accidents. Four categories of human factor, task factor, environment factor, and organization factor and 14 sub-categories of physical state, psychological state, knowledge/experience/ability, information/communication, regulation/procedure, specific character of task, infrastructure, device/MMI, working environment, external environment, education, direction/management, system/atmosphere, and welfare/opportunity along with 131 specific factors was suggested by carefully reviewing 8 representative published taxonomy of Casualty Analysis Methodology for Maritime Operations (CASMET), Cognitive Reliability and Error Analysis Method (CREAM), Human Factors Analysis and Classification System (HFACS), Integrated Safety Investigation Methodology (ISIM), Korea-Human Performance Enhancement System (K-HPES), Rail safety and Standards Board (RSSB), $TapRoot^{(R)}$, and Technique for Retrospective and Predictive Analysis of Cognitive Errors (TRACEr). Then these were applied to the case of the railway accident occurred between Komo and Kyungsan stations in 2003 for verification. Both cause decision chart and why-because tree were developed and modified to aid the analyst to find causal factors from the suggested taxonomy. The taxonomy was well suited so that eight causes were found to explain the driver's error in the accident. The taxonomy of PSFs suggested in this study could cover from latent factors to direct causes of human errors related with railway accidents with systematic categorization.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1998.05b
• /
• pp.305-311
• /
• 1998

최근들어 고려된 LBB(Leak Before Break) 적용요건중 동적파괴시힘 절차에는 울진 3&4호기 이후 파단전누설개념이 적용되는 배관이 탄소강으로 제작될 경우. 이 배관이 Dynamic Strain Aging (DSA)에 의해 파괴저항치가 감소되지 않는다는 것이 정량적으로 입증되지 않는 한, 동 배관의 파괴 물성치 결정시 DSA의 영향이 고려되어야 하며, DSA 영향을 평가하기 위해서는 동적과괴시험이 수행되어야 함을 요건화 하고 있다. 본 연구에서는 DSA 효과에 의한 파괴저항(J-R) 특성의 저하가차세대원전 원자로냉각재배관 파단전누설개넘(LBB) 적용시 설계 안전여유도에 영향을 미치지 않는 정도임을 평가하는데 있다. 따라서 ASME Section III에서 탄소강으로 분류하고 있는 강종별 파괴인성 변화를 고찰하고, 차세대원전 주냉각재배관 재료인 SA508 Class la의 최대 파괴인성 감소치를 예측하여, 울진 3&4호기에서 측정된 엘보우용 SA516-Gr.70 강의 DSA 영향 평가 결과와 비교 분석하여 차세대원전 주냉각재배관의 DSA영향을 평가하였다. 도출된 결론으로는 DSA 영향을 고려한 SA508 Class la의 J 및 dJ/dA 값은 극히 보수적으로 추정할 때 50% 이상 감소하는 것으로 예측된다. 이러한 DSA 영향을 고려하였을 경우 배관재 모재의 파괴인성치는 Weld-SAW의 J/T 값 수준으로 감소하였다. 그러나 현 LRB 해석이 가장 낮은 J/T값을 갖는 Weld-SAW Auto의 균열길이 2a인 J/T선도에 의거하여 수행되고 있다는 점을 고려한다면 비록 DSA가 배관재에 영향을 주는 가장 보수적인 값(J 및 dJ/dA값을 50% 이상)을 사용한다고 하더라도 차세대원전 LBB 적용에 문제가 되지 않음을 알 수 있다. 즉 차세대원자로 주냉각재배관에 LBB를 적용하는데는 DSA 영향은 상대적으로 중요하지 않다는 결론을 얻었다. 표면에 수소화물이 농축되어 있는 hydride layer가 형성됨을 관찰하였으며 ～5,000ppm 이상의 경우에는 수소화물의 방향성이 random하였으며 특히, ZIRLO$^{TM}$ 시편의 경우에서는 원주방향으로 길게 이어진 수소화물과 기계적 성질에 치명적인 반경방향의 수소화물이 평행하게 배열된 것을 관찰하였다.하였을 때는 Li$_2$O의 첨가에 의해 치밀화가 주로 일어났고, 반면에 $N_2$-7vol.%H$_2$ 분위기에서 소결하면 Li$_2$O의 첨가에 의해 작은 기공은 소멸되고 큰 기공이 생성되었다.지나치게 모국어의 영향만 강조하고 다른 요인들에 대해서는 다분히 추상적인 언급으로 끝났지만 이 분석을 통 해서 배경어, 목표어, 특히 중간규칙의 역할이 괄목할 만한 것임을 가시적으로 관찰할 수 있 다. 이와 같은 오류분석 방법은 학습자의 모국어 및 관련 외국어의 음운규칙만 알면 어느 학습대상 외국어에라도 적용할 수 있는 보편성을 지니는 것으로 사료된다.없다. 그렇다면 겹의문사를 [-wh]의리를 지 닌 의문사의 병렬로 분석할 수 없다. 예를 들어 누구누구를 [주구-이-ν가] [누구누구-이- ν가]로부터 생성되었다고 볼 수 없다. 그러므로 [-wh] 겹의문사는 복수 의미를 지닐 수 없 다. 그러면 단수 의미는 어떻게 생성되는가\ulcorner 본 논문에서는 표면적 형태에도 불구하고 [-wh]의미의 겹의문사는 병렬적 관계의 합성어가 아니라 내부구조를 지니지 않은 단순한 단어(minimal $X^{0}$ elements)로 가정한다. 즉, [+wh] 의미의 겹의문사는 동일한 구성요 소를 지닌 병렬적 합성어([$[W1]_{XO-}$ $[W1]_{XO}$ ]$_{XO}$)로

• Journal of Internet Computing and Services
• /
• v.18 no.1
• /
• pp.65-75
• /
• 2017

The Detection Model is the model to find the result of a certain purpose using artificial intelligent, data mining, intelligent algorithms In Cyber Security, it usually uses to detect intrusion, malwares, cyber incident, and attacks etc. There are an amount of unlabeled data that are collected in a real environment such as security data. Since the most of data are not defined the class labels, it is difficult to know type of data. Therefore, the label determination process is required to detect and analysis with accuracy. In this paper, we proposed a KDFL(K-means and D-S Fusion based Labeling) method using D-S inference and k-means(unsupervised) algorithms to decide label of data records by fusion, and a detection model architecture using a proposed labeling method. A proposed method has shown better performance on detection rate, accuracy, F1-measure index than other methods. In addition, since it has shown the improved results in error rate, we have verified good performance of our proposed method.

• Korean Journal of Cognitive Science
• /
• v.7 no.1
• /
• pp.57-74
• /
• 1996

This paper presents a comparison of recognition performance of the features used inthe recent handwritten korean character recognition.This research aims at providing the basis for feature selecion in order to improve not only the recognition rate but also the efficiency of recognition system.For the comparison of feature performace,we analyzed the characteristics of theose features and then,classified them into three rypes:global feature(image transformation)type,statistical feature type,and local/ topological feature type.For each type,we selected four or five features which seem more suitable to represent the characteristics of korean alphabet,and performed recongition experiments for the first consonant,horizontal vowel,and vertical vowel of a korean character, respectively.The classifier used in our experiments is a multi-layered perceptron with one hidden layer which is trained with backpropagation algorithm.The training and test data in the experiment are taken from 30sets of PE92. Experimental results show that 1)local/topological features outperform the other two type features in terms of recognition rates 2)mesh and projection features in statical feature type,walsh and DCT features in global feature type,and gradient and concavity features in local/topological feature type outperform the others in each type, respectively.

• Journal of Advanced Marine Engineering and Technology
• /
• v.40 no.6
• /
• pp.510-515
• /
• 2016

Recently, corrosion occurred on the piles of a marine bridge located on the NamHae expressway in Korea. A sacrificial anode cathodic protection system was installed to prevent corrosion damage in the marine bridge. In the case of the marine bridge in this study, the sacrificial anode cathodic protection system was applied at the tidal and splash zones of the piles because the upper part of the structure was not corroded, and because corrosion occurs at the tidal and splash zones due to sea tides. To verify the performance of the sacrificial anode cathodic protection system 8 years later, cathodic protection (CP) current, CP potential, and degree of depolarization were measured. The experimental results on the performance of the sacrificial anode cathodic protection system from a total of 60 piles were classified into 4 categories: good CP effect results (13 piles), partial CP effect results (27 piles), temporarily erroneous results (5 piles), and need for maintenance because of delamination (15 piles). It was determined that additional repairs are required, such as the application of bulk anodes and jacket casings, for piles where the CP effect is poor.

• Journal of KIISE:Software and Applications
• /
• v.33 no.9
• /
• pp.787-801
• /
• 2006

Ontology merging describes the process of integrating two ontologies into a new ontology. How this is done best is a subject of ongoing research in the Semantic Web, Data Integration, Knowledge Management System, and other ontology-related application systems. Earlier research on ontology merging, however, has studied for developing effective ontology matching approaches but missed analyzing and solving methods of problems of merging two ontologies given correspondences between them. In this paper, we propose a specific ontology merging process and a generic operator, T-MERGE, for integrating two source ontologies into a new ontology. Also, we define a taxonomy of merging conflicts which is derived from differing representations between input ontologies and a method for detecting and resolving them. Our T-MERGE operator encapsulates the process of detection and resolution of conflicts and merging two entities based on given correspondences between them. We define a data structure, MergeLog, for logging the execution of T-MERGE operator. MergeLog is used to inform detailed results of execution of merging to users or recover errors. For our experiments, we used oriental philosophy ontologies, western philosophy ontologies, Yahoo western philosophy dictionary, and Naver philosophy dictionary as input ontologies. Our experiments show that the automatic merging module compared with manual merging by a expert has advantages in terms of time and effort.

• Korean Journal of Cognitive Science
• /
• v.26 no.3
• /
• pp.301-322
• /
• 2015

This study examined the effect of emotional sounds on satisfaction of search (SOS). SOS occurs when detection of a target results in a lesser chance of finding subsequent targets when searching for an unknown number of targets. Previous studies have examined factors that may influence the phenomenon, but the effect of emotional sounds is yet to be identified. Therefore, the current study investigated how emotional sound affects magnitude of the SOS effect. In addition, participants' eye movements were recorded to determine the source of SOS errors. The search display included abstract T and L-shaped items on a cloudy background and positive and negative sounds. Results demonstrated that negative sounds produced the largest SOS effect by definition, but this was due to superior accuracy in low-salient single target trials. Response time, which represents efficiency, was consistently faster when negative sounds were provided, in all target conditions. On-target fixation classification revealed scanning error, which occurs because targets are not fixated, as the most prominent type of error. These results imply that the two dimensions of emotion - valence and arousal - interactively affect cognitive performance.