• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2018.05a
• /
• pp.166-168
• /
• 2018

해기사의 행동오류는 해양사고를 야기하는 최종 결과물이다. 해기사의 행동오류는 SRK(Skill, Rule, Knowledge) 행동이론을 통해서 규명할 수 있고, 이러한 행동오류는 해양사고 기록을 통해서 식별할 수 있다. 본 발표의 목적은 과거 해양사고 기록을 토대로 해기사들의 다양한 행동오류와 이들 행동오류로 인하여 발생한 해양사고를 연계하여 해기사들의 행도오류를 평가하기 위한 방법과 현재까지의 결과를 소개하는데 있다. 라스무센이 제한한 행동이론에 해양사고 기록을 적용하여 해양사고 발생 당시의 해기사들의 행동오류를 식별할 수 있었다. 식별한 변수를 이용하여 해기사들의 행동오류를 추정한 결과, 사고원인과 결과 및 행동오류 사이에 강한 상관관계를 나타냈다. 추후, HMM(Hidden Markov Model)을 이용한 예측 기법을 적용하여 행동오류에 따른 해양사고 예방 방안을 제안할 예정이다.

• Progress in Medical Physics
• /
• v.10 no.2
• /
• pp.95-101
• /
• 1999

The aims of this report are to classify the incorrect data of patients and the errors of dose and dose distribution observed in QA activities on teleradiotherapy chart, and to analyze their frequency. In our department, radiation physicists check several sheets of patient chart to reduce numeric errors before starting radiation therapy and at least once a week, which include history, port diagram, MU calculation or treatment planning summary and daily treatment sheet. The observed errors are classified as followings. 1) Identity of patient, 2) Omitted or unrecorded history sheet even though not including the item related to dose, 3) Omission of port diagram, or omitted or erroneous data, 4) Erroneous calculation of MU and point dose, and important causes, 5) Loss of summary sheet of treatment planning, and erroneous data of patient in the sheet, 6) Erroneous record of radiation therapy, and errors of daily dose, port setup, MU and accumulated dose in the daily treatment sheet, 7) Errors leading inexact dose or dose distribution, errors not administerd even though its possibility, and simply recorded errors, 8) Omission of sign. Number of errors was counted rather than the number of patients. In radiotherapy chart QA from Jun 17, 1996 to Jul 31, 1999, no error of patient identity had been observed. 431 Errors in 399 patient charts had been observed and there were 405 physical errors, 9 cases of omitted or unrecorded history sheet, and 17 unsigned. There were 23 cases (5.7%) of omitted port diagram, 21 cases (5.2%) of omitted data and 73 cases (18.0 %) of erroneous data in port diagram, 13 cases (3.2 %) treated without MU calculation, 68 cases (16.3 %) of erroneous MU, 8 cases (2.0%) of erroneous point dose, 1 case (0.2 %) of omitted treatment planning summary, 11 cases (2.7%) of erroneous input of patient data, 13 cases (3.2%) of uncorrected record of treatment, 20 cases (4.9%) of discordant daily doses in MU calculation sheet and daily treatment sheet, 33 cases (8.1%) of erroneous setup, 52 cases (12.8%) of MU setting error, 61 cases (15.1%) of erroneous accumulated dose. Cases of error leading inexact dose or dose distribution were 239 (59.0 %), cases of error not administered even though its possibility were 142 (35.1 %), and cases of simply recorded error were 24 (5.9 %). The numeric errors observed in radiotherapy chart ranged over various items. Because errors observed can actually contribute to erroneous dose or dose distribution, or have the possibility to lead such errors, thorough QA activity in physical aspects of radiotherapy charts is required.

• Journal of Intelligence and Information Systems
• /
• v.29 no.3
• /
• pp.267-286
• /
• 2023

The conversational agents such as AI speakers utilize voice conversation for human-computer interaction. Voice recognition errors often occur in conversational situations. Recognition errors in user utterance records can be categorized into two types. The first type is misrecognition errors, where the agent fails to recognize the user's speech entirely. The second type is misinterpretation errors, where the user's speech is recognized and services are provided, but the interpretation differs from the user's intention. Among these, misinterpretation errors require separate error detection as they are recorded as successful service interactions. In this study, various text separation methods were applied to detect misinterpretation. For each of these text separation methods, the similarity of consecutive speech pairs using word embedding and document embedding techniques, which convert words and documents into vectors. This approach goes beyond simple word-based similarity calculation to explore a new method for detecting misinterpretation errors. The research method involved utilizing real user utterance records to train and develop a detection model by applying patterns of misinterpretation error causes. The results revealed that the most significant analysis result was obtained through initial consonant extraction for detecting misinterpretation errors caused by the use of unregistered neologisms. Through comparison with other separation methods, different error types could be observed. This study has two main implications. First, for misinterpretation errors that are difficult to detect due to lack of recognition, the study proposed diverse text separation methods and found a novel method that improved performance remarkably. Second, if this is applied to conversational agents or voice recognition services requiring neologism detection, patterns of errors occurring from the voice recognition stage can be specified. The study proposed and verified that even if not categorized as errors, services can be provided according to user-desired results.

• Journal of Korean Society of Archives and Records Management
• /
• v.20 no.3
• /
• pp.1-21
• /
• 2020

Although the National Archives of Korea has been receiving transfer of standard electronic documents with a retention period of more than 30 years from the central administration since 2015, errors and inefficiencies persist. Despite this, however, there remain no studies that analyze or address these issues. This study aims to bridge this gap and found that most errors in the transfer process occur at the production stage. To address such, this study proposes a four-step solution. First, before electronic document approval, the text and attached files are checked for defects to prevent errors. Second, as soon as the signature is made, digital signatures are applied on a file-by-file basis to ensure integrity. Third, integrity verification and transfer inspection are automatically performed through digital signature investigation and defect check procedure during transfer and preservation. Fourth and last, the criteria of acquiring records and integrity guarantee technologies are properly applied in production stage with proper management and supervision.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.633-635
• /
• 2022

If an error occurs during operation of the video wall controller, the control system creates a log file and records the log. In order to minimize the load on the system due to log recording, the log level is set so that the log is not recorded as much as possible under normal operating conditions. When an error occurs, detailed logs are recorded by changing the log level to analyze and respond to the cause of the error. So work efficiency is reduced, and operator intervention is inevitable to change the log level. In this paper, we propose a model that automatically sets the log level according to the operating situation using Contextual Bandit.

• Proceedings of the Optical Society of Korea Conference
• /
• 2007.02a
• /
• pp.275-276
• /
• 2007

본 보고서에서는 삼진 연속 영 제한 부호를 광 디스크에 채용되게 하기 위해 여러 광 신호처리 기법의 변화을 소개한다. 삼진 광 정보의 기록 재생을 위해 광 기록 재생 시스템은 파형 등화법과 멀티 펄스 기록 보정법은 물론 적합한 오류 정정법을 필요로 하고 있다.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.382-385
• /
• 2018

Dead Link의 노출 최소화는 웹 검색 서비스의 품질 유지에 있어 매우 중요하다. 따라서 색인 내 Soft 404 오류의 정확한 판별은 필수적이지만, 리다이렉션 정보에 의존하거나 텍스트 혹은 HTML 자질 만을 고려하는 기존 방법의 활용만으로는 판별 가능한 Soft 404 오류의 유형이 한정될 수 있다는 문제가 있다. 이에 본 연구에서는 보다 범용성이 높은 Soft 404 오류 판별 기술의 개발을 위해, 404 오류 안내 페이지 고유의 형태적 특성을 오류 판별에 사용할 것을 제안한다. 제안 방법은 오류 안내 문서의 형태적 특성을 이미지 인식 모형에 기반해 학습한 후 이를 Soft 404 오류 판별에 사용하며, 리다이렉션 등 특정 정보에 의존하는 기존 방법에 비해 보다 폭넓게 적용 가능하다는 장점이 있다. 실험에서 제안 방법은 87.6%의 정확률과 92.7%의 재현율을 기록하는 등 높은 인식 성능을 보였다.

• The Korean Journal of Archival Studies
• /
• no.70
• /
• pp.85-117
• /
• 2021

The purpose of this study is to investigate and analyze the current status of unit tasks, unit task operation, and record management problems used by local governments, and to present improvement measures using text-based big data technology based on the implications derived from the process. Local governments are in a serious state of record management operation due to errors in preservation period due to misclassification of unit tasks, inability to identify types of overcommon and institutional affairs, errors in unit tasks, errors in name, referenceable standards, and tools. However, the number of unit tasks is about 720,000, which cannot be effectively controlled due to excessive quantities, and thus strict and controllable tools and standards are needed. In order to solve these problems, this study developed a system that applies text-based analysis tools such as corpus and tokenization technology during big data analysis, and applied them to the names and construction terms constituting the record management standard. These unit task operation support tools are expected to contribute significantly to record management tasks as they can support standard operability such as uniform preservation period, identification of delegated office records, control of duplicate and similar unit task creation, and common tasks. Therefore, if the big data analysis methodology can be linked to BRM and RMS in the future, it is expected that the quality of the record management standard work will increase.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.04a
• /
• pp.1047-1048
• /
• 2009

인공위성은 전자와 양자들로 인한 방사현상, 위치에 따른 온도변화 및 진공에 의한 영향 등 극한 우주 환경에서 동작하므로 항상 위험에 노출되어 있다. 특히 저궤도 관측위성은 지구와의 통신이 제한적이기 때문에 위성에 문제가 발생할 경우 신속한 대응이 어려워 질 수 있다. 따라서 위성을 설계할 때에 다양한 오류 상황을 예상하고 이에 대비하여 비상 시 운영 가능한 방안을 마련해야 한다. 본 논문에서는 소프트웨어적인 비상 운영 방법을 제안한다. 위성에서 발생하는 오류를 기록하여 지상으로 전달하는 방법과 긴급한 문제가 발생 하였을 경우 주요 파라 미터를 확인하여 위성에서 조치를 취할 수 있는 방법에 대해 소개하도록 하겠다.

• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.488-491
• /
• 2019

문법 오류 교정은 문법적으로 오류가 있는 문장을 입력 받아 오류를 교정하는 시스템이다. 문법 오류 교정을 위해서는 문법 오류를 제거하는 것과 더불어 자연스러운 문장을 생성하는 것이 중요하다. 이 연구는 적대적 생성 신경망(GAN)을 이용하여 정답 문장과 구분이 되지 않을 만큼 자연스러운 문장을 생성하는 것을 목적으로 한다. 실험 결과 GAN을 이용한 문법 오류 교정은 MaxMatch F_0.5 score 기준으로 0.4942을 달성하여 Baseline의 0.4462보다 높은 성능을 기록했다.