http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
송예슬,박지홍,심지영 한국문헌정보학회 2015 한국문헌정보학회지 Vol.49 No.2
The purpose of this study is to suggest an appropriate index for evaluating preferences of Web contents by examining the h-index and its variants. It focuses on how successfully each index represents relative user preference towards topical subjects. Based on data obtained from a popular IT blog (engadget.com), subject values of the h-index and its variants were calculated using 53 subject categories, article counts and the ‘Likes’ counts aggregated in each category. These values were compared through critical analysis of the indices and Spearman rank correlation analysis. A PFNet (Pathfinder Network) of subjects weighted by hT values was drawn and cluster analysis was conducted. Based on the four criteria suggested for the evaluation of Web contents, we concluded that the hT-index is a relatively appropriate tool for the Web contents preference evaluation. The hT-index was applied to visually represent the relative weight (topic preference by user ‘Likes’ count) for each subject category of the real online contents after suggesting the relative appropriateness of the hT-index. Applying scientometric indicators to Web information could provide new insights into, and potential methods for, Web contents evaluation. In addition, information on the focus of users’ attention would help online informants to plan more effective content strategies. The study tries to expand the application area of the h-type indices to non-academic online environments. The research procedure enables examination of the appropriateness of the index and highlights considerations for applying the indicators to Web contents.
송예슬(Ye-Seul Song) 한국산학기술학회 2021 한국산학기술학회논문지 Vol.22 No.12
치료과정에서 위험은 항상 존재하며, 정확한 현황 파악이 이루어져야 예방 가능한 병원 손상을 줄일 수 있으므로 해당 연구를 계획하였다. 2012년부터 2018년까지의 퇴원손상심층조사 원시자료를 이용하고, 복합표본설계 요소를 반영하여 빈도분석, 카이제곱 검정, T-test 검정을 실시하였다. 병원 손상을 재원 중 외상, 약물과 관련된 병원 손상, 재난과 관련된 병원 손상, 장치와 관련된 병원 손상, 수술 및 처치와 관련된 병원 손상 총 5가지 범주로 분류하였고, 전체 환자 중 1.64%가 이에 해당하였다. 병원 손상이 있는 환자의 주진단은 손상, 중독 및 외인에 의한 특정 기타 결과가 가장 많았으며, 세분화하여 비교한 결과, 병원 손상 유무에 따라 주진단에 뚜렷한 차이가 있음을 알 수 있었다. 소화기계와 근골격계 수술에서 수술 관련 병원 손상이 많았으며, 성별, 연령, 입원 경로가 유의확률 0.05에서 병원 손상에 유의한 영향을 주는 것으로 나타났다. 또한, 병원 손상은 치료결과에도 영향을 주며, 병원 손상이 없는 환자에 비해 병원 손상이 있는 환자의 재원기간이 약 11.42배 증가함을 알 수 있었다. 이 연구를 통해 병원 손상의 전반적인 사례를 확인하고, 의료기관의 적극적인 예방과 개선에 도움이 되고자 한다. A frequency analysis, chi-square test, and T-test were conducted to reflect complex sample design elements. The hospital treatment-related medical errors were classified into five categories: trauma during hospitalization, medication-related, misadventure-related, device-related, and surgical and other procedure-related errors, involving 1.64% of all patients. The most common primary causes of hospital-related medical errors were injury, poisoning, and certain other consequences of external causes. As a result of a subdivided comparison, it was found that there was a clear difference in the main diagnosis according to the presence or absence of hospital treatment-related medical error. Surgical-errors in gastrointestinal and orthopedic surgeries, and gender, age, and admission routine had a significant impact on hospital damage at P<0.05. In addition, it was found that such errors affect treatment outcomes, increasing the length of hospital stay by about 11.42 times in these patients compared to patients with no hospitalization-related medical error. Through this study, it was possible to confirm the overall incidence of hospitalization-related medical errors. This study is intended to help the active prevention of such errors and bring about an improvement in the treatment process in medical institutions.
송예슬(Yeseul Song),전정현(Junghyun Chun),감미아(Miah Kam),송성전(Sungjeon Song),이지연(Jiyeon Lee) 한국HCI학회 2014 한국HCI학회 학술대회 Vol.2014 No.2
디지털도서관의 활용도와 중요성이 높아짐에 따라, 사용성(Usability)이 중요한 화두가 되었다. 한편, 사용자 인터페이스는 사용성에 가장 직접적인 영향을 미치는 요소로, 누구나 쉽고 편리하게 서비스를 사용할 수 있도록 설계되어야 한다. 본 연구는 디지털도서관의 사용자 인터페이스 평가지표를 개발하고, 이를 국립중앙도서관 디지털도서관에 적용해 문제점을 분석하고 개선방안을 제시하였다. 평가지표는 문헌연구 및 사례분석을 통해 개발하였으며, 디지털도서관의 특수성을 반영하고자 하였다. 디지털도서관 운영 목적 및 사용자 지원, 시스템메뉴와 네비게이션, 동작과 상호작용, 페이지가 제공하는 정보, 시각적 디자인, 피드백 및 도움, 검색 총 7 가지 분류항목에 따라 세부기준을 마련하였다. 이를 바탕으로 실시한 국립중앙도서관 디지털도서관 분석을 통해서는 총 267 개의 문제점을 발견하였으며, 각각에 대한 구체적인 개선방안을 제시해 디지털도서관 개선 작업 시 실무에서 즉시 활용할 수 있도록 하였다. As popularity and importance of digital libraries are getting higher, usability of the digital libraries are becoming increasingly more crucial. The user interface is one of the most influential factors affecting the usability, thus the user interface should be designed to support easy and convenient usage of the corresponding system. In this study, we developed user interface evaluation heuristics, and applied them to The National Library of Korea’s digital library. The heuristics were developed through literature review and case analysis, while considering the characteristics of digital libraries. The evaluation heuristics include specific criteria according to the seven classification schemes such as purpose of the digital library and user support; menu and navigation; action and interaction; provided information; visual design; feedback and help; and information retrieval. Based on the heuristics, 267 problems were found in The National Library of Korea’s digital library. Practical and specific improvement suggestions were made for each problem to enable immediate implementation of the suggestions when the renovation of the digital library commences.
옥수빈,이대호 한국정보과학회 2023 정보과학회논문지 Vol.50 No.3
Image description, which has been receiving much attention with the development of deep learning, uses computer vision methods that identify the contents of images and natural language processing methods that represent descriptive sentences. Image description techniques are utilized in many applications including services for visually impaired people. In this paper, we summarize image description methods within three categories; template-based methods, visual/semantic similarity search-based methods, and deep learning-based methods, and compare their performances. Through performance comparison, we try to provide useful information by offering basic architectures, advantages, limitations, and performances of the models. We especially survey the deep learning-based methods in detail because the performances of these methods are significantly improved compared to other methods. Through this process, we aim to organize the overall contents of image description techniques. For the performance of each study, compare the METEOR and BLEU scores for the commonly used Flickr30K and MS COCO datasets, and if the results are not provided, check the test image and the sentences generated for it.