• Journal of the Korean Society for Library and Information Science
• /
• v.57 no.3
• /
• pp.351-379
• /
• 2023

The goal of this study is to investigate the domestic and international research trends on metaverse related researches. To achieve this goal, a set of 913 journal articles were collected from KCI (Korea Citation Index), 232 articles from WoS (Web of Science), and 277 articles from WoS-CPCI (Conference Proceeding Citation Index). A descriptive analysis shows the number of researches has been increased radically, and the mostly researched subject areas are interdisciplinary, computer science, and education in KCI, business and economics in WoS, and computer science in WoS-CPCI. The co-occurrence network analysis using author keywords revealed that technology related terms such as virtual reality and augmented reality showed high centrality measures in all of the databases, and the cluster analysis resulted in education and metaverse platform related keywords cluster from KCI, bibliometric analysis related keywords cluster from WoS, and all the metaverse technology related keywords cluster from WoS-CPCI.

• Science of Emotion and Sensibility
• /
• v.15 no.2
• /
• pp.239-248
• /
• 2012

This paper looks to extract personality traits from the drama characters within a drama script, and to apply it them to a character that has an artificial emotion. The method of applying the personality of a character from a drama script is as follows. First, we separate a drama script into several pieces, by the characters therin. Next, we extract emotion-related terms by matching morphemes analysis and by using an emotion terms database. Next, we analyze a dominant emotion using extracted emotion terms. Finally last, we apply the analyzed dominant emotion to an equation pertaining to artificial emotion. We made progress in developing user evaluation that features blind testing, to verify that the artificial emotion character bears the personality of a drama character. We apply three drama character personalities to artificial emotion characters bearing the same appearance. The user had to match three artificial emotion characters and drama characters according to personality. The users had a high percentage of correct answers, thus confirming the efficacy of our method of applying a personality, using information from a drama script.

• The Journal of the Korea Contents Association
• /
• v.18 no.4
• /
• pp.636-652
• /
• 2018

This study is about applying the concept of moral injury in reinterpreting and understanding the experience of female North Korean refugees who have experienced psychological trauma and are struggling with adjusting in their everyday lives. The database KISS and DBpia were used to search for academic peer-reviewed articles published before August 13th in 2017, using 'female or women North Korean,' 'female or women relocated North Korean,' and 'female or women North Korean defector' as keywords. In total 51 studies were finally selected for content analysis and the result were the following. First, listing from the most often occurring study topic to the least occurring topic in order: post-defection adjustment related, mothering experience, experience of North Korean regime and Korean War, work or job or employment related, recovering from psychological trauma as well as growth and resilience related, experiences in China, trauma and psychological difficulties, anxiety treatment process and effects. Second, words related to psychological aspect of moral injury appeared 260 times in 51 selected studies, changes in the existential belief appeared 82 times, cognitive aspects of moral injury appeared 71 times, and behavioral aspects 21 times. Third, two main contexts emerged to be the antecedent that led to moral injury and they were 'identity denial' and 'dismantled family.' In attempts to find opposite cases or stories where aspects of moral injury were not found despite the existence of its antecedents and it was 'identity restructuring.' Based on the finding, practice implications that could help female North Korean refugees prevent or recover from moral injury were suggested.

• Journal of the Korean Society for information Management
• /
• v.41 no.1
• /
• pp.411-438
• /
• 2024

This study aims to identify and comprehensively view health information-related research trends using a bibliometric analysis. To this end, 1,193 papers from 2002 to 2023 related to "health information" were collected through the Korea Citation Index (KCI) database and analyzed in diverse aspects: research trends by period, academic fields, intellectual structure, and keyword changes. Results indicated that the number of papers related to health information continued to increase and has been decreasing since 2021. The main academic fields of health information-related research included "biomedical engineering," "preventive medicine/occupational environmental medicine," "law," "nursing," "library and information science," and "interdisciplinary research." Moreover, a co-word analysis was performed to understand the intellectual structure of research related to health information. As a result of applying the parallel nearest neighbor clustering (PNNC) algorithm to identify the structure and cluster of the derived network, four clusters and 17 subgroups belonging to them could be identified, centering on two conglomerates: "medical engineering perspective on health information" and "social science perspective on health information." An inflection point analysis was attempted to track the timing of change in the academic field and keywords, and common changes were observed between 2010 and 2011. Finally, a strategy diagram was derived through the average publication year and word frequency, and high-frequency keywords were presented by dividing them into "promising," "growth," and "mature." Unlike previous studies that mainly focused on content analysis, this study is meaningful in that it viewed the research area related to health information from an integrated perspective using various bibliometric methods.

• Journal of the Korean Geographical Society
• /
• v.43 no.2
• /
• pp.188-201
• /
• 2008

With recent rapid globalization, accurate information for the foreign countries is increasingly important. Errors based on inaccurate information and unequal international relationships complicate the situations. In this article, authors analyzed the Japanese place names which appear on the Google Earth images and place name database of the NGA. Google Earth already becomes a tremendous soft power in internet society; therefore, accurate information on the satellite image is more necessary than ever. This article finds that many types of errors exist in the place names in Google Earth image service. Also many place names are listed with Japanese pronunciation in GNS database as variants. The Japanese place names have not been used in topographic maps published since 1910s and 1930s. Japanese place names were widely used in US military maps published in 1946. 1:250,000 maps published in 1954, however, doesn't seem to use Japanese pronunciation any more.

• KIISE Transactions on Computing Practices
• /
• v.22 no.2
• /
• pp.77-82
• /
• 2016

This paper presents the development of the Korean medicine symptom diagnosis system. In the Korean medicine symptom diagnosis system, the patient explains their symptoms and an oriental doctor makes a diagnosis based on the symptoms. Natural language processing is required to make a diagnosis automatically through the patients' reports of symptoms. We use morphological analysis to get understandable information from the natural language itself. We developed a diagnosis system that consists of NoSQL document-oriented databases-MongoDB. NoSQL has better performance at unstructured and semi-structured data, rather than using Relational Databases. We collect patient symptom reports in MongoDB to refine difficult medical terminology and provide understandable terminology to patients.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.18 no.2
• /
• pp.271-288
• /
• 2007

The purpose of this study is to investigate how one can find all journals of a country covered in a database such as SCI. We searched SCI and JCR Science Edition 2006 databases for Korean journals using four definition criteria-Country of a journal publication, Editorship, Unique words appeared in a journal title designating a country, a number of papers from a country in a journal. We also checked related literatures and websites to find information on SCI Korean journals. The study shows that 40 Korean journals were covered in SCI in 2006. SCI identified 38 of them as journals from South Korea, and JCR 2006 included 36. 11 journals had unique words indicating that they are Korean journals. 9 journals were Korean journals among the top 20 SCT journals by the number of papers with Korean addresses, and 36 journals were included among the top 500 SCI journals.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.65-70
• /
• 1999

음소 단위로 레이블링된 데이터베이스는 음성연구에 있어 매우 중요하다. 그러나 수작업에 의한 음소분할 및 레이블링 작업은 많은 시간과 노력이 필요하기 때문에 자동 음소분할 및 레이블링 시스템에 대한 많은 연구가 진행되고 있다. 저자들은 자동레이블링 시스템에서 레이블링 분할의 단위로monophone과 triphone의 장점을 포함하는 문맥 종속 반음소 단위 모델을 이용한 자동 음소분할 및 레이블링 시스템을 제안한바 있다[1]. 본 논문에서는 문맥종속 반음소 단위 자동음소분할 및 레이블링 시스템의 성능을 개선하기 위하여, 반음소의 단위를 개선하였다. 기존에 제안된 반음소 단위는 음소의 중점을 기준으로 left/right의 반음소 단위로 양분하였다. 본 논문에서는 음소의 길이가 120ms 이상일 경우 음소의 천이구간의 특성을 잘 나타낼 수 있도록, 음소의 앞뒤구간 각각 60ms를 전반음소와 후반음소로 나누고, 나머지 안정구간을 별도의 모델로 구성하였다. 본 논문에서 제안한 반음소 단위의 성능을 평가하기 위하여 PBW 452단어를 발성한 남자 30명분의 데이터를 이용하여 레이블링 시스템을 훈련하고, 훈련에 사용하지 않은 남자 4명분의 데이터를 이용하여 테스트 하였다. 실험결과, 기존의 반음소 단위에 비하여 10ms에서 $69.09\%$로 $1.65\%$, 20ms에서 $85.32\%$로 $1.02\%$의 성능향상을 가져왔다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.18 no.5
• /
• pp.586-591
• /
• 2008

This paper studied the emotion recognition system combined with robust speech recognition system in order to improve the performance of emotion recognition system. For this purpose, the effect of emotional variation on the speech recognition system and robust feature parameters of speech recognition system were studied using speech database containing various emotions. Final emotion recognition is processed using the input utterance and its emotional model according to the result of speech recognition. In the experiment, robust speech recognition system is HMM based speaker independent word recognizer using RASTA mel-cepstral coefficient and its derivatives and cepstral mean subtraction(CMS) as a signal bias removal. Experimental results showed that emotion recognizer combined with speech recognition system showed better performance than emotion recognizer alone.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.7
• /
• pp.50-56
• /
• 1998

본 논문은 자동 음소 분할기의 음소 경계 오류를 보상하기 위한 후처리 (Postprocessing)에 관한 연구이다. 이는 현재 음성 합성을 위한 음성/언어학적 연구, 운율 모델링, 합성단위 자동 생성 연구 등에 대량의 음소 단위 분절과 음소 레이블링된 데이터의 필요성에 따른 연구의 일환이다. 특히 수작업에 의한 분절 및 레이블링은 일관성의 유지가 어렵고 긴 시간이 소요되므로 자동 분절 기술이 더욱 중요시 되고 있다. 따라서, 본 논문은 자동 분절 경계의 오류 범위를 줄일 수 있는 후처리기를 제안하여 자동 분절 결과를 직접 합성 단위로 사용할 수 있고 대량의 합성용 운율 데이터 베이스 구축에 유용함을 기술한다. 제안된 후처리기는 수작업으로 조정된 데이터의 특징 벡터를 다층 신경회로망 (MLP:Multi-layer perceptron)을 통해 학습을 한 후, ETRI(Electronics and Telecommunication Research Institute)에서 개발된 음성 언어 번역 시스템을 이용한 자동 분절 결과와 후처리기인 MLP를 이용하여 새로운 음소 경계를 추출한다. 고립단어로 발성된 합성 데이터베이스에서 후처리기로 보정된 분절 결과는 음성 언어 번역 시스템의 분할율보 다 약 25%의 향상된 성능을 보였으며, 절대 오류(｜Hand label position-Auto label position ｜)는 약 39%가 향상되었다. 이는 MLP를 이용한 후처리기로 자동 분절 오류의 범위를 줄 일 수 있고, 대량의 합성용 운율 데이터 베이스 구축 및 합성 단위의 자동생성에 이용될 수 있음을 보이는 것이다.