초록

본 연구의 목적은 국내 디지털 도서관 관련된 연구의 동향을 분석하는 것이다. 연구의 목적을 달성하기 위하여 국내 문헌정보학 분야 대표 학술지 4종으로부터 최근 10년(2007년-2016년) 동안 발표된 디지털 도서관 관련 논문 272편을 대상으로 해당 논문의 저자가 직접 부여한 영문키워드 578개를 수집하였다. 수집된 자료는 NetMiner V.4를 사용하여 연결중심성과 매개중심성을 분석하였다. 연구 결과로는, 첫째, 연구 주제의 양적인 측면에서는 ‘Academic Library’, ‘Reference Service’, ‘Public Library’, ‘E Resource’, ‘E Book’ 순으로 나타났고, 둘째, 연구 주제의 영향력 측면에서는, ‘Academic Library’, ‘Reference Service’, ‘Information Behavior’, ‘E Resource’ 등의 순으로 나타났다. 마지막으로 연구 주제의 확장성 측면에서는, ‘Academic Library’, ‘Metadata’, ‘Information Behavior’, ‘E Resource’, ‘Librarian’ 등의 순으로 파악되었다.

Abstract

The aim of this study was to analyze the research trends on the digital libraries in Korea. To achieve this objective, a total of 578 author-assigned English keywords were collected from the 272 major LIS journal articles published in Korea during last ten years-period, i.e., 2007-2016. The collected data were analyzed using NetMiner V.4 to discover their ‘degree centrality’ and ‘betweenness centrality’. As the results, ‘Academic Library’, ‘Reference Service’, ‘Public Library’, ‘E Resource’, and ‘E Book’ showed the most frequently conducted research topics, and ‘Academic Library’, ‘Reference Service’, ‘Information Behavior’, and ‘E Resource’ were the most influencing research topics. Finally, ‘Academic Library’, ‘Metadata’, ‘Information Behavior’, ‘E Resource’, and ‘Librarian’ seemed the most widely intervening research topics in this research.

초록

최근 들어 다양한 분야에서 딥러닝이 혁신적인 기계학습 기법으로 급속하게 확산되고 있다. 이 연구에서는 딥러닝 연구동향을 분석하기 위해서 자아 중심 주제 인용분석 기법을 변형하여 응용해보았다. 이를 위해 Web of Science에서 ‘deep learning’으로 탐색하여 검색된 문헌 중 소수의 씨앗 문헌으로부터 인용 관계를 통해 분석 대상 문헌을 확보하는 방법을 시도하였다. 씨앗 문헌을 인용하는 최근 논문들을 딥러닝 분야의 현행 연구를 반영하는 자아 문헌집합으로 설정하였다. 자아 문헌으로부터 빈번히 인용된 선행 연구들은 딥러닝 분야의 연구 주제를 나타내는 인용 정체성 문헌집합으로 설정하였다. 자아 문헌집합에 대해서는 공저 네트워크 분석을 비롯한 정량적 분석을 실시하여 주요 국가와 연구 기관을 파악하였다. 인용 정체성 문헌들에 대해서는 동시인용 분석을 실시하고, 도출된 문헌 군집을 인용하는 주요 키워드인 인용 이미지 키워드를 파악하여 주요 문헌과 주요 연구 주제를 밝혀내었다. 마지막으로 특정 주제에 대한 인용 영향력이 성장하는 추세를 반영하는 인용 성장지수 CGI를 제안하고 측정하여 딥러닝 분야의 선도 연구 주제가 변화하는 동향을 밝혔다.

Abstract

Recently, deep learning has been rapidly spreading as an innovative machine learning technique in various domains. This study explored the research trends of deep learning via modified ego centered topic citation analysis. To do that, a few seed documents were selected from among the retrieved documents with the keyword ‘deep learning’ from Web of Science, and the related documents were obtained through citation relations. Those papers citing seed documents were set as ego documents reflecting current research in the field of deep learning. Preliminary studies cited frequently in the ego documents were set as the citation identity documents that represents the specific themes in the field of deep learning. For ego documents which are the result of current research activities, some quantitative analysis methods including co-authorship network analysis were performed to identify major countries and research institutes. For the citation identity documents, co-citation analysis was conducted, and key literatures and key research themes were identified by investigating the citation image keywords, which are major keywords those citing the citation identity document clusters. Finally, we proposed and measured the citation growth index which reflects the growth trend of the citation influence on a specific topic, and showed the changes in the leading research themes in the field of deep learning.

초록

Abstract

This study examined two characteristics of library and information science (LIS) journals in Korea through analysis of journal coupling with common authors. This study also illustrated the trend of Korean LIS research in the recent decade. The notable change is that record management and archival studies settle in LIS domain as a major research area. We introduced two indicators, Publishing Preference Index (PPI) and Researcher Attraction Index (RAI), based on the degree of common authors among journals. Both indicators revealed notable changes in author coupling, including reversal of PPIs in some journals, which can be interpreted as proof of changes in their author groups. The RAI analysis, which measured the degree of journals’ attractiveness to Korean LIS researchers and author sharing between two journals, illustrated the journals’ states in a domain; this result can help find both an isolated journal and strongly bonded journals in the specific domain. Journal coupling with common authors introduced in this study proved to be an effective investigative method for illustrating journals’ states in a specified domain as well as a multidisciplinary area.