The half of a century has passed since Library and information science (LIS) was introduced in Korea. LIS has developed and rapidly expanded its area of research with a rapid increase of information, media diversity, and the development of information...
The half of a century has passed since Library and information science (LIS) was introduced in Korea. LIS has developed and rapidly expanded its area of research with a rapid increase of information, media diversity, and the development of information technology in the 1990s and 2000s. The purpose of this study was to examine the intellectual structure of domestic LIS in the 1990s and 2000s using author bibliographic coupling analysis (ABCA) and an alternative approach, word-author co-occurrences analysis (WACA). ABCA is good for representing research fronts and identifying rising authors in a field. The WACA method proposed in this study is based on citation information and textual information. Key words in titles were used as indicators of content for a research topic and cited authors were used as indicators for bibliographic relationships.
This study was conducted for two periods, 1991-2000 and 2001-2010, using four representative journals with regard to LIS registered with the National Research Foundation of Korea: the 「Journal of the Korean Society for Library and Information Science」, the 「Journal of the Korean Biblia Society for Library and Information Science」, the 「Korean Library and Information Science Society」, and the 「Journal of the Korean Society for Information Management」. A contribution weight based on the number of papers in the 1990s and the 2000s was used for selecting core authors. Seventy authors with the contribution weight of 4.33 or higher were selected in the 1990s and seventy authors with the contribution weight of 9.66 or higher were selected in the 2000s. An authors’ profile matrix, cosine similarity matrix, and Pearson’s correlation coefficient matrix, and an author matrix with the ABCA method were constructed for each period using the cited authors, and an authors’ profile matrix, cosine similarity matrix, and Pearson’s correlation coefficient matrix, and an author matrix with the WACA method were constructed for each period using both cited authors and title key words.
Cluster analysis and multi-dimensional scaling (MDS) analysis were performed to examine core subject areas and to map authors in two-dimensional space, and author clusters were indicated on a MDS map. Next, Pathfinder network (PFnet) was used to visualize intellectual interchanges among authors and parallel nearest neighbor clustering (PNNC) was used to examine subject clusters in PFnet. Centrality analysis in weighted networks was used to reveal the top subject areas for global centrality.
The results of the 1990s and the 2000s intellectual structures in the field of LIS using ABCA and WACA were as follows.
First, 17 sub-disciplinary areas were found in the 1990s cluster analysis. These sub-disciplinary areas were ‘library management/policy’, ‘classification’, ‘catalog’, ‘information retrieval’, ‘digital library’, ‘information service’, ‘information material/media’, ‘automatic classification/automatic index/automatic abstract’, ‘school library’, ‘information distribution/authorship’, ‘user study’, ‘index/abstract/thesaurus’, ‘informetrics’, ‘bibliography’, ‘LIS general’, ‘information system/database’, and ‘information education’.
The core sub-disciplinary areas in 1990s were ‘information retrieval’, ‘classification’, ‘catalog’, ‘library management/policy’. Network analysis showed that ‘information retrieval’, ‘classification’, ‘catalog’, ‘digital library’ were top subject areas for global centrality and played the role of bridge that interconnected sub-disciplinary areas. Also, the result of the 1990s MDS analysis showed authors were divided into two major research field, library science and information science.
Second, 21 sub-disciplinary areas in the 2000s cluster analysis were found. These sub-disciplinary areas were ‘library management/policy’, ‘information service’, ‘LIS general’, ‘classification’, ‘catalog’, ‘information retrieval’, ‘information material/media’, ‘automatic classification/automatic index/automatic abstract’, ‘digital library’, ‘school library’, ‘information distribution/authorship’, ‘user study’, ‘index/abstract/thesaurus’, ‘informetrics’, ‘bibliography’, ‘information system/database’, ‘information education’, ‘metadata’, ‘record management/preservation’,‘open access’, ‘ontology’, and ‘web/internet’.
The core sub-disciplinary areas in 2000s were ‘library management/policy’, ‘information service’, ‘LIS general’. Network analysis showed that ‘library management/policy’, ‘information service’, ‘LIS general’ were top subject areas for global centrality and ‘information material/media’, ‘information system/database’ played the role of bridge that interconnected sub-disciplinary areas.
Third, the results of comparison between the 1990s and 2000s intellectual structures showed that new sub-disciplinary areas in the 2000s were identified by segmentation and specialization of LIS after the advent of the information society. Emerging sub-disciplinary areas were ‘metadata’, ‘ontology’, ‘record management/preservation’, ‘open access’, and ‘web/internet’. And among the 2000s sub-disciplinary areas, ‘information service’, ‘informetrics’, ‘school library’, and ‘information education’ expanded and ‘information retrieval’, ‘index/abstract/thesaurus’, ‘classification’ and ‘catalog’ decreased, compared with the 1990s sub-disciplinary areas. Meanwhile, ‘library management/policy’ and ‘information services’ formed a large part of the LIS field for 20 years.
The core sub-disciplinary areas and the sub-disciplinary areas that played the role of bridge were changed over the course of time. In 1990s, ‘information retrieval’, ‘classification’, ‘catalog’, ‘library management/policy’ were core sub-disciplinary areas and ‘classification’, ‘catalog’, ‘digital library’ played the role of bridge. However, in 2000s, ‘library management/policy’, ‘information service’, ‘LIS general’ were core sub-disciplinary areas and ‘information material/media’, ‘information system/database’ played the role of bridge.
Finally, the results of comparison between the ABCA and WACA techniques showed that the WACA technique was more effective in reflecting subject relationships among authors than the ABCA technique because key words were expected to identify a research topic in better way. Thus, WACA showed more specific research areas and detected any change in research trends promptly. Especially when citation data were insufficient, because of an emerging research field, the WACA method was expected to identify intellectual structure more effectively than the ABCA method.
As can be seen from the results above, this study revealed current research trends, major subject specializations, core authors, and their relationships within the specialties using the two methods over 20 years. The results of this study could be used as baseline data for LIS researchers. Additionally, the results demonstrate that the ABCA and WACA methods could reflect current research trends, sub-disciplines, and key authors effectively. Thus, ABCA and WACA will be useful methods for identifying the intellectual structure of other fields.