引文分析學術領域中,「書目耦合關係」是較先被提出的議題,但隨著另一議題「共同引用關係」被提出之後,書目耦合關係即漸漸式微,而共同引用關係取而代之成為研究主流。但經由文獻回顧,發現近年來之研究逐漸著力於探討兩關係之差異,這也意味著兩關係間仍存在著研究空間及價值值得去開發及挖掘。 引文分析領域研究以往所探討議題絕大多是以「共同引用關係」為主,而本研究則針對「書目耦合關係」與「共同引用關係」所產生的智識構圖與領域議題內容進行解讀比較,經由比較結果提供引文分析領域新的研究思維與方向,以增進領域研究成果的深廣度,提昇後續研究的學術基礎。 本研究利用自Citeseer科技文獻資料庫蒐集之文獻進行過濾取得實驗資料集,將此資料集分別以「書目耦合關係」與「共同引用關係」進行引文分析流程,最後將兩方法所產生之智識構圖進行五項指標比較,說明兩者之間孰優孰劣,並分析解讀文獻內容。總結本研究結果,遂發現兩智識構圖,在「因素數量」及「因素本質」指標上,共同引用關係優於書目耦合關係;「因素內容」中,若研究者欲瞭解近期領域發展及新興議題,則採用書目耦合關係較佳,若研究者欲綜觀領域發展歷史、演進情況,則採用共同引用關係較為適當;最後在「因素重疊」與「圖形結構」上,兩者不分軒輊,皆能清楚呈現知識領域中各議題之關聯情形與現況。
In the citation analysis fields, "bibliographic coupling" appears prior to "co-citation". While co-citation have been used widely, whereas bibliographic coupling is discussed and used rarely. However, we saw more researches have been devoted to study the differences between the two citation methods recently. It has been purported that the bibliographic coupling may supplement the co-citation in some aspects. In contrast with most of the previous researches that focus on the co-citation analysis, this research compares the intellectual structures and domain contents derived from the bibliographic coupling and co-citation analysis. This study obtains dataset collected from the Scientific Literature Digital Library – Citeseer. We use the same dataset in the analysis utilizing both bibliographic coupling and co-citation, and compare the intellectual structures derived from these two analyses in terms of five indicators. We then illustrate the differences between the two outcomes and further analyze the content and interpret the result. In summary, we find the analysis using the co-citation is better than that of bibliographic coupling in two aspects – "Size of specialties" and "Nature of specialties". In the "Individual document's memberships" indicator, bibliographic coupling is suitable to researchers who want to understand the recent developments and emerging issues in the field but co-citation is more appropriate when researchers want to have an overview of the evolution history of the field. Lastly, they are equally good in the aspects of "Overlap between specialties" and "Overall structure" because both display clearly the association between various issues and the present situation of the knowledge domain.