開放取用系統與商業資料庫收錄資料之正確性與重複性

Bibliographic Data Accuracy and Overlap in Open Access System and Commercialized Database

本研究旨在比較商業與開放取用兩種學術傳播工具，透過實際操作商業引文資料庫Web of Science與Scopus與搜尋引擎Microsoft Academic，匯集式機構典藏系統Open DOAR和物理學專科特性的Astrophysics Data System開放取用引文系統，分析其收錄書目資料的正確性與重複性。本研究以諾貝爾物理學2001年至2013年得獎者之著作為研究樣本，進行書目資料剖析，包括排序、比對、刪除、聚集與統計，並相互交叉比對分析結果，比較各商業資料庫與開放取用系統之優劣，並根據分析結果提出建議。本研究結果期望能提供圖書館選擇引文索引資料庫與建置機構典藏系統，或引文資料庫與系統未來發展之參考。此外，亦可作為研究人員進行學術傳播與學術單位進行學術評鑑採用之指標與工具之建議。 This study investigates scholarly communication systems’ data accuracy, internal duplication, and between-system overlap in coverage through a comparison of bibliographic records from three open access systems and two commercial databases. The former group of open access systems includes an Internet search engine Microsoft Academic, an aggregating institutional repository OpenDOAR, and an academic open access system in physics the Astrophysics Data System; and the latter group includes Web of Science and Scopus. For the study, searches for the publications by winners of Nobel Prizes in Physics from 2001 to 2013 that were conducted in the five abovementioned citation systems. Bibliographic records representing the sampled publications were retrieved and downloaded from each system. The analytical tasks of sorting, matching, elimination, aggregation, computation, and comparison to assess individual systems’ data accuracy, internal duplication rates, and the percentages of overlap between systems were then performed. The findings of the study may provide: (1) valuable information for libraries to acquire citation index databases, build institutional repositories, or create future citation index systems on their own; and (2) a useful resource in scholarly communication and academic assessment.