引用文獻索引資料庫在Institute for Scientific Information(ISI)建置了Science Citation Index(SCI)與Social Science Citation Index(SSCI)以後，逐漸為學術界所重視，也帶動了傳統商業公司發展引用文獻索引資料庫之風潮，其中又以Science Citation Index-Expanded(SCIE)及後起之Scopus最為人稱道。但由於傳統商業公司對學者及圖書館收費過於高昂，引起學者及圖書館的反動，興起一陣由計畫及少數商業公司所發展，開放存取引用文獻索引資料庫之風潮，其中又以Google公司製作的Google Scholar，及NEC公司隨著計畫建置的CiteSeer最受人注目。 / 本研究採取實作法為研究方法，評比四個引用文獻索引資料庫的檢索介面及檢索細項之優劣。並以美國計算機械學會(ACM)頒發的杜林獎之50位得主為樣本，對SCIE、Scopus、CiteSeer及Google Scholar四個引用文獻索引資料庫進行作者檢索，逐一過濾檢索結果後，針對正確的檢索結果進行分析，比較四個引用文獻索引資料庫內部重複性與完整性，並交叉比對四個引用文獻索引資料庫兩兩比較之重複性、獨特性及完整性，並歸納造成此研究結果之原因。 / 研究結果發現SCIE與Scopus的檢索方式較容易，不會造成使用者太大的負擔，檢索方式也較為多元詳盡，其中又以Scopus的作者檢索使用最方便；而Google Scholar及CiteSeer皆主要利用一簡潔的檢索列，較難精準的檢索出所需資料。收錄資料完整度方面，Google Scholar收錄資料最多元，SCIE則涵蓋最完整之學術資源。交叉比對結果可得知，Google Scholar之資料獨特性最高；CiteSeer之收錄資料完整度最低。此外除了SCIE以外，其他三個引文索引資料庫皆收錄大量的網路資源。此外，美國計算機械學會的出版品則在四個引文索引資料庫中，皆扮演重要角色。 / 根據研究結果，對此四個引文索引資料庫提出建議，希望傳統商業引文索引資料庫能增加索引網路資源，並調整收費政策；開放存取引文索引資料庫應改正其書目著錄格式；希望圖書館能增加對引文索引資料庫使用之推廣，並教導使用者正確利用開放存取引文索引資料庫。 / 引文索引資料庫索引之文獻，已對學術評鑑造成很大的影響。圖書館應實地使用並引導使用者正確的利用引文索引資料庫，及使用網路資源的正確觀念。如此方可協助使用者不在浩瀚之網路資源中迷失。 / After Institute for Scientific Information(ISI) made Science Citation Index(SCI) and Social Science Citation Index(SSCI), Scholars progressively took notice of citation indexing databases. Commercial Companies also had begun to expand citation indexing database like the famous products are Science Citation Index-Expanded(SCIE). However, the commercial companies charged too much for using the database. So it excited the development of open access(OA) citation indexing database, instant of Google Scholar and CiteSeer. / OA means that people can use these citation indexing database for free. This paper aims to adopt comparison as four databases’ retrieval interface, and unique and overlap of documents of the subjects of computing machinery and electrical engineering. The research subjects are composed of OA and traditional commercial citation indexing database in the follow: SCIE, Scopus, Google Scholar, and CiteSeer. Moreover, this research retrieved all documents of Turing award winners in the four citation indexing databases, in order to examine these four citation indexing databases’ unique and overlap. / As a consequence, this study provides the findings as follows : Firstly, traditional commercial citation indexing databases (SCIE and Scopus) have the easier retrieval interface and various searching forms. The Google Scholar collects more multiform resources of retrieval results, and SCIE completed collects scholarly literatures. We make a comparison to find that Google Scholar has much more unique data, but CiteSeer is completely less in four citation indexing databases. Besides SCIE, another three citation indexing databases conclude a large number of internet data. Finally, publications of The Association of Computing Machinery(ACM) play an important role in the four citation indexing databases.

摘要 《紅樓夢》是一部具有高度思想性和高度藝術性的文學鉅著，其前進思想和表現的寫作技巧，無可置疑的領先同時代的作家和作品。因為其具有獨特的藝術魅力，所以不但廣泛的流傳民間，也成功地站上世界文學之林。 《紅樓夢》雖然膾炙人口且流傳已逾兩百餘年，然而本書真正的作者是誰，卻一直是學者專家們爭論的話題。在大家的印象中，紅樓夢前八十回由清朝曹雪芹所寫，而後四十回則由高鶚所續編完成，但是研究紅樓夢的學者對於此一說法，仍抱著懷疑的態度，不斷的尋求證據以解答此問題的真相。 近年來，學者憑靠著殘存的證據，試圖以各種研究方法予以合理的推論，然時空變遷，只能恢復部分的歷史真相，無法給予完整的復原，而《紅樓夢》的作者究竟是誰，至今尚未有一個大家認同的答案。 本論文嘗試以品種比較、樣本重複性及品種涵蓋率等統計方法，配合電腦的檢索，藉由分析寫作風格及其用字習慣，以統計分析的角度來推論《紅樓夢》的作者。 關鍵詞：紅樓夢、品種問題、樣本重複性、卜瓦松過程。 / Abstract "The Dream of Chamber" is a greatly artistic novel in Chinese literature. Undoubtedly, the writing style and the delicate design of this book lead the other authors and novels at the same time. Because of its distinctive charm, it is wide-spreading not only in China but also in the other country. Although "The Dream of Chamber" has been spread more than two hundred years, however it also exists a mystery─"Who is the real author of this book？". Most people believe that Sher-Chin Tsao wrote the first 80 chapters, and Gao-E wrote the last 40 chapters. But many have doubt about this statement. People try to find evidence in order to solve this problem, but still have not a persuasive answer. In this report, we attempt to solve this riddle by statistical analysis, including the methods of species comparing, species overlap, and sample coverage etc., besides, we use computer to search words. We try to infer the author of "The Dream of Chamber" from the statistical point of view. Keyword: The Dream of Chamber, species comparison, sample coverage. Poisson Process.

自2003年布達佩斯宣言公佈起，國際間學術文獻開始開放取用的趨勢。於此背景下，本研究以諾貝爾生物醫學獎近十年23位得主為研究樣本，評比在商業資料庫(SCIE、Scopus)及開放取用系統(生物醫學類：Pubmed、Highwire；綜合類：Google Scholar)的文獻收錄狀況，除了比較其內部重複性與完整性，並交叉比對五個資料庫與系統的重複性、獨特性及完整性，同時也觀看能否取得全文的比率，來了解現今開放取用文獻的狀況，進而觀察開放取用系統和商業資料庫兩者是否可以互補，或是開放取用系統有代替商業資料庫的可能性。 研究結果顯示五個資料庫及系統的檢索形式多元。針對作者檢索而言，Scopus最完善，資料收錄也較齊全；SCIE及Pubmed兩者則是檢索結果最為相似。如果以學術出版收錄而言，則是Highwire較完整；至於Google Scholar的獨特性較高。整體而言，開放取用系統比商業資料庫的全文收錄比例高，但Scopus是收錄最多全文的資料庫。本研究同時也發現PNAS是五個資料庫與系統之重複來源及獨特來源。另外，使用PubMed及Highwire檢索生物醫學文獻會比Google Scholar來得專業。 根據研究結果建議，商業資料庫可考慮將網路開放資源納入收錄範圍，以便妥善整理及應用網路資源的書目及全文。開放取用系統則應改善索引書目之正確性及著錄完整性。另外，針對圖書館的服務宜採取以下之因應措施：(1)加強推廣商業資料庫之正確檢索方式及使用時機；(2)教導如何正確使用開放取用系統的檢索模式；(3)平衡商業資料庫和開放取用系統的使用，以達成圖書館經費的合理運用。 本研究後續可延伸至生物醫學領域的臨床及實證醫學上，以了解生物醫學中兩個最具時效性的學術文獻系統是否達到開放取用的立即性及實用性。再者，使用者對開放取用的滿意度研究是學術出版界急欲了解的課題，也是後續研究可加強努力的方向。 / The International Scholarly Communication has gradually forwarded open access system since the publication of Budapest Declaration in 2003. Under this research background, this study uses biomedical Nobel Prize winners in recent years for the study of 23 samples of appraisal in the commercial database (SCIE, Scopus) and open access systems (biomedical categories: Pubmed, Highwire ; Comprehensive: Google Scholar) literature collection status, in addition to comparing repeatability and integrity of its internal and cross-comparison of the five databases and system repeatability, uniqueness and integrity, while also viewing the ability to obtain the ratio of text to understand current status of open access literature, and then observe the open access systems and commercial databases whether the two can complement each other, or open access database system instead of commercial possibilities. The results showed that five databases have different retrieval systems in many different forms. For the purposes of retrieval, Scopus collections are more complete; SCIE and Pubmed are the most similar two databases in the search results. Inclusion academic publishing purposes, Highwire is the most complete one. For Google Scholar, the collection’s uniqueness is the highest. Overall, comparing the open access system with commercial database, open access system contains a high proportion of full text. Scopus is the most one of full text collections. The PNAS study also found that five of the duplicate database and system sources and unique source. In addition, the use of PubMed and Highwire retrieved biomedical literature is more professional than Google Scholar. According to the study results suggest that commercial databases can be considered included in the scope of network resources into the open, in order to properly organize network resources and application of bibliographic and full-text. Open access system should improve the accuracy and bibliographic indexing bibliographic completeness. In addition, for the library service should take the following measures in response to: (a) enhance the promotion of commercial database retrieval methods and the use of proper timing; (2) to teach the proper use of open access system retrieval mode; (3) Balance Business open access database and use of the system, in order to achieve rational use of library funds. The follow-up research of this study can be extended to the field of clinical and biomedical evidence-based medicine research. The follow-up research results can be used to understand the biomedical literature’ timeliness, whether the system reaches an open access immediate or practicality. Furthermore, users' satisfaction with open access scholarly publishing research is also an anxious subject to know, and the follow-up study will strengthen efforts.