醫學博碩士論文關鍵詞與MeSH詞彙之對應研究－以臺北醫學大學為例

Mapping Medical Theses Keywords with MeSH Vocabulary─A Case Study of Taipei Medical University

謝育慈(Yu-Tzu Hsieh)

指導教授 ： 王美玉

淡江大學/文學院/資訊與圖書館學系數位出版與典藏數位學習碩士在職專班/碩士(2016年)

https://doi.org/10.6846/TKU.2016.00730

摘要

由於醫學資訊變動快速，時時有醫學新知的出現，對於臨床醫學人員、醫學院師生而言，醫學資訊必需是可取得的、有權威性的、可靠的、正確且即時的。因此，除了醫學學術期刊外，具有原創性和詳細完整研究內容的醫學博碩士論文，是重要的醫學資訊管道。「關鍵詞」檢索是目前網際網路中常見的檢索策略。由網路使用者自行決定關鍵詞彙或片語，是最快速且最便捷的查詢方法之一。關鍵詞乃敘述文獻重點的詞彙，大多數為作者在自然無受控情況下所給予的自由詞，屬於非控制詞彙。
相反的，控制詞彙是在嚴謹制度的把關下所提供的詞彙語意，其上下層級架構嚴謹且有其限制。醫學領域中最重要的控制詞彙表為美國國家醫學圖書館所發行的MeSH。MeSH將主題詞分為16大類，每大類下還有子類，因此透過醫學博碩士論文關鍵詞出現的頻率與其在MeSH上下主題層級的分布狀況亦可一窺探研究主題的趨勢。
本研究以臺北醫學大學醫學博碩士論文關鍵詞為研究對象，探討其與MeSH的對應程度，並研究臺北醫學大學醫學博碩士論文近十年來的主題分佈概況以及主題研究趨勢。
本研究資料分析結果發現民國93至102十年間的臺北醫學大學醫學博碩士論文，大約有50%的關鍵詞符合MeSH的主標目或近似詞，而在比較不符合MeSH主標目、同義詞或補充概念的關鍵詞中有很多是因為作者使用縮寫語或拼錯字。此外，透過約占50%符合MeSH的主標目或近似詞之臺北醫學大學醫學博碩士論文關鍵詞與MeSH的樹狀結構做主題分析，發現化學品和藥物、疾病、解剖學以及現象與過程，此四種主題為這十年間最重要的研究方向。若以這十年間臺北醫學大學醫學博碩士論文的關鍵詞分析，則發現熱門的研究主題詞彙為细胞凋亡、乳癌、血小板、結締組織生長因子以及 樹突狀細胞。
由本研究發現可推知MeSH的主題詞彙在描述醫學研究主題方面具足一定的程度，但是博碩士論文的作者對於MeSH的認識可再加強。關鍵詞能夠幫助到學術作者提升文獻的能見度及被檢索機率，以促進與同儕之間的學術交流，因此宣導與教育學術作者如何下關鍵詞也是不能忽視的課題之一。

關鍵字

MeSH ； 關鍵詞 ； 自由詞彙 ； 制詞彙 ； 對應分析 ； 主題分析

並列摘要

Because medical information is constantly changing, it is very crucial for medical practitioners and students to have access to up-to-date, authentic and authoritative medical information. Aside from journal publications, theses and dissertations, which consist of original thoughts and through coverage of research results, are of critical importance to medical information retrieval. At present, keyword search is one of the most common techniques used for retrieving data or information. Users can decide keywords or phrases by themselves, which makes keyword search to be one of the most convenient search methods. Keywords are significant words as a point of reference to overall contents of articles, and most of them are uncontrolled vocabularies chosen by authors in their own categorization.
In contrast with uncontrolled vocabularies, controlled vocabularies are carefully selected list of words and phrases projecting organized indexing systems. In the medical sciences, Medical Subject Headings (MeSH), created by United States National Library of Medicine, is a comprehensive controlled vocabulary recourse organized in 16 categories and various sub-categories. Given the above mentioned, investigating occurrences of keywords from theses and dissertations and comparing distributions of those keywords to the structure of MeSH may be beneficial to understand research trends.
This study aims to investigate the natural (free-text) keywords given in the theses and dissertations in medical science from Taipei Medical University (TMU). The purpose of this study is twofold. First, a comparison is made between those natural keywords and the controlled vocabularies from Medical Subject Headings (MeSH) to see the mapping between these two sets of words. Second, an analysis is conducted to understand the research trends and topics between 2004 to 2013 of TMU theses and dissertations in medical sciences
The results are as follows. First, there is about 50% of the time when those natural keywords can be mapped to the corresponding descriptors or synonyms in MeSH. The chance level performances are resulted from the use of acronyms or the misspelling. Second, by subject analysis of the mapped keywords, it is found that the top four subject areas are chemicals and drugs, diseases, anatomy and phenomena and processes. In terms of keyword analyses, some of the most frequently used keywords are apoptosis, breast cancer, platelet, connective tissue growth factor and dendritic cell.
The hierarchically-organized terminology of biomedical information in MeSH is informative. However, it is discovered that authors of theses and dissertations may need more exposure to MeSH for a more refined understanding of keywords. Creating appropriate keywords are essential to optimize search volume. Leaning how to use keywords for more exposure is a profound issue for every researcher.

並列關鍵字

Keyword ； MeSH ； Natural vocabulary ； Controlled vocabulary ； Mapping Analysis ； Subject Analysis

參考文獻

邱舜逸（2010）。劇本關鍵詞彙於分鏡場景製作之應用研究（未出版之碩士論文）。崑山科技大學，臺南市。

陳昭珍（1998）。控制詞彙與非控制詞彙主題索引問題之探討。圖書館學刊，13，103-128。

張郁蔚、呂淑慧（2013）。大學圖書館館員於職場之資訊尋求行為研究。大學圖書館，18（1），14-32。

楊美華（2012）。大學圖書館在新學術傳播體系的角色。大學圖書館，16（2），1-19。

鄒明城（2010）。網頁地理資訊檢索與探勘─以民宿主題為例。資訊管理學報，17（3），19-44。