個人研究影響點數之計算與應用：以某醫學期刊的作者為例

Calculation and application of personal research impact factor: an example used for authors of a medical informatics journal

錢才瑋(Tsair-Wei Chien) ； 周偉倪(Willy Chou)

《醫療資訊雜誌》 27卷4期 (2018/12) Pp. 1-12+14

摘要

許多學者每年六月皆關心與期待湯森路透公司公佈SCI各期刊的影響點數（JIF）。對於評估研究者個人學術成就的H指數（H-index），雖可務實地參考論文被引用次數，卻是累計過去的成就而非年年評估過去二年且一次公佈成果來得受人關注與側目。本研究利用醫學學術期刊（Pubmed）論文的搜尋功能，下載一定期間內的個人發表論文數及其次年被引用數，建立學者個人的影響點數，稱作個人Impact Factor（PIF）。首先建立一具演算法，能夠分配論文作者間的貢獻權重：第一作者高於通訊（暫以最後作者認定），通訊作者又高於第二及第三等作者等，其貢獻值呈等比級數遞減。蒐集醫療資訊雜誌歷年來總計674名作者英文姓名，於Medline下載2015及2016年台灣作者發表的26,872篇論文，其中1,308篇172作者具有醫療資訊雜誌作者的英文名字。本研究（1）利用社會網絡分析繪製該172作者的共同集群現象，（2）計算172作者的PIF。（3）繪製172作者的視覺化PIF比較圖。結果顯示，（1）在Pubmed上，以發表於2016年醫療資訊雜誌的作者最多（23，13.4%），以許明暉（144，4.8%）及徐建業（139，4.7%）的作者集群人數最眾。（2）以吳佳真的PIF最高（22.9）。（3）於谷歌地圖容易呈現172作者的視覺化PIF比較圖。呈現個人學術成就在同儕間的比較，示範一具可行的研究影響點數計算，以供未來評量期刊（或醫療機構）研究人員的學術成就。

關鍵字

影響點數 ； 谷歌地圖 ； 社群網絡分析 ； 醫學研究期刊 ； 視覺化集群

並列摘要

Many researchers expect the publication of journal impact factors(IF) released by Tomson Reuters in June each year. Although H-index is used for evaluating personal research achievement in academics, the effect has not been annually stirred by the wave of deep concerns and intentions drawn from authors as IF did in the past. We applied the search query functionality of Medline at Pubmed.com to download the citation records in 2017 for each article published in 2015 and 2016. The personal Medline IF(PIF) for each designated author was calculated afterward. An algorithm was generated in the current study to objectively allocate the contributions for authors: the 1^(st) author earns more credits than the last author(AKA, corresponding author) and other 2^(nd) or third ones. This is, the trend of contribution to the article is decreased in geometric progression. A total of 674 authors collected from The Journal of Taiwan Association for Medical Informatic(JTAMI) since 2005 were matched by those in 26,872 articles published by authors at Taiwan in both years of 2015 and 2016. Of which, 1,308 papers were selected and 172 authors identified. We applied (1) social network analysis(SNA) to plot clusters for the 72 authors in collaborations, (2) calculated the PIF for 172 authors, and (3) drew a visual display for the 172 PIFs in comparison. The results show that (1) the most number of authors who published articles in JTAMI comes from 2016(23, 13.4%); the authors Min-Huei Hsu(144,4.8%) and Chien-Yeh Hsu(139, 4.7%) possess the most number of authors in clusters; (2) Chia-Chen Wu gains the highest PIF (22.9) among 172 authors; (3) a visual representation dynamically displays a comparison diagram on Google Maps for 172 authors. The PRIF allows the JHM peers to compare personal academic achievement each other. We demonstrated a feasible approach of the PRIF calculation and provided any journal (or healthcare institute) with a tool to objectively evaluate member's PRIF in the future.

並列關鍵字

impact factor ； Google maps ； social network analysis ； Medline ； visualized clusters