PICO语言学家与BabelMeSH：用于MEDLINE/PubMed的循证多语言搜索工具的开发与部分评估

PICO Linguist and BabelMeSH: development and partial evaluation of evidence-based multilanguage search tools for MEDLINE/PubMed.

作者信息

Fontelo Paul, Liu Fang, Leon Sergio, Anne Abrahamane, Ackerman Michael

机构信息

High Performance Computing and Communications, Lister Hill National Center for Biomedical Communications, NIH, Bethesda, MD 20894, USA.

出版信息

Stud Health Technol Inform. 2007;129(Pt 1):817-21.

PMID:17911830

Abstract

PICO Linguist and BabelMeSH are multilanguage search tools intended for users whose native language is not English. A database of medical terms was created using concept identification equivalents of English terms to other languages. The primary sources of vocabularies were UMLS, MeSH, WHO EMRO and UMLF. The search interface changes according to the language selected which allows search terms to be entered in the native language. The user can limit the search output according to the language of publication but citations retrieved are in English only. Links may be provided to journals if published online. Evaluation of the French and Spanish versions using journal key words and a list of common diseases showed 77.5% and 86.5% accuracy respectively. User feedback was positive. PICO Linguist and BabelMeSH could be useful and convenient tools in finding current evidence sources in the medical literature especially for non-English medical terms that may be difficult to express in English.

摘要

PICO Linguist和BabelMeSH是面向母语非英语用户的多语言搜索工具。利用英语术语与其他语言的概念识别对应词创建了一个医学术语数据库。词汇的主要来源是统一医学语言系统（UMLS）、医学主题词表（MeSH）、世界卫生组织东地中海区域办事处（WHO EMRO）和统一医学语言形式（UMLF）。搜索界面会根据所选语言进行变化，这使得搜索词能够以母语输入。用户可以根据出版物的语言限制搜索输出，但检索到的参考文献仅为英文。如果期刊在线发表，可能会提供链接。使用期刊关键词和常见疾病列表对法语和西班牙语版本进行评估，结果显示准确率分别为77.5%和86.5%。用户反馈良好。PICO Linguist和BabelMeSH在查找医学文献中的当前证据来源时可能是有用且便捷的工具，特别是对于可能难以用英语表达的非英语医学术语。