支持以 Arden 语法进行结构化输出的知识库实证分析。
Empirical analysis of knowledge bases to support structured output in the Arden syntax.
作者信息
Jenders Robert A
机构信息
Center for Biomedical Informatics & Department of Medicine, Charles Drew University, Los Angeles, CA, USA.
出版信息
Stud Health Technol Inform. 2013;192:1058.
BACKGROUND
Structured output has been suggested for the Arden Syntax to facilitate interoperability.
OBJECTIVE
Tabulate the components of WRITE statements in a corpus of medical logic modules (MLMs)in order to validate requiring structured output.
METHODS
WRITE statements were tabulated in 258 MLMs from 2 organizations.
RESULTS
In a total of 351 WRITE statements, email destinations (226) predominated, and 39 orders and 40 coded output elements also were tabulated. Free-text strings predominated as the message data.
CONCLUSIONS
Arden WRITE statements contain considerable potentially structured data now included as free text. A future, normative structured WRITE statement must address a variety of data types and destinations.
背景
有人建议采用结构化输出的Arden语法以促进互操作性。
目的
将医学逻辑模块（MLM）语料库中WRITE语句的组件制成表格，以验证对结构化输出的要求。
方法
对来自2个组织的258个MLM中的WRITE语句进行制表。
结果
在总共351条WRITE语句中，电子邮件目的地（226个）占主导地位，还列出了39个医嘱和40个编码输出元素。自由文本字符串作为消息数据占主导地位。
结论
Arden WRITE语句包含大量目前作为自由文本的潜在结构化数据。未来的规范性结构化WRITE语句必须处理各种数据类型和目的地。
Zotero 插件