基于最大熵原理的医疗文本信息结构化

Structured Medical Text Based on Maximum Entropy Principle

医疗文本信息结构化是处理海量医疗文本的重要手段之一。最大熵模型提供了一种自然语言处理的方法,提出了一种结合医疗标准术语的医疗文本信息结构化方案。该方案结合最大熵模型在处理规则知识上的优势,以及专业标准术语,实现了叙述性医疗报告到基于医学标准术语的结构化报告的转化,表达准确率达到93.5%。

Structured medical text is one of the important means to deal with massive health care texts. Maximum entropy model provides a natural language processing method,and proposes a structured program of medical text information combined with standard terminology for health care. In combination of professional standard terminology and maximum entropy model with the advantages of dealing with rule and knowledge,the program achieves the transformation from a narrative medical report to a structured medical report based on medical standard terminology,with 93.5% of accuracy rate of expression.