Adverse Drug Events (ADE) Corpus

开发基准语料库以支持从医疗案例报告中自动提取与药物相关的不良反应。大量关于药物相关安全问题（例如副作用）的信息发布在医疗案例报告中，由于其非结构化性质，这些信息只能由人类读者探索。这里介绍的工作旨在生成一个系统注释的语料库，该语料库可以支持从医学病例报告中自动提取药物相关不良反应的方法的开发和验证。文档在多轮系统中进行了双重注释，以确保注释的一致性。带注释的文档最终被协调以生成具有代表性的共识注释。为了演示一个示例用例场景，使用语料库来训练和验证模型，以针对非信息性句子对信息性进行分类。使用简单特征训练并通过 10 倍交叉验证进行评估的最大熵分类器的 F₁ 得分为 0.70，表明该语料库的潜在有用应用。

Develop a benchmark corpus to support the automatic extraction of drug-related adverse events from medical case reports. A large volume of information concerning drug-related safety issues (e.g., side effects) is published in medical case reports; due to their unstructured nature, this information can only be explored by human readers. The work introduced herein aims to generate a systematically annotated corpus that supports the development and validation of methods for automatically extracting drug-related adverse events from medical case reports. Documents underwent dual annotation in a multi-round annotation framework to ensure annotation consistency. The annotated documents were ultimately reconciled to generate representative consensus annotations. To demonstrate an exemplary use case scenario, the corpus was used to train and validate a model for classifying sentences as informative or non-informative. A Maximum Entropy classifier trained with simple features and evaluated via 10-fold cross-validation achieved an F1-score of 0.70, demonstrating the potential utility of this corpus.

搜集汇总

背景与挑战

背景概述

Adverse Drug Events (ADE) Corpus是一个用于支持从医疗案例报告中自动提取药物相关不良反应的基准语料库，经过系统双重注释以确保一致性。该语料库可用于训练和验证模型，例如进行信息性句子分类，示例应用显示F₁得分为0.70，适用于生物医学文本处理任务如命名实体识别和关系抽取。