Improving the Transferability of Clinical Note Section Classification Models with BERT and Large Language Model Ensembles
Improving the Transferability of Clinical Note Section Classification Models with BERT and Large Language Model Ensembles
复制标题
使用 BERT 和大型语言模型集成提高临床记录部分分类模型的可迁移性
DOI：
10.18653/v1/2023.clinicalnlp-1.16
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Timothy Miller
中科院分区：
文献类型：
--
作者：
Weipeng Zhou;M. Afshar;Dmitriy Dligach;Yanjun Gao;Timothy Miller
电子健康记录中的文本被组织成部分，将这些部分分类为部分类别对于后续任务很有用。在这项工作中，我们试图通过将监督学习模型中的数据集特定知识与大型语言模型（llm）中的世界知识相结合来提高截面分类模型的可移植性。令人惊讶的是，我们发现零射击llm优于应用于域外数据的基于bert的监督模型。我们还发现它们的优势是协同的，因此一个简单的集成技术可以带来额外的性能收益。
Text in electronic health records is organized into sections, and classifying those sections into section categories is useful for downstream tasks. In this work, we attempt to improve the transferability of section classification models by combining the dataset-specific knowledge in supervised learning models with the world knowledge inside large language models (LLMs). Surprisingly, we find that zero-shot LLMs out-perform supervised BERT-based models applied to out-of-domain data. We also find that their strengths are synergistic, so that a simple ensemble technique leads to additional performance gains.