基于预训练语言模型的公众健康问句分类

Classification of Public Health Questions Based on Pre-trained Language Model

以“公众健康问句分类”任务算法评测大赛参赛项目为例,阐述通过对开源标注数据进行训练提升模型对公众健康问句识别能力的方法,包括模型构建方法、数据描述与预处理、模型算法等,为相关研究提供参考。

Taking the competition event of“Classification of Public Health Questions”algorithm evaluation contest as an example,the paper presents a method to improve the model's ability to identify public health questions by training open-source annotated data,expounds the model construction method,data description and preprocessing,model algorithm and so on,and provides references for related study.