FrancophonIA/PxCorpus

PxCorpus是一个法语spoken medical drug prescriptions的语料库，包含了55名专家和非专家参与者的4小时转录和注释对话。这个数据集旨在用于自然语言处理模型的训练，特别是那些处理药物处方自动转录和语义标注的模型。数据集在获取协议上经过了医学专家的审核，确保了隐私和规定的遵守，可以自由分发。总共有2067个录音，由38%的非专家、25%的医生和36%的医学从业者完成，所有录音都经过了人工转录和语义标注。

PxCorpus is a corpus of spoken medical drug prescriptions in French, containing 4 hours of transcribed and annotated dialogues from 55 participants, including both experts and non-experts in drug prescriptions. This dataset is intended for training NLP models, particularly those dealing with automatic transcription and semantic annotation of drug prescriptions. The data acquisition protocol has been reviewed by medical experts to ensure compliance with privacy and regulations, allowing for free distribution. There are a total of 2067 recordings completed by 38% non-experts, 25% doctors, and 36% medical practitioners, all of which have been manually transcribed and semantically annotated.