FlauBERT vs. CamemBERT: Understanding patient's answers by a French medical chatbot
作者：
Highlights：
• Medical chatbots may help healthcare providers save time and orient patients.
• Chatbots use Natural Language Understanding (NLU) to analyze patients' answers.
• NLU helps building language models that predict intents and slots from sentences.
• Two French language models were compared regarding intent and slot prediction.
• FlauBERT outperformed CamemBERT with all neural network architectures used for intent and slot prediction.
摘要
•Medical chatbots may help healthcare providers save time and orient patients.•Chatbots use Natural Language Understanding (NLU) to analyze patients' answers.•NLU helps building language models that predict intents and slots from sentences.•Two French language models were compared regarding intent and slot prediction.•FlauBERT outperformed CamemBERT with all neural network architectures used for intent and slot prediction.
论文关键词：Intent and slot prediction,FlauBERT,CamemBERT,Language models,Natural Language Understanding,Neural network architectures
论文评审过程：Received 10 June 2021, Revised 15 February 2022, Accepted 23 February 2022, Available online 2 March 2022, Version of Record 12 March 2022.
论文官网地址：https://doi.org/10.1016/j.artmed.2022.102264
