Lundi 11 Février

Retour à la vue des calendrier
Lundi 11 Février
Heure: 14:00 - 16:00
Lieu: Salle B107, bâtiment B, Université de Villetaneuse
Résumé: Review of neural automatic summarization of conversations for a medical chatbot
Description: Jessica Lopez Espejel Conversation analysis plays an important role in the development of simulation devices for training health professionals (doctors, nurses). The goal of the thesis is to develop an original automatic summarization method for medical conversations between a patient and a health professional based on recent convolutional and recurrent neural summarization advances. The proposed method should be adapted to the specific problems of summarizing dialogues (and trained with abundant data from medical conversations between doctor and patients). The hypothesis is that pertinent blocks of dialogue might be extracted by tagging medical topics to dialog blocks and detecting topic evolution through deep learning.