Distinct but correct:generating diversified and entity-revised medical response  

作　　者：Bin LI Bin SUN Shutao LI Encheng CHEN Hongru LIU Yixuan WENG Yongping BAI Meiling HU 

机构地区：[1]College of Electrical and Information Engineering,Hunan University,Changsha 410082,China [2]School of Mathematics,Sun Yat-sen University,Guangzhou 510275,China [3]JD Technology,Beijing 101100,China [4]National Laboratory of Pattern Recognition,Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China [5]Xiangya Hospital of Central South University,Changsha 410008,China [6]Teaching and Research Section of Clinical Nursing,Xiangya Hospital of Central South University,Changsha 410008,China

出　　处：《Science China(Information Sciences)》2024年第3期90-109,共20页中国科学（信息科学）（英文版）

基　　金：supported by National Key Research and Development Project (Grant No.2018YFB1305200);National Natural Science Foundation of China (Grant No.62171183);Project of Hunan Provincial Health Commission (Grant No.202114010841)。

摘　　要：Medical dialogue generation(MDG)is applied for building medical dialogue systems for intelligent consultation.Such systems can communicate with patients in real time,thereby improving the efficiency of clinical diagnosis.However,predicting correct entities and correctly generating distinct responses remain a great challenge.Inspired by actual doctors'responses to patients,we consider MDG a two-stage task:entity prediction and dialogue generation.For entity prediction,we design an ent-mac post pre-training strategy by leveraging external medical entity knowledge to enhance the pre-trained model.For dialogue generation,we propose an entity-aware fusion MDG method in which predicted entities are integrated into the dialogue generation model through different encoding fusion mechanisms,using information from different sources.Because the diverse beam search algorithm can produce responses with entities that deviate from the predicted entities,an entity-revised diverse beam search is proposed to correct the entities entailed in the generated responses and make the generated responses more distinct.The experimental results on the China Conference on Knowledge Graph and Semantic Computing 2021(A/B tests)and the International Conference on Learning Representations 2021(online test)datasets show that the proposed method outperforms several state-of-the-art methods,which demonstrates its practicability and effectiveness.

关 键 词：medical entity prediction ent-mac post pre-training strategy entity-aware fusion medical dialogue generation encoding fusion mechanism entity-revised diverse beam search 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术] R-05[自动化与计算机技术—计算机科学与技术]