Research On A Dental Consultation Training System Based On Natural Language Processing

With the continuous improvement of people’s living standards,more and more people begin to pay attention to oral health issues.In addition,the dental care industry is facing a shortage of physicians.On the one hand,the process of training a qualified dentist is long,complicated and costly;on the other hand,the existing consultation training system for dentists is weak in interaction,and it is difficult to effectively train doctors’ consultation ability.In order to solve the above problems,this paper implements a dental consultation training system based on natural language processing technology,which can effectively train doctors’ consultation ability.On the model,this article uses the RoBERTa-BiLSTM model to extract relevant features and determine whether two texts are similar;On the system,this article is based on common clinical cases and utilizes a trained matching model to implement the consultation function.The main research work is as follows:(1)Collect 60 common oral clinical cases,construct them into a dental case library,and store them in the database.Extract physician questions from the case database and create a dental question dataset for training text matching models.In response to the problem of insufficient training corpus in the question dataset,this article adopts data augmentation to expand the data and provide sufficient training corpus for the pre trained model.(2)This paper proposes a text matching task based on the RoBERTa-BiLSTM pre-training model,and conducts comparative experiments in the open domain dataset LCQMC,the medical domain dataset cMedQQ,and the self-made stomatological question dataset.Among them,in the self-made oral medical question data set,the ratio of training set,verification set,and test set is 8:1:1,and the ratio of positive and negative samples is close to 1:1.The accuracy of the model is higher than that of ABCNN,BIMPM,BERT,AlBERT and RoBERTa respectively increased by 10.82%,5.41%,1.46%,2.63% and 0.73%,and finally reached 92.25%.(3)In order to design and implement the stomatological consultation and training system,this paper mainly introduces the design and implementation process of the system in detail from the aspects of the overall design of the system,database design,and the implementation of functional modules,and conducts functional and non-functional tests on the system.After that,the system is deployed online in a containerized manner,and finally allows users to complete the questionnaire,providing valuable advice for improving the system.The core function of the system is the consultation training function,which uses a trained text matching model to find questions similar to user input from the case database,and returns the corresponding patient answers to the user interface,thereby achieving simulated consultation and improving the doctor’s consultation ability.