Design And Implementation Of Teaching Assistant Platform Based On Intelligent Review

The online homework review system is a teaching network platform that provides services for class students and teachers.Teachers use this network platform to arrange homework,and the system distributes homework to designated classes or students.Teachers can obtain the statistics of assignments,select submitted assignments for review,and provide feedback to students.Students bind their account to the class,obtain all coursework online,and the system updates the homework status while submitting the homework,effectively avoiding missing homework and realizing interactive processing of homework.The teaching assistant platform uses the web terminal as the background management system,adopts the micro-service architecture and Vue.js framework for development,manages the education bureau,schools,classes,teachers and students and application information,and explicitly controls the enterprise We Chat applet and We Chat applet.As a working platform for teachers,the main functions of enterprise We Chat include homework assignment,homework details,homework statistics and homework review,etc.;as a learning platform for parents and students,We Chat mini-programs mainly include homework list,homework completion and homework review.Job wall,etc.;the back end adopts Spring Boot framework,and the database adopts My SQL,Redis and Mongo DB to meet the needs of big data storage and real-time cloud computing of the whole system,and is committed to providing teachers and students with complete online homework full-scenario solutions.The teaching assistant platform provides teachers with a math quick calculation intelligent review service.The service mainly includes three modules,namely text detection,text recognition and calculation deduction.The text detection module of this service adopts DBNet based on segmentation algorithm,and the CBAM attention mechanism module is added to the network structure to improve the connection of each feature in channel and space,which is more conducive to extracting the effective features of the target;the text recognition module of this service Using CRNN and CTC frameworks,for the scene of distorted fonts in primary school homework,STN(Spatial Transformation Network)is added to process distorted characters,and Bi LSTM and the improved En CTC model are used to improve the recognition ability of the network;the calculation and deduction module of the service is designed And implemented a custom string calculator to calculate the text of the calculation questions identified in the text recognition stage to verify the correctness of the calculation results.This mathematical quick-calculation intelligent review service can automatically identify and grade assignments,with a correction accuracy of 98%.