Design And Implementation Of Automatic Evaluation System For English Learning Situation

Examination has been an effective means of evaluating students’ learning situation for a long time.However,in the process of traditional manual scoring and modern online scoring,there are problems such as heavy workload for teachers to grading papers,and the evaluation of students’ learning situation can only be done manually by teachers,which adds extra workload for teachers.Under the background of "double reduction",how to effectively improve the quality of teaching while reducing the burden effectively puts forward higher requirements for teaching work,which requires teachers to have a more accurate cognition of students’ learning situation,so as to carry out targeted teaching and truly achieve the purpose of teaching students in accordance with their aptitude.Through the analysis of the whole process of examination and learning evaluation,combined with artificial intelligence technology,the automation of scoring and evaluation is realized.It is expected to effectively reduce the workload of teachers in scoring and evaluation,and help teachers to better grasp the learning situation of students,and devote more limited working time to the design and optimization of the teaching process,so as to achieve the goal of reducing the burden without reducing the quality under the background of "double reduction".Combined with the English subject,research and analyze the evaluation objectives and evaluation methods of the English examination.Focusing on the four teaching objectives of vocabulary,grammar,reading comprehension,and written expression,research and analyze the manual evaluation standards and automatic evaluation related technologies for the five types of questions: multiple choice,blank filling,short answer,translation and composition,and design and build the evaluation characteristics and automatic evaluation models for each type of question.The evaluation features of short answer questions and translation questions are extracted from literal,structural and semantic aspects,and machine learning models are selected to construct corresponding automatic evaluation models.The evaluation features of composition questions are extracted from three levels: deep,shallow and theme,and an automatic evaluation model of composition questions is constructed by combining deep learning and statistical methods.An automatic evaluation system of English learning situation is designed and implemented on the basis automatic evaluation methods of each question type and in combination with the actual evaluation needs of English subjects.Based on the students’ answers to the English test papers,the students’ English learning situation is automatically evaluated from the perspective of question type and ability target,and corresponding evaluation reports are output.The feasibility of the proposed automatic evaluation method is verified through experiments,and the test verifies the availability of the automatic evaluation system for English learning,and it can better meet the requirements of actual evaluation.