Automatic Grading Of Junior Algebra Subjective Questions Baesd On Computational Reasoning

The transition from education to intelligence is one of the focuses of current research in the field of education,which aims to apply artificial intelligence and machine learning technology to the field of education to improve teaching effectiveness,improve teaching efficiency,promote intelligent,personalized,and fair education.In particular,the innovative application of automatic test paper marking effectively applies artificial intelligence technology to the field of education,aiming to achieve step level test paper marking and automatic grading of student answers through machine learning and natural language processing technologies,in order to improve efficiency and accuracy.This thesis mainly studies the automatic grading of algebra subjective questions in junior high school mathematics,based on the correct and wrong judgment of students’ answers in junior high school algebra and the logical judgment between the answering steps,and studies the representation of basic knowledge of junior high school algebra,the unification of algebraic symbols,and the semantic understanding of questions and the answering process using natural language understanding(NLP)technology;Research on correct and incorrect judgment based on mathematical expressions;Research logic judgment based on student pre sequence steps.The main research contents are as follows:1.Algebraic expression knowledge representation in junior high school.First of all,it is necessary to evaluate junior high school algebra,and it is necessary for the system to correctly recognize and understand the semantic information of these expressions.The main research content of this article is to convert mathematical expressions into a syntax tree structure,and on this basis,perform formal comparisons and determine the equivalence of values.2.Natural language understanding of elementary algebraic texts.Based on the atlas of junior high school mathematics knowledge,natural language processing technology is used to extract mathematical entities such as algebraic equations,functions,polynomials,and other mathematical entities in junior high school algebra,as well as relationships between entities such as equation relationships,equation group relationships,and inverse number relationships,and convert natural language text descriptions into entity relationship triplets.3.Algebraic expression correctness judgment based on standard answers.For students with standard answers,research is conducted to determine whether they are right or wrong based on the mapping and matching between student answers and standard answers.4.Computational reasoning logic judgment based on student pre sequence steps.For student answers that do not have a standard answer or have a low degree of matching with the standard answer,by analyzing the internal logical structure of the student’s answer,research logic judgment methods based on pre sequence steps,perform local logic and global logic judgment,and then synthesize the correct and wrong judgment results to obtain the final judgment output result.Based on the above method,a step level automatic grading system for subjective junior high school algebra questions was designed and implemented.The system was tested on junior high school algebra data sets,and the grading accuracy was 94%.