Research On Automatic Scoring Algorithm For Subjective Questions In Electrical Field Based On Syntactic Analysis And Ontology

With the rise of technologies such as computers,big data,and artificial intelligence,smart education uses modern information technology to realize the information and intelligence of education.As an important part of wisdom education,automatic grading can not only reduce the workload of teachers in the process of marking papers but also avoid the problem of grading deviation caused by personal subjective factors.In the field of electrical engineering,smart education is applied to college teaching,examinations for registered electrical engineers,and induction training for power grid employees,which helps trainees improve their knowledge system in the electrical field.At this stage,the research on the automatic scoring algorithm for objective questions has been very mature,but the research on the automatic scoring algorithm for subjective questions is still not perfect.For disciplines in the electrical field,the existing automatic scoring algorithms for subjective questions do not fully understand the semantics of professional terms in the electrical field,and there is inaccurate word segmentation of professional terms,large errors in semantic analysis,and incomplete consideration of semantic relations in the calculation of sentence similarity,etc.problem.This paper deeply researches the automatic scoring algorithm for subjective questions and analyzes the advantages and disadvantages of the automatic scoring algorithm in detail.At the same time,combined with the research project of "Research and Practice of Intelligent Correction System in Smart Campus Environment" participated in the postgraduate period,it is proposed based on syntactic analysis and ontology.The automatic scoring algorithm for subjective questions in the electrical field.The research work of this article is as follows:(1)Combining the characteristics of the electrical field,constructing the electrical field ontology.In this paper,electrical domain ontology is introduced into the automatic scoring process of subjective questions,and proposing a method for calculating the similarity of professional terms based on the ontology,which solves the problem of adequate semantic understanding and incomplete word segmentation of professional terms in specific fields.(2)Aiming at the ubiquitous polysemous phenomenon of common words,a word similarity algorithm combining "Synonyms Cilin" and "How Net" is designed to effectively improve the accuracy of the calculation results of common words similarity.(3)This paper proposes an automatic scoring algorithm combining word order similarity,word shape similarity and collocation pair similarity.According to the results of syntactic analysis,a method for extracting collocation word pairs is designed,which can obtain the internal logical relationship and deep semantic relationship between sentence components.At the same time,the algorithm uses the surface structure information of the text,which solves the large error of semantic analysis and the consideration of semantic relations in sentence similarity calculation Incomplete problem.(4)Based on the proposed scoring algorithm based on syntactic analysis and ontology,an automatic scoring prototype system in the electrical field is realized.Finally,we verify the effectiveness of the algorithm through experiments.The experimental results show that the scoring algorithm based on syntactic analysis and ontology can more accurately calculate the similarity between the student’s answer and the reference answer,making the algorithm’s scoring result closer to the manual scoring result.