Design And Implementation Of Big Data Basic Teaching Platform Based On Automatic Judgment

In order to cope with the current situation of big data talent shortage,some domestic universities have successfully applied for the undergraduate degree of big data and set up a big data curriculum system for undergraduates.The course "Big Data Foundation" aims to cultivate familiarity with the process of big data processing,and set up corresponding experimental courses for school hours.Due to the lack of experimental content that meets the teaching objectives of Big Data Foundation and the need for teaching management and real-time scoring,it is of practical value to design and develop a teaching platform based on automatic scoring for Big Data Foundation.Starting from the teaching objectives of the course "Big Data Basis",six experimental contents covering the whole life cycle process of big data development are designed to improve the understanding of big data technology and to train their programming and practical abilities.According to the needs of teaching and experiment,the platform designs modules of teaching management,experiment management and automatic grading,focusing on the design and implementation of automatic grading.According to the requirements of teaching and experiment,the system architecture and functional modules of the large data basic teaching platform based on automatic scoring are designed.The paper focuses on the design and implementation of automatic scoring scheme.According to the characteristics of the experiment content and imitating the idea of artificial evaluation,the overall scoring scheme based on the combination of dynamic and static is put forward: the result score is obtained by dynamic testing of the program;the point score and structure score are obtained by static analysis of the program;and the final score is obtained by synthesizing the three-party score.According to the design of experiment content and the complexity of experiment result,similarity and DB measure are used as the result comparison scheme.Compared with one-to-one comparison method,the efficiency and rationality of dynamic test part are improved.In the static analysis,the code itself is judged in an all-round way by a scoring scheme combining the key points scoring with the structural scoring.In structure scoring,the code is first transformed into abstract grammar tree,and the abstract grammar tree is de-redundant according to user-defined rules,and matching score is made by tree editing distance.The matching algorithm is improved and the sub-tree is divided to reduce the time complexity of scoring scheme.By testing the teaching platform of Big Data Foundation based on automatic scoring,the expected requirements have been fulfilled,and the use of the platform in two semesters of actual teaching has also been recognized by teachers and students.