融合多模态知识解析与大模型生成的智能考试系统设计与实现

打开文本图片集

中图分类号：TP311 文献标识码：A 文章编号：2096-4706（2025）20-0079-09

Abstract： Taking the course of Database Principles and Applications as the application scenario，this paper designs and implements an inteligent examination systemthat integrates multimodal knowledgeanalysisandlarge model technology.The system covers threecore functional modules： intelligent questionseting，AI markingand intelligent questionanswering.In terms ofimplementation，thefront end is developed using the Vueframework，andthe back endis builtbasedonthe Spring Bootarchitecture.Among them，the intellgentquestion seting function isrealized by CozeandDoubao large model.The automaticscoringofsubjectivequestionsiscompletedbycustomscoringrulescombined withthesemantic understandingability of DeepSeek large model.The intellgent question answering system is based on the Dify platformand the Tongyi Qianwen large model.Thetestresultsshowthatthesystemcan effectivelyreduce theburdenof teachers'teaching management，mprove students'autonomous learning experience，and provide a practical technical solution for smart education.

Keywords： intellgent examination system; multimodal knowledge analysis; large model automatic scoring; knowledge base question answering

0 引言

在线考试系统作为教育信息化的重要组成部分，已广泛应用于各类课程教学评估。（剩余11430字）