生成式人工智能分析工具在高考评价中的应用:以Kimi为例  

Research on the Application of Generative Artificial Intelligence in the Evaluation of the New College Entrance Examination:Taking Kimi as an Example

作　　者：许志勇[1] 范英军[2] Xu Zhiyong;Fan Yingjun(Tianjin Municipal Educational Admission&Examinations Authority,Tianjin,300387;Tianjin No.41 High School,Tianjin,300204)

机构地区：[1]天津市教育招生考试院,天津300387 [2]天津市第四十一中学,天津300204

出　　处：《考试研究》2024年第6期42-50,共9页Examinations Research

基　　金：教育部教育考试院“十四五”规划支撑专项课题“基于学科核心素养的新高考分数标准参照解释的研究”(课题批准号:NEEA2021043)研究成果之一。

摘　　要：选用Kimi作为生成式人工智能AIGC分析工具,以某市2023年高中学业水平等级性考试历史学科试卷和考生答题卡为分析内容,探讨AIGC在高考评价中的应用。研究表明:Kimi协同工作模式可以优化试卷结构分析,提升考生水平临界分数划定效度,提高评价结果反馈的时效性和针对性;但存在问题判定过分广泛、不能深入推理发现材料之间的关系、有时会给出错误结论或理由、无法明确进行微调等问题。Chosing Kimi as the analysis tool,and the test paper of History subject and candidates’answer sheets of 2023 High School Academic Level Grade Examination of a city as the analysis contents,this paper focuses on the application of generative artificial intelligence in the evaluation of the new college entrance examination.The study shows that Kimi collaborative working mode can optimize the structure analysis of the examination paper,enhance the validity of the demarcation of the critical score of the candidates’level,and improve the timeliness and relevance of the feedback in the evaluation results so as to improve the validity and efficiency.Some problems,such as problem judgment is too broad,it is unable to reason deeply to discover relationships between materials,sometimes it gives wrong conclusions or reasons,it is not sure how to fine-tune,should be paid attention to under the current use of AIGC.

关 键 词：生成式人工智能 高考评价 试卷结构分析 评价反馈 

分 类 号：G424.74[文化科学—课程与教学论]