摘要
当前新一代信息技术迅猛发展,人工智能已成为变革教育评价的重要力量。人工智能赋能教育评价可以提升智能教育评价的可信度与有效性,实现教育评价规模化与个性化的统一。智能教育评价依赖于教育知识与教育评价知识的归纳与表征,具有教育专家评价知识引导的特质。在知识引导智能教育评价中,教育评价知识标准与规范形成筛选机制,引导智能教育评价的数据感知与融合;教育评价知识刻度与分级标准形成匹配机制,引导智能评价模型选择与参数控制;教育评价知识的目标与关联形成反馈机制,引导评价结果的知识建构,生成新的可信评价知识。通过人机协同形成粗粒度专家知识与细粒度数据相结合的教育评价指标体系,进而构建知识引导数据筛选与模型选择机制和知识约束的神经网络模型参数控制机制是实现这一新型教育评价的重要路径。
With the current rapid development of the new generation of information technology,artificial intelligence technology has become an important force in transforming educational evaluation.Artificial intelligence empowering education evaluation can enhance the credibility and effectiveness of intelligent education evaluation,and realize the unity of education evaluation scale and personalization.Intelligent educational evaluation relies on the generalization and characterization of educational knowledge and educational evaluation knowledge,with the qualities of knowledge-guided evaluation by educational experts.The knowledge-guided intelligent educational evaluation consists of a screening mechanism formed by the standards and specifications of educational evaluation knowledge to guide the perception and integration of data for intelligent educational evaluation;a matching mechanism formed by the scales and grading standards of educational evaluation knowledge to guide the selection and parameter control of the intelligent evaluation model;and a feedback mechanism formed by the goals and associations of the educational evaluation knowledge to guide the knowledge construction of the evaluation results and the generation of new credible evaluation knowledge.Establishing an educational evaluation index system combining coarse-grained expert knowledge and fine-grained data with human-machine co-construction,and then constructing knowledge-guided data screening and model selection mechanism and knowledgeconstrained neural network model parameter control mechanism is the important path to implement this new education evaluation.
作者
周东波
赵帅
肖明
ZHOU Dongbo;ZHAO Shuai;XIAO Ming(the National Engineering Research Center of Educational Big Data,Faculty of Artificial Intelligence in Education,Central China Normal University;the Faculty of Artificial Intelligence in Education,Central China Normal University;the Office of Information Technology,Central China Normal University,Wuhan 430079)
出处
《教育研究与实验》
CSSCI
北大核心
2023年第4期118-127,共10页
Educational Research and Experiment
基金
科技创新2030新一代人工智能重大项目“混合增强在线教育关键技术与系统研究”(2020AAA0108804)
国家自然科学基金重大项目“人工智能赋能教与学的基础理论与关键技术研究”(62293550),国家自然科学基金面上项目“大学生个性化日常行为习惯的可解释发现框架与数据驱动干预机制”(62177017)的研究成果
关键词
教育评价
知识引导
数据驱动
人工智能
educational evaluation
knowledge-guided
data driven
artificial intelligence