语义、语用“双管齐下”:优质信息技术课堂互动话语自动化分析及特征发现  

Semantic and Pragmatic Dual Approach:Automated Analysis and Feature Discovery of High-Quality Classroom Interaction Discourse in Information Technology

作　　者：胡艺龄[1] 汪沁婵 吴忭[1] 彭俊 HU Yiling;WANG Qinchan;WU Bian;PENG Jun(East China Normal University,Shanghai,200062;City University of Macao,Macao,999078)

机构地区：[1]华东师范大学,上海200062 [2]澳门城市大学,中国澳门999078

出　　处：《现代远距离教育》2025年第1期77-87,共11页Modern Distance Education

基　　金：全国教育科学“十四五”规划2023年度国家一般项目“数据赋能的教师数字素养培养模式与可视化环境设计研究”(编号:BCA230274)。

摘　　要：作为教育学领域中的经典研究议题,传统的教师课堂话语分析普遍聚焦话语的语义特征而忽视语用特征,导致话语形式特征丢失、对教师话语策略归纳片面。对此,本研究首先以已有教师课堂话语研究为基础,总结出一套语义和语用“双管齐下”的编码框架体系。然后,对收集到的高中信息技术课堂教师互动话语,运用无监督学习模型和预训练深度模型等自然语言处理技术完成教师互动话语的自动编码,其语用编码和语义编码准确率分别达到81.35%和72.47%。最后,总述优质信息技术课堂中教师话语的语用和语义特征,并聚类分析三个类型课堂中教师的常用话语策略,为新课标、新教材下的高中信息技术教学提供实践参考。As a classic research topic in the field of pedagogy,traditional analysis of teacher pedagogical discourse has generally focused on the semantic features of discourse while neglecting its pragmatic characteristics,leading to a loss of formal discourse features and a partial understanding of teachers’discourse strategies.In response,this study first builds on existing research on teacher pedagogical m discourse to develop a dual-coding framework that addresses both semantic and pragmatic features.Then,using natural language processing techniques,such as unsupervised learning models and pre-trained deep learning models,we automatically code teacher interaction discourse collected from high school information technology classes.The accuracy rates for pragmatic and semantic coding reach 81.35%and 72.27%,respectively.Finally,we summarize the semantic and pragmatic features of teacher discourse in high-quality information technology classrooms and perform cluster analysis on the common discourse strategies employed by teachers in three types of classes,providing practical insights for high school information technology teaching under new curricula and textbooks.

关 键 词：教师话语 话语分析 自然语言处理 语用 语义 

分 类 号：G43[文化科学—教育学]