Discovery Of Teaching Behavior Model And Teaching Practice Based On Data

With the construction and development of smart campuses,education big data is growing exponentially,and data-driven education-related research is also rapidly developing.Therefore,a perfect education data era research system is established to realize data-driven model discovery and teaching innovation to enhance Teaching quality has important theoretical value and practical significance for the deep integration of education informatization.However,there are still few data-driven research methods in data-driven education research.The existing results are more focused on online education teaching research and student results data research.The research on offline smart classroom teacher teaching process data is less..Based on this,this study analyzes the process data collected from the self-developed teaching platform tool to explore the teaching behavior mode.The main work is as follows:First,based on the research on the educational function and type of teaching behavior,paying attention to the editing behavior and non-editing behavior of teachers when applying teaching tools,based on the accompanying collection of teaching behavior records in smart classrooms,the behavior is divided into digital teaching tool behavior,digital labeling tool behavior and Other tool behaviors,combined with the sequence behavior,interactivity and different sequence length characteristics of the teaching behavior sequence,are extracted as a series of teaching behavior sequence feature vectors,using DTW to complete the similarity measurement between the different length feature vectors,and generating DTW for all data sets Cumulative distance matrix realizes the division of teaching behavior patterns through clustering of cluster centers.Secondly,according to the result of teaching behavior clustering,the teaching behavior path is visualized on each cluster of teaching behavior sequence,and a directed graph of the teaching behavior path is obtained.It is found that there is a single-tool unlabeled application mode and single-tool labeled application in the application of the teaching platform by teachers There are five application modes including mode,dual-tool independent application mode,dual-tool associated application mode and multi-tool associated application mode.Different application modes reflect the difference in the effective application level of teachers to teaching tools.Thirdly,through the analysis of classroom teaching practice,the validity of the analysis of the process data of the teaching mentioned in this article is verified,and the rationality of the division of the teacher's application model and the interpretability of the corresponding teaching scenarios are further explored.Finally,through the development of teaching activities supported by the teacher's application model,and the evaluation of the effective application of teaching tools by the directed graph of the teaching behavior path,it caused the transformation of the form of learning from "passive acceptance" to "active discussion" in the classroom.The display form is changed from "individual spot check" to "all participation",and the effective application of teacher teaching tools is changed from "questionnaire result" to "mining process" evaluation,thereby promoting the role of teachers and students,classroom forms,Demonstrate the reconstruction of feedback,teaching tools,and teaching activities.The research results of this paper have been integrated into the self-developed classroom analysis system,and data-driven classroom analysis research and development have achieved preliminary application.