Research On Automatic Analysis Of Online Collaborative Learning Interaction Quality Based On Text Mining

Interaction is the basic unit and key of online collaborative learning activities and the core element of learning environment.Interaction analysis is the core of exploring the law of collaborative learning and understanding the dynamic changes of interaction process.At the same time,it is also the diagnostic basis for learners to master the goal of collaborative learning.The proposal of "Li Kedong problem" makes educational scholars realize that text interaction lacks in-depth analysis and discussion,and the interaction quality is not ideal.How to realize the in-depth analysis of text interaction process and help improve the quality of text interaction has become the focus of research.Facing the current situation of interactive quality analysis of online collaborative learning,there are incomplete analysis theories,different analysis methods and lack of automation tools.This research is committed to realizing the automatic analysis of interactive quality of online collaborative learning.Taking "theoretical model algorithm model platform development" as the main line,this research attempts to implement the automatic analysis of interactive quality from the theoretical level to the practical level by designing the index system for interactive quality evaluation,constructing the analysis method with good performance,and developing the tools and platform for automatic analysis.In the theoretical model design stage,taking the inquiry community as the core theoretical basis,clarify cognitive interaction,teaching interaction,social interaction and emotional interaction as the analysis dimensions of interaction quality,and determine the indicators to measure different analysis dimensions,so as to provide a clear analysis direction for interaction quality.At the same time,in order to realize the effective implementation of interaction quality analysis dimension and its measurement index from theory to practice,an interaction quality analysis coding framework for online collaborative learning is designed under the guidance of analysis dimension and measurement index.The online collaborative learning interaction quality analysis index system provides new ideas and analysis guidance for the current online learning interaction quality automatic analysis,and provides a theoretical basis for the subsequent algorithm model construction and platform development.In the algorithm model construction stage,the text classification technology of machine learning is adopted to obtain the algorithm model with the best performance of text classification in each analysis dimension of interaction quality through the steps of data acquisition,text classification and annotation,data preprocessing,word embedding,model training and model evaluation,so as to provide technical support for the realization of interactive text automatic classification,Finally,the interactive quality calculation method is designed to achieve the purpose of interactive quality calculation and provide the calculation basis for measuring the interactive quality.In the development stage of the visualization platform,combined with the constructed analysis index system,algorithm model,visualization technology and web page development technology,an automatic analysis platform for the interactive quality of online collaborative learning is constructed.The platform realizes the automatic analysis of the interactive quality and the visualization of the analysis results.Teachers can grasp the interactive quality in real time through the platform,so as to provide scientific and powerful support for the subsequent management of the interactive process of online collaborative learning.