Networked teaching data analysis method and system based on multiple modes

Abstract

The invention belongs to the technical field of data processing, and discloses a networked teaching data analysis method and system based on multiple modes, and the method comprises the steps: carrying out the feature screening through employing a maximum information coefficient MIC, and removing irrelevant factors; after MIC analysis is used for feature screening, screened features form a featurespace again, a random forest is used for regression, and a final evaluation model is obtained; a method of combining a learning analysis technology and a data mining algorithm is adopted; according to the method, learning ability data, physiological data and learning behavior data generated when students learn on a theoretical course online learning platform are integrated and analyzed, a theoretical online course learning effect evaluation model is established, the learning effect of the students is evaluated, and an evaluation result is output in the forms of graphs, numbers and the like byapplying a visualization technology. According to the invention, a machine learning technology is utilized to establish a multi-modal information fusion theoretical course evaluation system, and theoretical and technical method support is provided for an online course learning process.

Description

technical field [0001] The invention belongs to the technical field of data processing, and in particular relates to a method and system for analyzing networked teaching data based on multimodality. Background technique [0002] At present, the closest existing technology: network teaching is different from the traditional teacher-student face-to-face teaching mode, and it is of great significance to build an effective evaluation system model for network courses. At present, a large number of studies have used learning analysis techniques such as correlation analysis, regression analysis, and data mining algorithms to collect, measure, analyze, and report students' learning behavior data, understand and optimize the teaching process and situations, and provide support for teaching decisions and academic warnings. Improve teaching effect. [0003] However, the existing technology mainly involves the acquisition and analysis of students' learning behavior data. At the same ti...

AI Technical Summary

Problems solved by technology

[0005] (1) The existing technology mainly involves the acquisition and analysis of students' learning behavior data, and the evaluation model established at the same time is relatively fixed, and there are certain evaluation prediction errors in different batches and environments

[0006] (2) In the existing network online education, teachers and students are in a state of quasi-separation, and students mostly study alone geographically, lacking the emotional attention of teachers, and it is difficult to communicate with other students in depth, and they cannot experience the classroom presence and sense of traditional education The sense of collective belonging strengthens the students' sense of loneliness, which can easily lead to learning burnout; affect the learning effect of students

[0007] (3) The existing technology mainly involves the acquisition, analysis and evaluation of student learning behavior data, which is lagging and difficult to actively intervene in the learning process of students

[0012] Applying the evaluation model based only on the first batch of data to the subsequent student evaluation will directly cause the instability of the evaluation results, and because learners will continue to generate dynamic big data during the learning process, the generation of these data means the richness of feature data , but it also brings the complexity of feature analysis, how to determine the impact factors of different batches of the same feature on the final model construction, the evaluation and comparison of models combined in different combinations in different batches, and how to base on These evaluation models based on different combination methods are used to screen quantitative features, which require us to have a rigorous logical structure

Images