Design And Implementation Of Student Portrait Generation System

In recent years,education informatization has developed rapidly.In order to effectively monitor students’ learning progress and recommend suitable learning content for them,education data mining has gradually emerged.Among them,student portraits,as a key component of education data mining,play an important role in the field of education.This paper conducted in-depth research on the construction and generation of student portrait models,and found that there are the following problems in the current stage of student portrait research: computer science experts mainly study portraits from the cognitive level,such as knowledge point recommendation and grade prediction,lacking in-depth research on student portraits? education experts propose portraits with depth,but are limited to theoretical aspects and lack technical implementation.Therefore,this paper combines knowledge from both fields to propose a comprehensive and multi-dimensional student portrait model,which not only shows the diversity of students’ competencies,but also explains their academic performance.At the same time,the paper generates data for each dimension of the portrait through machine learning algorithms,and visualizes the portrait data in the system,enabling teachers to intuitively view student portraits and better understand their learning situation.The specific main work of this paper is as follows:1.Creatively design the dataset.In response to the problem of messy education data,this paper used labeling to process the knowledge points and problem data of high school mathematics courses.Then,the labels were imported into the questionand-answer platform to collect various information about students to obtain basic attributes,and complex attributes were constructed in combination with question labels.Finally,the data was processed and cleaned to pave the way for portrait generation.2.Construct student portraits based on educational theories.In response to the problem of insufficient depth in portrait construction in computer science,this paper builds a comprehensive and in-depth portrait based on educational theories.The portrait includes ten dimensions,using the ”grade dimension” to measure students’ academic performance,and using nine core competency dimensions of ”mathematical modeling and abstraction,mathematical operations,logical reasoning,focus,creativity,learning attitude,stress resistance,emotional expression,and confidence level” to represent key features of students’ learning and growth,while explaining their academic performance.3.Predict student performance dimension based on the Boruta-PSO-XGBoost model.This paper used the Boruta algorithm to select features,on the one hand,to screen out important features,and on the other hand,to verify the rationality of complex attributes.Then,the PSO-XGBoost algorithm was used to predict the performance dimension,and comparison experiments and ablation experiments were conducted with different models.The model proposed in this paper is superior to traditional machine learning models in precision,recall,and F1 score.4.Predict student core competency dimensions based on a hybrid neural network model.In order to improve the accuracy of core competency prediction,this paper used an FNN network to process static data and a CNN-LSTM network to process dynamic data,and the two were combined to generate a hybrid neural network model.The model proposed in this paper is superior to basic deep learning models in precision,recall,F1 score,and accuracy.5.Design and implement a student portrait generation system.The paper divides the system into an information input module,a portrait generation module,and a front-end display module.In the system development process,the paper used the Spring Boot framework to build and implement various functions.To ensure that the system’s performance meets the requirements,the paper conducted functional testing and non-functional testing to ensure that it meets the expected performance standards,thus providing a high-quality and stable user experience.