Preschool education teaching quality evaluation system based on data analysis

Abstract

The invention provides a preschool education teaching quality evaluation system based on data analysis, and relates to the technical field of preschool education and artificial intelligence combination, and the system comprises a data collection module which is used for collecting basic information, teaching behavior data and teaching effect information of teachers; the knowledge graph construction module is used for constructing a multi-dimensional knowledge graph based on the data collected by the data collection module; the teaching behavior analysis engine module is used for analyzing the relationship between the teaching behavior and the teaching effect of the teacher based on the data in the knowledge graph by utilizing a teaching behavior analysis model; the teaching behavior analysis model is constructed based on a graph convolutional neural network; and the teaching behavior analysis model takes the knowledge graph as input and outputs an evaluation result of teaching quality. According to the method, the teaching behavior knowledge graph for preschool education teachers is constructed, and the teaching behaviors, the teaching methods and the teaching effects of the teachers are comprehensively recorded and analyzed, so that the teaching quality of the teachers is objectively evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of the combination of preschool education and artificial intelligence, and in particular to a preschool education teaching quality evaluation system based on data analysis. Background Art

[0002] In the field of preschool education, the teaching quality of teachers directly affects the learning outcomes and all-round development of young children. Traditional teaching quality assessment methods often rely on administrative observations, peer reviews or parent feedback, which have certain limitations. For example, administrative observations may only be able to observe fragments of teachers' teaching due to time constraints, peer reviews may be affected by subjective judgments, and parent feedback may lack professionalism and systematicness. In addition, these traditional methods are usually difficult to record and analyze teachers' teaching behaviors and teaching methods in real time, and it is also difficult to provide specific and personalized suggestions for teaching improvements.

[0003] Although there are some educational technology tools that claim to be able to evaluate teaching quality, these tools tend to focus on the analysis of student learning outcomes rather than on the in-depth analysis and evaluation of teachers' teaching behaviors. Therefore, developing a system that can comprehensively and objectively evaluate teachers' teaching quality and provide specific improvement suggestions is of great significance to improving the quality of preschool education. Summary of the invention

[0004] In view of this, the present invention provides a preschool education teaching quality evaluation system based on data analysis. By constructing a teaching behavior knowledge map for preschool education teachers, it comprehensively records and analyzes teachers' teaching behaviors, teaching methods and teaching effects, thereby objectively evaluating teachers' teaching quality and providing teachers with personalized teaching improvement suggestions, aiming to achieve scientific evaluation and effective guidance of the entire teaching process of teachers.

[0005] To this end, the present invention provides the following technical solutions:

[0006] The present invention provides a preschool education teaching quality evaluation system based on data analysis, the system comprising:

[0007] Data collection module: used to collect teachers’ basic information, teaching behavior data and teaching effect information;

[0008] Knowledge graph construction module: constructs a multi-dimensional knowledge graph based on the data collected by the data collection module; the knowledge graph includes: teacher information, teaching behavior, teaching effect and their mutual relationship;

[0009] Teaching behavior analysis engine module: using the teaching behavior analysis model, based on the data in the knowledge graph, the relationship between the teacher's teaching behavior and teaching effect is analyzed; the teaching behavior analysis model is constructed based on the graph convolutional neural network; the teaching behavior analysis model uses the knowledge graph as input and outputs the evaluation results of the teaching quality.

[0010] Furthermore, the teaching behavior data includes: classroom interaction data, teaching strategy data and teaching material data.

[0011] Furthermore, the teaching effect data includes: student feedback data, parent feedback data and student performance data.

[0012] Furthermore, the data collection module is also used to clean, standardize and convert the collected data after it is collected.

[0013] Furthermore, a multi-dimensional knowledge graph is constructed, including:

[0014] Entity recognition and classification: Identify entities and their attributes from the collected data, including teachers, teaching behaviors, and teaching effects;

[0015] Relationship definition and extraction: determine the relationships between entities and extract these relationships from the data; the relationships between entities include: usage relationships, connecting teachers and specific teaching behaviors; influence relationships, connecting teaching behaviors and teaching effects;

[0016] Knowledge graph construction: Using graph database, the identified entities and relationships are constructed into a knowledge graph;

[0017] Entity linking and fusion: Entities from different data sources but representing the same real object are linked and fused.

[0018] Furthermore, constructing a multidimensional knowledge graph also includes: Knowledge graph update and maintenance: regularly update the knowledge graph to reflect changes in teachers' teaching behaviors as well as new teaching methods and teaching effectiveness data.

[0019] Furthermore, the teaching behavior analysis model is used to analyze the relationship between teachers’ teaching behavior and teaching effect based on the data in the knowledge graph, including:

[0020] Data preprocessing: converting the data in the knowledge graph into a format suitable for graph convolutional neural network processing, including constructing adjacency matrix and feature matrix;

[0021] Model training: training teaching behavior analysis model based on data preprocessed;

[0022] Teaching quality evaluation: The features learned by the trained model are used for classification or regression tasks to predict the teaching effect, and the teaching quality is evaluated based on the predicted teaching effect.

[0023] Furthermore, the teaching quality assessment also includes: formulating personalized improvement suggestions based on each teacher’s teaching behavior characteristics and corresponding analysis results.

[0024] Furthermore, the system also includes: a report generation module: automatically generates a comprehensive report including teacher teaching behavior analysis, teaching quality evaluation and improvement suggestions.

[0025] Furthermore, the system also includes: a visualization display module: visually outputting the report.

[0026] Advantages and positive effects of the present invention:

[0027] The present invention uses advanced knowledge graphs and artificial intelligence technologies to construct an innovative preschool education teaching quality evaluation system. First, the teaching behavior data of teachers are collected and sorted in detail, including but not limited to classroom management methods, interaction types, feedback frequency, etc., and these data are converted into a graphical knowledge structure, that is, a knowledge graph. On the basis of the knowledge graph, the present invention further uses a machine learning model to conduct an in-depth analysis of the data. The model is able to extract valuable information from massive data and identify key factors that affect teaching effectiveness. Through the analysis of the machine learning model, the teaching behavior of teachers can be fully recorded and deeply understood, thereby providing an objective and accurate teaching evaluation.

[0028] The preschool education teaching quality evaluation system of the present invention can not only conduct teaching evaluation, but also provide personalized teaching improvement suggestions based on the specific situation of each teacher. These suggestions are based on in-depth analysis and understanding of teaching behavior, aiming to help teachers optimize teaching methods and improve teaching effects. At the same time, the real-time update function of the system ensures that teachers can obtain the latest teaching feedback and suggestions in a timely manner, further improving the pertinence and effectiveness of teaching. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0030] Figure 1 It is a structural block diagram of a preschool education and teaching quality evaluation system based on data analysis in an embodiment of the present invention;

[0031] Figure 2 A flowchart of constructing a knowledge graph in an embodiment of the present invention;

[0032] Figure 3 This is a schematic diagram of a knowledge graph in an embodiment of the present invention;

[0033] Figure 4 Flow chart of teaching behavior analysis in an embodiment of the present invention. DETAILED DESCRIPTION

[0034] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0035] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0036] With the development of information technology, educational data analysis has shown great potential in improving teaching quality. The advancement of data collection technology makes it possible to collect a large amount of teaching behavior data at low cost and high efficiency, and the application of data analysis technology, especially knowledge graph technology, provides new possibilities for in-depth understanding of the teaching process and evaluation of teaching quality. In this invention, the information such as teachers' teaching behavior, teaching methods and teaching effects is organized in a structured manner through knowledge graphs, which is convenient for in-depth analysis and mining.

[0037] like Figure 1 As shown, an embodiment of the present invention provides a preschool education teaching quality evaluation system based on data analysis, the system comprising:

[0038] Data collection module 100: used to collect basic information of teachers, teaching behavior data and teaching effect information;

[0039] The basic information of teachers may include: the teacher's age, gender, educational background, teaching experience, qualification certificates, etc. It may also include the teacher's main subjects, teaching grade, cumulative teaching time, past teaching results, etc.

[0040] Teaching behavior data may include: classroom interaction data, teaching strategy data, and teaching material data. In specific implementations, audio and video sensors can be deployed to monitor the frequency and quality of interaction between teachers and students. For example, audio recognition technology can be used to detect the frequency of teachers' questions, the activeness of students' responses, and the depth and breadth of classroom discussions. The different teaching strategies and methods used by teachers in class can be recorded through video analysis or teaching logs. For example, record the time and effectiveness of teachers' teaching using different methods such as games, stories, and songs. Teaching material data, such as lesson plans, slides, teaching aids, etc., can evaluate the rationality of their design and the impact on student learning outcomes.

[0041] Teaching effectiveness data, including: student feedback data, parent feedback data and student performance data. In specific implementation, electronic questionnaires or interactive platforms can be used to regularly collect students' and parents' opinions and feedback on teaching, including students' preferences for teachers' teaching methods, the perceived difficulty of learning content, and evaluation of classroom atmosphere, and parents' satisfaction and opinions on teachers' teaching effectiveness. Student performance data includes students' academic performance, social skills, emotional development, creativity and other aspects.

[0042] After the above data are collected, the collected data will be preprocessed, including but not limited to: cleaning, standardization and conversion, to facilitate subsequent data analysis and knowledge graph construction. Specifically, the qualitative description of teaching behavior is converted into quantifiable indicators, such as interaction frequency, diversity of teaching strategies, etc.

[0043] Knowledge graph construction module 200: This module is responsible for constructing a multi-dimensional knowledge graph based on the data collected by the data collection module 100. This knowledge graph contains entities such as teacher information, teaching behavior, teaching effect and their interrelationships, which is used to deeply analyze and evaluate the teaching quality of teachers.

[0044] In specific implementation, Figure 2 As shown, a multidimensional knowledge graph can be constructed in the following way:

[0045] S11. Entity recognition and classification: Identify entities and their attributes from the collected data;

[0046] In this embodiment, entities include: teacher entity, teaching behavior entity and teaching effect entity. Among them, the attributes of the teacher entity include: age, gender, educational background, teaching experience, qualification certificate, teaching subject, teaching grade, cumulative teaching time, past teaching results, etc. The attributes of the teaching behavior entity include: classroom interaction frequency, teaching strategy type, teaching material use, question frequency, student response activity, classroom discussion depth and breadth, etc. The attributes of the teaching effect entity include: student feedback, parent feedback, academic performance, social skills, emotional development, creativity, etc.

[0047] S12. Relationship definition and extraction: Determine the relationships between entities and extract these relationships from the data.

[0048] For example, you can define a "use" relationship to connect teachers and specific teaching methods, indicating which methods the teacher uses in the teaching process; define an "influence" relationship to connect teaching methods and teaching effectiveness, indicating how certain teaching methods affect students' learning outcomes; define a "belongs to" relationship to connect students and classes, indicating which class a student belongs to. For example, extract the relationship that "Teacher Li" uses "group discussion" and "multimedia projection" as teaching methods from the data, and analyze how these methods affect the learning outcomes of "first-year" students. The teaching effectiveness can be reflected through the collected student feedback data and student homework data.

[0049] S13. Knowledge graph construction: Use graph database or other suitable data structures to construct the identified entities and relationships into a knowledge graph.

[0050] like Figure 3 As shown in the figure, according to the examples in the first two steps, a knowledge graph containing nodes such as "Teacher Li", "Group Discussion", "Multimedia Projection" and "Geometric Shapes" can be constructed. In this graph, the "Use" relationship connects "Teacher Li" with "Group Discussion" and "Multimedia Projection", while the "Influence" relationship connects these teaching methods with student learning outcomes.

[0051] S14. Entity linking and fusion: Entities from different data sources but representing the same real object are linked and fused to ensure the accuracy and consistency of the knowledge graph.

[0052] For example, if another teaching method "gamified learning" used by "Teacher Li" is found in another data source, this information needs to be linked and integrated with the existing "Teacher Li" entity.

[0053] S15. Knowledge graph update and maintenance: The knowledge graph will be updated regularly to reflect changes in teachers’ teaching behaviors, as well as the addition of new teaching methods and teaching effectiveness data.

[0054] As time goes by, data is collected on "Teacher Li" adopting a new teaching method, "flipped classroom", or the learning outcomes of students in the "first grade class" are improved. This new information needs to be updated in the knowledge graph in a timely manner.

[0055] Teaching behavior analysis engine module 300: Utilizes the teaching behavior analysis model to analyze the relationship between teachers’ teaching behavior and teaching effectiveness based on the data in the knowledge graph.

[0056] Among them, the teaching behavior analysis model can be constructed based on a graph convolutional neural network (GCN), which consists of an input layer, multiple graph convolutional layers, an activation function, a pooling layer, and an output layer. The graph convolutional layer is the core of the network, which updates the feature representation of the current node by aggregating the feature information of neighboring nodes. Each graph convolutional layer can be regarded as a filter that learns how to propagate and transform features on the graph structure. The depth of the graph convolutional network (i.e., the number of layers) depends on the complexity of the task and the scale of the data. In general, the network depth can range from a few layers to dozens of layers. For teaching behavior analysis, because the data set is relatively small, the task does not require an extremely high level of abstraction and does not require a too deep network. In the embodiment of the present invention, the experiment starts with a network of 2 to 5 layers and is adjusted according to the performance of the verification set.

[0057] In the specific implementation, the knowledge graph is used as the input of the teaching behavior analysis model, and the teaching behavior analysis model analyzes the impact of teaching behavior on teaching effect to obtain the evaluation result of teaching quality. Figure 4 As shown, you can follow the steps below:

[0058] S21. Data preprocessing: Convert the data in the knowledge graph into a format suitable for graph convolutional neural network processing, including constructing the adjacency matrix and feature matrix.

[0059] The adjacency matrix describes the direct connection relationship between vertices. In this embodiment, the adjacency matrix is ​​used to represent the relationship between the teacher entity, the teaching behavior entity and the teaching effect entity.

[0060] An example of constructing an adjacency matrix is ​​as follows:

[0061] Entity definition: Teacher entity: mark it as T. Teaching behavior entity: for the sake of specificity, assume that there are two teaching behaviors, marked as M1 (e.g. "group discussion") and M2 (e.g. "multimedia projection"). Teaching effect entity: consider two effects, marked as E1 (e.g. "student academic performance") and E2 (e.g. "student social skills").

[0062] Relationship definition: Assume that teacher T uses two teaching methods M1 and M2. Assume that M1 has a direct effect on E1 and M2 has a direct effect on E2.

[0063] Adjacency matrix construction: Based on the above definition, we can construct a 6×6 adjacency matrix (because there are 6 entities: T, M1, M2, E1, E2), where the rows and columns represent these entities, as shown in Table 1.

[0064] Table 1

[0065]

[0066]

[0067] explain:

[0068] The relations T to M1 and T to M2 are marked as 1, indicating that teacher T uses both teaching methods.

[0069] The relationship from M1 to E1 is marked as 1, indicating that M1 (group discussion) directly affects E1 (student academic performance).

[0070] The relationship from M2 to E2 is marked as 1, indicating that M2 (multimedia projection) directly affects E2 (students’ social skills).

[0071] This adjacency matrix provides a clear framework to show how the teaching behaviors used by teachers directly affect student learning outcomes. By analyzing such a matrix, we can better understand the impact of different teaching methods on learning outcomes and adjust teaching strategies accordingly.

[0072] Constructing a feature matrix: A feature matrix (sometimes also called an incidence matrix) is a matrix that describes the relationships between vertices and edges in a graph, where rows represent entities (vertices) and columns represent relationships (edges). In this embodiment, the feature matrix can represent the relationships between different elements in the teaching process.

[0073] An example of constructing a feature matrix is ​​as follows:

[0074] In the feature matrix, if the vertex is part of an edge, the corresponding matrix element is 1, otherwise it is 0. Therefore, the feature matrix shown in Table 2 can be constructed.

[0075] Table 2

[0076]

[0077]

[0078] explain:

[0079] The relationships T to M1 and T to M2 indicate that the teacher uses these two teaching methods, so the positions at row T and column M1 and row T and column M2 are 1.

[0080] The relationship between M1 to E1 and M2 to E2 indicates that the teaching method directly affects the learning effect, so the position in the M1 row and E1 column and the M2 row and E2 column is 1.

[0081] S22. Model training: training the teaching behavior analysis model based on the data after data preprocessing.

[0082] In the specific implementation, the constructed adjacency matrix and feature matrix are used as model inputs. GCN automatically learns high-dimensional feature representations of nodes (entities) through the information aggregation mechanism of neighbor nodes. These features capture the local structure and global position information of the network. Specifically, for each teacher node, GCN considers the information of its neighbor nodes (such as the students taught, the teaching methods used, the courses taught, etc.). By aggregating the features of these neighbor nodes, GCN is able to learn the comprehensive feature representation of the teacher, which includes the teacher's teaching style, preferred teaching methods and their effects. These automatically learned feature representations capture the local structure of each teacher in the teaching network, such as how active they are in teaching, how frequently they interact with students, etc., as well as global position information, such as the teacher's influence and teaching scope in the entire education system.

[0083] By using the multi-layer structure of GCN, complex dependencies between entities can be inferred. For example, by analyzing the features learned by multi-layer GCN, it is possible to identify how certain specific teaching strategies (such as gamification, art activities, role-playing, etc.) affect student performance. For example, it may be found that students are more engaged and have better homework completion when using role-playing teaching, indicating that this teaching strategy is positively correlated with improving student performance. It is also possible to capture the relationship between the difficulty of different courses and the teaching methods used. For example, for more difficult courses, more interactive teaching may lead to better student understanding.

[0084] When training the model, a cross entropy loss function is used, which can be binary cross entropy or multivariate cross entropy depending on the classification task. Adam is used for model optimization. The model performance is evaluated on the validation set, hyperparameters are adjusted, and then a final evaluation is performed on the test set.

[0085] S23. Teaching quality evaluation: By using the features learned by GCN for classification or regression tasks, the teaching effect is predicted, and the teaching quality is evaluated based on the predicted teaching effect.

[0086] In the specific implementation, the features learned by GCN are used to train the regression model to predict the specific impact of different teaching behaviors on teaching effectiveness, such as predicting the improvement of students' performance in standardized tests after adopting the gamified learning method. In addition to academic performance, other factors reflecting teaching effectiveness can also be predicted, such as the improvement of students' social skills and creativity development, so as to comprehensively evaluate the teaching quality. A teaching quality classification model is established to divide the teaching quality of teachers into "excellent", "good", "average" and other levels. By analyzing the teaching behavior of teachers and their corresponding teaching effects, the system automatically evaluates the teaching quality of the teacher and classifies it.

[0087] S24. Analysis and improvement suggestions: Based on each teacher’s teaching behavior characteristics and corresponding analysis results, formulate specific improvement suggestions.

[0088] For example, if the model finds that teacher A has a low interaction frequency, which has a negative impact on teaching effectiveness, it can provide A with suggestions on how to increase student interaction, such as through group discussions or interactive teaching activities.

[0089] When generating personalized recommendations, you can also break down the improvement suggestions into specific action steps, such as providing specific interactive activities, feedback mechanisms, or classroom management strategies. And set clear goals and implementation timelines for each improvement suggestion to track progress and effectiveness.

[0090] Report generation module 400: automatically generates a comprehensive report including teacher teaching behavior analysis, teaching quality evaluation and improvement suggestions.

[0091] Visualization display module 500: presents the analysis results in the form of charts, reports or dashboards, so that teachers and education administrators can intuitively understand the relationship between teaching behavior and teaching effect.

[0092] In the above embodiment, by integrating knowledge graphs and artificial intelligence technology, a comprehensive, objective and real-time solution is provided for the evaluation of preschool education teaching quality. This system can not only help teachers better understand and improve their teaching behaviors, but also provide strong decision-making support for preschool education managers, and jointly promote the development and progress of preschool education.

[0093] For ease of understanding, three specific application scenario examples are given below:

[0094] Application scenario 1: Teacher self-improvement

[0095] A preschool teacher wants to improve his teaching skills and classroom management ability. The preschool education teaching quality evaluation system based on data analysis in the above embodiment can be used to record his teaching behavior and teaching methods. Specifically, the teacher uses wearable devices or automatic recording and broadcasting systems in the classroom to collect his teaching behavior data. The system automatically uploads and records the teaching behavior data to the knowledge graph. The teacher views his teaching behavior analysis report through the system interface, including the frequency of interaction, the use of teaching methods, etc. The system provides personalized teaching improvement suggestions based on the results of the knowledge graph analysis, such as strategies to increase student participation. Teachers adjust their teaching plans and methods according to the suggestions and improve themselves.

[0096] Application scenario 2: School teaching quality monitoring

[0097] School administrators hope to monitor and improve the teaching quality of school teachers to ensure teaching standards and effectiveness. Specifically, the school deploys the preschool education teaching quality evaluation system based on data analysis in the above embodiment, and configures necessary data collection tools for each teacher. The system collects and integrates teachers' teaching behaviors, teaching methods, and student feedback data. Administrators view the teaching behavior knowledge map and teaching quality analysis report of all teachers in the school through the system interface. The system identifies areas that need improvement and provides decision-making support for administrators. Based on the analysis results and suggestions, administrators organize teacher training and teaching seminars.

[0098] Application scenario three: teacher training and development

[0099] Teacher training institutions want to provide customized teacher training courses based on empirical data. Specifically, the training institutions use the preschool education teaching quality evaluation system based on data analysis to analyze the early teaching behavior of teachers participating in the training. The system generates a teaching behavior analysis report and knowledge map for each teacher. The training institution designs a personalized training plan based on the analysis results. Teachers participate in the training and learn new teaching methods and strategies. After the training, the teachers apply the new strategies in their teaching, and the system collects data again and evaluates the results.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A preschool education teaching quality evaluation system based on data analysis, characterized in that: The system comprises: Data collection module: used to collect teachers’ basic information, teaching behavior data and teaching effect information; Knowledge graph construction module: constructs a multi-dimensional knowledge graph based on the data collected by the data collection module; the knowledge graph includes: teacher information, teaching behavior, teaching effect and their mutual relationship; Teaching behavior analysis engine module: using the teaching behavior analysis model, based on the data in the knowledge graph, the relationship between the teacher's teaching behavior and teaching effect is analyzed; the teaching behavior analysis model is constructed based on the graph convolutional neural network; the teaching behavior analysis model uses the knowledge graph as input and outputs the evaluation results of the teaching quality.

2. The system according to claim 1, characterized in that The teaching behavior data includes: classroom interaction data, teaching strategy data and teaching material data.

3. The system according to claim 1, characterized in that The teaching effect data includes: student feedback data, parent feedback data and student performance data.

4. The system according to claim 1, characterized in that The data collection module is also used to clean, standardize and transform the collected data after it is collected.

5. The system according to claim 1, characterized in that Construct a multi-dimensional knowledge graph, including: Entity recognition and classification: Identify entities and their attributes from the collected data, including teachers, teaching behaviors, and teaching effects; Relationship definition and extraction: determine the relationships between entities and extract these relationships from the data; the relationships between entities include: usage relationships, connecting teachers and specific teaching behaviors; influence relationships, connecting teaching behaviors and teaching effects; Knowledge graph construction: Using graph database, the identified entities and relationships are constructed into a knowledge graph; Entity linking and fusion: Entities from different data sources but representing the same real object are linked and fused.

6. The system according to claim 5, characterized in that Building a multi-dimensional knowledge graph also includes: Knowledge graph update and maintenance: Regularly update the knowledge graph to reflect changes in teachers' teaching behaviors as well as new teaching methods and teaching effectiveness data.

7. The system according to claim 1, characterized in that Using the teaching behavior analysis model, the relationship between teachers’ teaching behavior and teaching effect is analyzed based on the data in the knowledge graph, including: Data preprocessing: converting the data in the knowledge graph into a format suitable for graph convolutional neural network processing, including constructing adjacency matrix and feature matrix; Model training: training teaching behavior analysis model based on data preprocessed; Teaching quality evaluation: The features learned by the trained model are used for classification or regression tasks to predict the teaching effect, and the teaching quality is evaluated based on the predicted teaching effect.

8. The system according to claim 7, characterized in that The teaching quality assessment also includes: formulating personalized improvement suggestions based on each teacher’s teaching behavior characteristics and corresponding analysis results.

9. The system according to claim 1, characterized in that The system also includes: a report generation module: automatically generates a comprehensive report including teacher teaching behavior analysis, teaching quality evaluation and improvement suggestions.

10. The system according to claim 9, characterized in that The system also includes: a visualization display module: visually outputting the report.

Images