Student group tutoring matching method based on psychological portraits and data driving

Abstract

The invention discloses a student group tutoring matching method based on psychological portraits and data driving. According to the method, a multi-modal data fusion and dynamic psychological portrait modeling mechanism is introduced, and a self-adaptive multi-objective optimization algorithm is combined, so that a novel student group tutoring matching method is provided, and a plurality of key problems in the prior art are solved. Existing student tutoring methods generally depend on static psychological assessment data and manual grouping, are difficult to effectively reflect dynamic changes of emotions and behaviors of students, and cannot automatically adapt to changing psychological states of the students. Through the method provided by the invention, the emotion, behavior, learning progress and other characteristics of the students can be automatically extracted based on the dynamic psychological portraits, and weighting is carried out according to the relative importance of different targets, so that more accurate group matching is provided.

Description

Technical Field

[0001] This invention relates to the fields of educational informatization and psychological data analysis technology, and in particular to a student group counseling matching method based on psychological profiling and data-driven approaches. Background Technology

[0002] With the continuous development of educational informatization and mental health management, student counseling is gradually shifting from a single manual assessment model to a data-driven intelligent analysis model. Traditional group counseling matching methods primarily rely on counselors' subjective judgments of students or fixed psychological assessment questionnaire results for grouping. While this approach is feasible in early childhood education environments, it has a weak ability to respond to dynamic changes in students' emotional states, behavioral fluctuations, and real-time feedback on their learning status, making it difficult to adapt to the complexity of contemporary students' psychological and behavioral characteristics. In particular, existing technologies still have significant shortcomings in multimodal data fusion and dynamic feature recognition.

[0003] Currently, most psychological counseling matching systems are based on static psychological assessment data to build matching models. Some studies use fixed-period psychological questionnaires or personality scales to classify students' psychological characteristics and then group them according to the scale results. However, such methods can only capture static slices of students' psychological states and cannot reflect the time-series changes in their emotions and behaviors. Students' psychological fluctuations under academic pressure, social relationships, or environmental changes often exhibit complex non-linear characteristics. Group matching mechanisms that rely solely on fixed data cannot accurately respond to such dynamic changes, resulting in poor targeting and stability of counseling groups.

[0004] Traditional matching algorithms are mostly based on simple similarity calculations or clustering analysis models. For example, K-means clustering, hierarchical clustering, or matching algorithms based on Euclidean distance are widely used in student grouping tasks. However, these algorithms generally assume that the feature dimensions are independent of each other and cannot handle the coupling relationship between emotional, behavioral, and learning features. Students' psychological states are often influenced by multiple factors simultaneously, such as the interaction effect between emotional fluctuations and behavioral patterns, and the dynamic feedback between learning engagement and emotional fluctuations. The interactive effects of these features cannot be effectively represented by traditional clustering algorithms, resulting in matching results that deviate from actual needs.

[0005] To address this issue, some studies have introduced machine learning methods, such as decision trees, support vector machines, and neural network models, to improve the accuracy of psychological profiling. However, these methods still largely remain at the "static learning" stage, meaning that once the model is trained, its parameters are fixed and cannot adaptively adjust to real-time changes in students' psychological states. When students' behavioral patterns shift, the model's output may quickly become invalid, leading to a decrease in the timeliness and accuracy of group counseling matching. Furthermore, existing technologies for constructing "psychological profiles" often remain at the feature extraction level, lacking dynamic representation mechanisms based on emotional and behavioral fluctuations. The inherent correlation between emotional and behavioral data often exhibits non-stationary characteristics; if fluctuation patterns in time series are not considered, the model may easily overlook potential psychological trends.

[0006] Existing technologies also have limitations in matching optimization strategies. Most psychological group matching systems are based on single-objective optimization, such as minimizing personality differences within the group or maximizing emotional similarity, while ignoring the multi-dimensional characteristics of counseling goals. In fact, the effectiveness of group counseling depends not only on the emotional compatibility among students but also on the combined influence of behavioral synergy and similar learning progress. Matching algorithms lacking a multi-objective balancing mechanism often optimize one objective at the expense of other dimensions, leading to an imbalance in the group structure during long-term counseling.

[0007] Current matching algorithms typically rely on human experience or fixed parameters for weight settings, making personalized adjustments difficult. The relative importance of emotions, behaviors, and learning progress changes dynamically across different coaching stages. For example, emotional stability may be more critical in the early stages of group coaching, while learning progress and behavioral consistency may have a more significant impact on group outcomes in the later stages. Existing fixed-weight mechanisms cannot adjust the weight allocation of optimization objectives over time, resulting in a lack of flexibility and adaptability in matching results. Summary of the Invention

[0008] One objective of this invention is to propose a student group counseling matching method based on psychological profiling and data-driven approaches. This method introduces multimodal data fusion and dynamic psychological profiling modeling mechanisms, combined with an adaptive multi-objective optimization algorithm, to propose a group matching method capable of real-time perception of students' psychological and behavioral fluctuations, thereby overcoming several key problems in existing counseling matching technologies. Through this method, the system can dynamically adjust group combinations based on students' emotional stability, behavioral consistency, and changes in learning progress during the counseling process, maintaining a coordinated balance among group members in terms of emotional, behavioral, and learning characteristics, thus significantly improving counseling effectiveness and individual fit.

[0009] The student group counseling matching method based on psychological profiling and data-driven methods according to embodiments of the present invention includes the following steps: Step 1: Collect students' multimodal data, including students' emotional data, behavioral data, and psychological measurement data. Standardize the multimodal data to generate standardized multimodal data. Step 2: Based on the standardized multimodal data, construct a fluctuation model, and use time series analysis and wavelet transform to model the students' emotional fluctuations and behavioral changes, generating emotional fluctuation data and behavioral fluctuation data of students during the tutoring process; Step 3: Based on the emotional fluctuation data and behavioral fluctuation data, establish a self-evolutionary optimization algorithm to dynamically model the students' emotional and behavioral fluctuations. The self-evolutionary optimization algorithm includes recording the students' emotional and behavioral fluctuations according to historical data, adjusting the fluctuation model using the optimization algorithm, and obtaining updated emotional and behavioral characteristics. Step 4: Based on the updated emotional and behavioral characteristics, generate a dynamic psychological profile of the student. The dynamic psychological profile reflects the student's personalized characteristics during the counseling process by integrating the student's emotional fluctuations, behavioral fluctuations, and psychological measurement data. Step 5: Based on the dynamic psychological profile of the students, construct a multi-objective optimization algorithm to generate group matching relationships, where the objectives include emotional stability, behavioral consistency, and learning progress; Step Six: During the counseling process, monitor students' emotional fluctuations and behavioral changes in real time. Based on real-time data, dynamically update students' dynamic psychological profiles through a feedback mechanism, and adjust the matching relationships of group members according to the updated profiles.

[0010] Optionally, step one specifically includes: Collect students' multimodal data, including students' emotional data, behavioral data, and psychometric data; The multimodal data is standardized, including standardizing different data types to generate standardized multimodal data.

[0011] Optionally, step two specifically includes: Based on the standardized multimodal data, time series analysis methods are used to model students' emotional and behavioral data; Wavelet transform technology is applied to decompose the data at multiple scales, extracting local features of students' emotional and behavioral fluctuations. Combined with time series analysis and wavelet transform results, a fluctuation model is constructed. Generate data on students' emotional and behavioral fluctuations during the tutoring process.

[0012] Optionally, step three specifically includes: Based on the emotional fluctuation data and behavioral fluctuation data, a self-evolutionary optimization algorithm is established. The self-evolutionary optimization algorithm dynamically adjusts the weights and parameters of the fluctuation model by recording the changes in emotional and behavioral fluctuations in historical data, and generates updated emotional and behavioral features.

[0013] Optionally, the self-evolutionary optimization algorithm is specifically as follows: The self-evolutionary optimization algorithm dynamically adjusts the weights and parameters of the fluctuation model by recording historical data of students' emotional and behavioral fluctuations to obtain updated emotional and behavioral characteristics. Based on students' historical emotional and behavioral fluctuation data, the parameters of the fluctuation model are optimized and adjusted. The algorithm calculates the current time. The error between the emotional and behavioral fluctuations and the predicted values ​​is used to update the model using an error feedback mechanism: ; in, These are the old parameters for the current model. For learning rate, Errors due to emotional fluctuations For the error of behavioral fluctuations, These are the updated model parameters.

[0014] Optionally, step four specifically includes: Based on the updated emotional and behavioral characteristics, a dynamic psychological profile of the student is generated. The dynamic psychological profile integrates students' emotional fluctuations, behavioral fluctuations, and psychological measurement data to reflect the students' individual characteristics during the counseling process. Based on the updated emotional and behavioral characteristics, a dynamic psychological profile of the student is generated.

[0015] Optionally, generating a dynamic psychological profile of a student specifically includes: Based on the updated emotional and behavioral characteristics, a dynamic psychological profile of the student is generated. This dynamic psychological profile reflects the student's individual characteristics during the counseling process by integrating the student's emotional fluctuations, behavioral fluctuations, and psychological measurement data. ; in, For students at all times Dynamic psychological profile, For the updated sentiment feature data, For the updated behavioral fluctuation data, For students at all times Psychological measurement data, For the weighting coefficients, satisfying ; By weighting and integrating these data, a personalized psychological profile of the student is generated.

[0016] Optionally, step five specifically includes: Based on the dynamic psychological profile of the students, a multi-objective optimization function is constructed using the students' emotional fluctuation data, behavioral fluctuation data, and learning progress data. The objectives of the optimization function include emotional stability, behavioral consistency, and learning progress. By calculating the weights of student emotional fluctuation data, behavioral fluctuation data, and learning progress data in each objective function and weighting them according to the relative importance of each objective, the particle swarm optimization algorithm is used to optimize the group matching relationship. The matching relationship between group members is iteratively adjusted to minimize the error of the objective function and generate the group matching relationship.

[0017] Optionally, the multi-objective optimization algorithm specifically includes: Using students' emotional fluctuation data, behavioral fluctuation data, and learning progress data, a multi-objective optimization function is constructed. The objectives of the optimization function include emotional stability, behavioral consistency, and learning progress. ; in, To optimize the total value of the function, and The scores were for emotional stability, behavioral consistency, and learning progress, respectively. The weighting coefficients for emotional stability, behavioral consistency, and learning progress goals, and satisfying the following conditions: ; The weighting coefficients are dynamically adjusted based on students' real-time tutoring data. ; in, For the first The goal is at any time The weight, The magnitude of change of the objective in the current iteration. For adaptive adjustment coefficients, For the first The goal is at any time weights, This represents the magnitude of change of the objective in the current iteration.

[0018] Optionally, step six specifically includes: During the tutoring process, students' emotional fluctuations and behavioral changes are monitored in real time. This real-time monitoring involves collecting students' emotional and behavioral data and updating this data regularly during the tutoring process. The dynamic psychological profile of students is dynamically updated through a feedback mechanism, which includes real-time analysis of students' emotional fluctuations and behavioral changes in order to adjust and improve the dynamic psychological profile of students. Adjust the matching relationships of group members based on the updated dynamic psychological profiles.

[0019] The beneficial effects of this invention are: This invention proposes a novel student group counseling matching method by introducing a multimodal data fusion and dynamic psychological profile modeling mechanism, combined with an adaptive multi-objective optimization algorithm, overcoming several key problems in existing technologies. Existing student group counseling matching methods generally rely on static psychological assessment data and manual grouping, making it difficult to effectively reflect the dynamic changes in students' emotions and behaviors, and failing to automatically adjust the group member structure during the counseling process. The method of this invention can automatically generate dynamic psychological profiles based on students' emotional fluctuation data, behavioral fluctuation data, and learning progress data, and then perform intelligent group matching based on these profiles, thereby providing more accurate matching results.

[0020] This invention utilizes a dynamic psychological profiling method to capture students' emotional fluctuations, behavioral changes, and learning progress in real time. This allows each student's psychological profile to be continuously updated during the tutoring process, reflecting their individualized psychological state in real time. This method is better able to handle the non-linear changes in students' emotions and behaviors than traditional static psychological assessment methods, thereby enhancing the adaptability and accuracy of tutoring matching. Furthermore, by incorporating an adaptive multi-objective optimization algorithm, it can balance multiple dimensions such as emotional stability, behavioral consistency, and learning progress, avoiding the limitations of single-objective optimization and significantly improving the effectiveness of tutoring group matching.

[0021] The multi-objective optimization algorithm of this invention optimizes group matching relationships, dynamically adjusting the matching relationships between group members based on students' real-time data, thereby ensuring that the tutoring group maintains its optimal state during long-term tutoring. Compared to traditional matching algorithms based on static data, this method better adapts to changes in students' states, enhancing the personalization and targeting of tutoring.

[0022] In practical applications, this invention addresses the problems of inaccurate group matching and lack of dynamic adjustment in existing methods. Existing tutoring matching methods typically rely on manually set rules and fixed data inputs, while this invention can automatically learn students' dynamic characteristics and make real-time decisions, making the tutoring process more intelligent and efficient. This method not only improves the matching accuracy of group tutoring but also optimizes the student's tutoring experience through a dynamic feedback mechanism.

[0023] This invention not only addresses the issues of stability and personalization deficiencies in existing group matching methods, but also enhances the intelligence and adaptability of student group tutoring matching by introducing dynamic psychological profiling and multi-objective optimization algorithms. It has broad application prospects in the field of student group tutoring matching, particularly in intelligent tutoring within educational environments and personalized learning support for students, providing more precise support for the formulation of tutoring strategies. Attached Figure Description

[0024] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0025] Figure 1 This is a flowchart of the student group counseling matching method based on psychological profiling and data-driven approaches proposed in this invention. Figure 2 This diagram illustrates the generation of matching relationships in the student group counseling matching method based on psychological profiling and data-driven approaches proposed in this invention. Detailed Implementation

[0026] Combination Figures 1-2 The present invention will be described in further detail below. These accompanying drawings are simplified schematic diagrams, illustrating only the basic structure of the invention and showing the main components relevant to the invention. Figure 1 and Figure 2 The student group counseling matching method based on psychological profiling and data-driven methods of the present invention includes the following steps:

[0027] Step 1: Collect students' multimodal data, including students' emotional data, behavioral data, and psychological measurement data. Standardize the multimodal data to generate standardized multimodal data. Step 2: Based on the standardized multimodal data, construct a fluctuation model, and use time series analysis and wavelet transform to model the students' emotional fluctuations and behavioral changes, generating emotional fluctuation data and behavioral fluctuation data of students during the tutoring process; Step 3: Based on the emotional fluctuation data and behavioral fluctuation data, establish a self-evolutionary optimization algorithm to dynamically model the students' emotional and behavioral fluctuations. The self-evolutionary optimization algorithm includes recording the students' emotional and behavioral fluctuations according to historical data, adjusting the fluctuation model using the optimization algorithm, and obtaining updated emotional and behavioral characteristics. Step 4: Based on the updated emotional and behavioral characteristics, generate a dynamic psychological profile of the student. The dynamic psychological profile reflects the student's personalized characteristics during the counseling process by integrating the student's emotional fluctuations, behavioral fluctuations, and psychological measurement data. Step 5: Based on the dynamic psychological profile of the students, construct a multi-objective optimization algorithm to generate group matching relationships, where the objectives include emotional stability, behavioral consistency, and learning progress; Step Six: During the counseling process, monitor students' emotional fluctuations and behavioral changes in real time. Based on real-time data, dynamically update students' dynamic psychological profiles through a feedback mechanism, and adjust the matching relationships of group members according to the updated profiles.

[0028] In this embodiment, step one specifically includes: Collect students' multimodal data, including students' emotional data, behavioral data, and psychometric data; The multimodal data is standardized, including standardizing different data types to generate standardized multimodal data.

[0029] In this embodiment, step two specifically includes: Based on the standardized multimodal data, time series analysis methods are used to model students' emotional and behavioral data; Wavelet transform technology is applied to decompose the data at multiple scales, extracting local features of students' emotional and behavioral fluctuations. Combined with time series analysis and wavelet transform results, a fluctuation model is constructed. Generate data on students' emotional and behavioral fluctuations during the tutoring process.

[0030] In this embodiment, step three specifically includes: Based on the emotional fluctuation data and behavioral fluctuation data, a self-evolutionary optimization algorithm is established. The self-evolutionary optimization algorithm dynamically adjusts the weights and parameters of the fluctuation model by recording the changes in emotional and behavioral fluctuations in historical data, and generates updated emotional and behavioral features.

[0031] In this embodiment, the self-evolutionary optimization algorithm is specifically as follows: The self-evolutionary optimization algorithm dynamically adjusts the weights and parameters of the fluctuation model by recording historical data of students' emotional and behavioral fluctuations to obtain updated emotional and behavioral characteristics. Based on students' historical emotional and behavioral fluctuation data, the parameters of the fluctuation model are optimized and adjusted. The algorithm calculates the current time. The error between the emotional and behavioral fluctuations and the predicted values ​​is used to update the model using an error feedback mechanism: ; in, These are the old parameters for the current model. For learning rate, Errors due to emotional fluctuations For the error of behavioral fluctuations, These are the updated model parameters.

[0032] In this embodiment, step four specifically includes: Based on the updated emotional and behavioral characteristics, a dynamic psychological profile of the student is generated. The dynamic psychological profile integrates students' emotional fluctuations, behavioral fluctuations, and psychological measurement data to reflect the students' individual characteristics during the counseling process. Based on the updated emotional and behavioral characteristics, a dynamic psychological profile of the student is generated.

[0033] In this embodiment, generating the student's dynamic psychological profile specifically includes: Based on the updated emotional and behavioral characteristics, a dynamic psychological profile of the student is generated. This dynamic psychological profile reflects the student's individual characteristics during the counseling process by integrating the student's emotional fluctuations, behavioral fluctuations, and psychological measurement data. ; in, For students at all times Dynamic psychological profile, For the updated sentiment feature data, For the updated behavioral fluctuation data, For students at all times Psychological measurement data, For the weighting coefficients, satisfying ; By weighting and integrating these data, a personalized psychological profile of the student is generated.

[0034] In this embodiment, step five specifically includes: Based on the dynamic psychological profile of the students, a multi-objective optimization function is constructed using the students' emotional fluctuation data, behavioral fluctuation data, and learning progress data. The objectives of the optimization function include emotional stability, behavioral consistency, and learning progress. By calculating the weights of student emotional fluctuation data, behavioral fluctuation data, and learning progress data in each objective function and weighting them according to the relative importance of each objective, the particle swarm optimization algorithm is used to optimize the group matching relationship. The matching relationship between group members is iteratively adjusted to minimize the error of the objective function and generate the group matching relationship.

[0035] In this embodiment, the multi-objective optimization algorithm specifically includes: Using students' emotional fluctuation data, behavioral fluctuation data, and learning progress data, a multi-objective optimization function is constructed. The objectives of the optimization function include emotional stability, behavioral consistency, and learning progress. ; in, To optimize the total value of the function, and The scores were for emotional stability, behavioral consistency, and learning progress, respectively. The weighting coefficients for emotional stability, behavioral consistency, and learning progress goals, and satisfying the following conditions: ; The weighting coefficients are dynamically adjusted based on students' real-time tutoring data. ; in, For the first The goal is at any time The weight, The magnitude of change of the objective in the current iteration. For adaptive adjustment coefficients, For the first The goal is at a certain moment weights, This represents the magnitude of change of the objective in the current iteration.

[0036] In this embodiment, step six specifically includes: During the tutoring process, students' emotional fluctuations and behavioral changes are monitored in real time. This real-time monitoring involves collecting students' emotional and behavioral data and updating this data regularly during the tutoring process. The dynamic psychological profile of students is dynamically updated through a feedback mechanism, which includes real-time analysis of students' emotional fluctuations and behavioral changes in order to adjust and improve the dynamic psychological profile of students. Adjust the matching relationships of group members based on the updated dynamic psychological profiles.

[0037] Example 1: This embodiment applies the student group counseling matching method based on psychological profiling and data-driven approaches of the present invention to the psychological health counseling process of a high school student. In this school, there are approximately 1500 students, of whom about 200 frequently face problems such as significant emotional fluctuations, inconsistent behavior, and large differences in learning progress. Traditional counseling methods mainly rely on manual grouping, with counselors grouping students based on psychological assessment results and behavioral performance. However, due to the differences in students' emotional, behavioral, and learning states, the counseling effect of traditional methods has not reached the ideal level, resulting in poor counseling experience and learning outcomes for students. Therefore, the school decided to try using the intelligent student group matching system of the present invention to improve the counseling effect through dynamic psychological profiling generation and multi-objective optimization algorithms.

[0038] In this embodiment, the school's counseling center first collected data on students' emotional fluctuations, behavioral fluctuations, and learning progress. Emotional fluctuation data was obtained through monthly mental health questionnaires, behavioral fluctuation data was collected through counselor observation reports and records of students' participation in group activities, and learning progress data was obtained through students' report cards and completion of learning tasks. Standardization and normalization of this data ensured its uniformity and comparability. Subsequently, the system used this data to generate a dynamic psychological profile for each student.

[0039] Each student's dynamic psychological profile not only reflects their current emotional, behavioral, and learning status but also reflects changes in the student in real time during the tutoring process. As tutoring progresses, the student's psychological profile is continuously updated, capturing dynamic changes in the student's emotions, behavior, and learning progress in real time. Based on this, the method of this invention applies a multi-objective optimization algorithm for student group matching. The system balances multiple objectives by comprehensively considering students' emotional stability, behavioral consistency, and learning progress, and optimizes them using a particle swarm optimization algorithm, thereby matching each student with the most suitable tutoring group.

[0040] In practice, the school's counseling center selected 50 students as experimental subjects, representing a diverse group with varying emotional states, behaviors, and learning progress. Using dynamically generated psychological profiles, the system matched these 50 students with the most suitable counseling groups. In traditional counseling methods, manually assigned groups by counselors often exhibit significant emotional fluctuations, behavioral inconsistencies, and differences in learning progress, leading to unsatisfactory counseling results. However, the method of this invention allows the system to dynamically adjust the members of each counseling group based on each student's psychological profile and real-time data, ensuring a high degree of matching in terms of emotions, behaviors, and learning progress, thereby improving the accuracy and relevance of the counseling.

[0041] During implementation, we compared the group matching effects of traditional manual grouping and the method of this invention, and obtained the following results. Under the traditional grouping method, students' emotions fluctuated significantly, with approximately 70% of students remaining emotionally unstable after tutoring, resulting in poor tutoring effectiveness. In contrast, in the tutoring group using the method of this invention, only 10% of students experienced significant emotional fluctuations, leading to a significant improvement in tutoring effectiveness. Regarding behavioral consistency, approximately 60% of students in the traditional grouping exhibited inconsistent behavior during tutoring, causing difficulties in interaction and collaboration among group members and affecting tutoring effectiveness. In the tutoring group using this invention, approximately 85% of students exhibited more consistent behavioral patterns, making group collaboration smoother. Regarding learning progress, the tutoring group members in the traditional grouping method showed significant differences in learning progress, causing some students to struggle to keep up and affecting overall learning effectiveness. In contrast, the tutoring group members matched using the method of this invention showed more consistent learning progress, resulting in a significant improvement in tutoring effectiveness.

[0042] To further verify the effectiveness of the method of this invention, we conducted a comparative analysis of the academic performance of 50 students. Before tutoring, the average academic score of the 50 students was 65 points. After three months of tutoring, the academic score improved to 85 points, an average improvement of 30%. Regarding emotional stability, after tutoring, approximately 80% of the students effectively controlled their emotional fluctuations, and their emotional stability scores generally improved. Regarding behavioral consistency, approximately 90% of the students showed significant improvement in behavioral consistency, with enhanced interactivity and cooperation during the tutoring process. Regarding learning progress, after tutoring, approximately 70% of the students completed the predetermined learning tasks, and the overall learning effect significantly improved.

[0043] Table 1

[0044] The data from this embodiment demonstrates that the tutoring matching method of the present invention effectively solves the problems of large emotional fluctuations, inconsistent behavior, and significant differences in learning progress in traditional tutoring methods. After applying the method of the present invention, students' emotional stability, behavioral consistency, and learning progress are significantly improved, resulting in a substantial improvement in tutoring effectiveness. This method not only enhances the effectiveness of group tutoring but also provides a feasible solution for future intelligent and personalized tutoring, showing broad application prospects. With the increase in data volume and the expansion of application scenarios, it is expected that the method of the present invention will be applied in more schools and educational institutions, especially in the fields of intelligent tutoring and personalized student learning support, providing more precise support for the education system.

Claims

1. A student group counseling matching method based on psychological profiling and data-driven approaches, characterized in that: Includes the following steps: Step 1: Collect students' multimodal data, including students' emotional data, behavioral data, psychological measurement data, and learning progress data. Standardize the multimodal data to generate standardized multimodal data. Step 2: Based on the standardized multimodal data, construct a fluctuation model, and use time series analysis and wavelet transform to model the students' emotional fluctuations and behavioral changes, generating emotional fluctuation data and behavioral fluctuation data of students during the tutoring process; Step 3: Based on the emotional fluctuation data and behavioral fluctuation data, establish a self-evolutionary optimization algorithm to dynamically model the students' emotional and behavioral fluctuations. The self-evolutionary optimization algorithm includes recording the students' emotional and behavioral fluctuations according to historical data, adjusting the fluctuation model using the optimization algorithm, and obtaining updated emotional and behavioral characteristics. Step 4: Based on the updated emotional and behavioral characteristics, generate a dynamic psychological profile of the student. The dynamic psychological profile reflects the student's personalized characteristics during the counseling process by integrating the student's emotional fluctuations, behavioral fluctuations, and psychological measurement data. Step 5: Based on the dynamic psychological profile of the students, construct a multi-objective optimization algorithm to generate group matching relationships, where the objectives include emotional stability, behavioral consistency, and learning progress; Step Six: During the counseling process, monitor students' emotional fluctuations and behavioral changes in real time. Based on real-time data, dynamically update students' dynamic psychological profiles through a feedback mechanism, and adjust the matching relationships of group members according to the updated profiles.

2. The student group counseling matching method based on psychological profiling and data-driven approach according to claim 1, characterized in that, Step one specifically includes: Collect students' multimodal data, including students' emotional data, behavioral data, and psychometric data; The multimodal data is standardized, including standardizing different data types to generate standardized multimodal data.

3. The student group counseling matching method based on psychological profiling and data-driven approach according to claim 1, characterized in that, Step two specifically includes: Based on the standardized multimodal data, time series analysis methods are used to model students' emotional and behavioral data; Wavelet transform technology is applied to decompose the data at multiple scales, extracting local features of students' emotional and behavioral fluctuations. Combined with time series analysis and wavelet transform results, a fluctuation model is constructed. Generate data on students' emotional and behavioral fluctuations during the tutoring process.

4. The student group counseling matching method based on psychological profiling and data-driven approach according to claim 1, characterized in that, Step three specifically includes: Based on the emotional fluctuation data and behavioral fluctuation data, a self-evolutionary optimization algorithm is established. The self-evolutionary optimization algorithm dynamically adjusts the weights and parameters of the fluctuation model by recording the changes in emotional and behavioral fluctuations in historical data, and generates updated emotional and behavioral features.

5. The student group counseling matching method based on psychological profiling and data-driven approach according to claim 4, characterized in that, The self-evolutionary optimization algorithm is specifically as follows: The self-evolutionary optimization algorithm dynamically adjusts the weights and parameters of the fluctuation model by recording historical data of students' emotional and behavioral fluctuations to obtain updated emotional and behavioral characteristics. Based on students' historical emotional and behavioral fluctuation data, the parameters of the fluctuation model are optimized and adjusted. The algorithm calculates the current time. The error between the emotional and behavioral fluctuations and the predicted values ​​is used to update the model using an error feedback mechanism: ； in, These are the old parameters for the current model. For learning rate, Errors due to emotional fluctuations For the error of behavioral fluctuations, These are the updated model parameters.

6. The student group counseling matching method based on psychological profiling and data-driven approach according to claim 1, step four specifically includes: Based on the updated emotional and behavioral characteristics, a dynamic psychological profile of the student is generated. The dynamic psychological profile reflects the individual characteristics of students during the counseling process by integrating their emotional fluctuations, behavioral fluctuations, and psychological measurement data. Based on the updated emotional and behavioral characteristics, a dynamic psychological profile of the student is generated.

7. The student group counseling matching method based on psychological profiling and data-driven approach according to claim 6, characterized in that, The generation of dynamic psychological profiles of students specifically includes: Based on the updated emotional and behavioral characteristics, a dynamic psychological profile of the student is generated. This dynamic psychological profile reflects the student's individual characteristics during the counseling process by integrating the student's emotional fluctuations, behavioral fluctuations, and psychological measurement data. ； in, For students at all times Dynamic psychological profile, For the updated sentiment feature data, For the updated behavioral fluctuation data, For students at all times Psychological measurement data, For the weighting coefficients, satisfying ; By weighting and integrating these data, a personalized psychological profile of the student is generated.

8. The student group counseling matching method based on psychological profiling and data-driven approach according to claim 1, characterized in that, Step five specifically includes: Based on the dynamic psychological profile of the students, a multi-objective optimization function is constructed using the students' emotional fluctuation data, behavioral fluctuation data, and learning progress data. The objectives of the optimization function include emotional stability, behavioral consistency, and learning progress. By calculating the weights of student emotional fluctuation data, behavioral fluctuation data, and learning progress data in each objective function and weighting them according to the relative importance of each objective, the particle swarm optimization algorithm is used to optimize the group matching relationship. The matching relationship between group members is iteratively adjusted to minimize the error of the objective function and generate the group matching relationship.

9. The student group counseling matching method based on psychological profiling and data-driven approach according to claim 8, characterized in that, The multi-objective optimization algorithm specifically includes: Using students' emotional fluctuation data, behavioral fluctuation data, and learning progress data, a multi-objective optimization function is constructed. The objectives of the optimization function include emotional stability, behavioral consistency, and learning progress. ; in, To optimize the total value of the function, and The scores were for emotional stability, behavioral consistency, and learning progress, respectively. The weighting coefficients for emotional stability, behavioral consistency, and learning progress goals, and satisfying the following conditions: ; The weighting coefficients are dynamically adjusted based on students' real-time tutoring data. ; in, For the first The goal is at a certain moment The weight, The magnitude of change of the objective in the current iteration. For adaptive adjustment coefficients, For the first The goal is at a certain moment weights, This represents the magnitude of change of the objective in the current iteration.

10. The student group counseling matching method based on psychological profiling and data-driven approach according to claim 1, characterized in that, Step six specifically includes: During the tutoring process, students' emotional fluctuations and behavioral changes are monitored in real time. This real-time monitoring involves collecting students' emotional and behavioral data and updating this data regularly during the tutoring process. The dynamic psychological profile of students is dynamically updated through a feedback mechanism, which includes real-time analysis of students' emotional fluctuations and behavioral changes in order to adjust and improve the dynamic psychological profile of students. Adjust the matching relationships of group members based on the updated dynamic psychological profiles.

Images