Learning intervention method and system in mixed learning environment

Abstract

The invention belongs to but not limited to the technical field of mixed learning, and particularly relates to a learning intervention method and system in a mixed learning environment, and the method comprises the steps: S1, constructing a portrait model; s2, establishing an intervention framework based on a learner portrait in a mixed learning environment; and S3, implementing learning intervention. The learning portrait model oriented to mixed learning comprises dimensions such as basic information of students, learning preparation, learning styles, learning behaviors, learning results and thinking ability, classroom and online learning data of the learners are collected and processed, overall, group and individual portraits of the learners are presented in a visual form, a teacher reprocesses the portraits, and the learning portrait model is used for learning the students. A learning report is generated and fed back to students, learner problems are diagnosed through portraits, a precise and personalized learning intervention framework and strategy are formed, and overall, group and individual learning intervention strategies are implemented in combination with portraits in the whole process of mixed learning.

Description

Technical Field

[0001] The present invention belongs to, but is not limited to, the technical field of blended learning, and particularly relates to a learning intervention method and system in a blended learning environment. Background Art

[0002] Existing learner portraits have differences in data format, grammar, and semantics, lacking a unified standard, making it difficult to ensure data exchange and interoperability across platforms and systems, which restricts data integration and sharing.

[0003] Current systems have technical bottlenecks in the collection, cleaning, and real-time analysis of learning behavior data, making it difficult to comprehensively and dynamically reflect the state of learners, affecting the accuracy of personalized portraits and the effect of real-time intervention. Although the intervention model based on learner portraits has initially verified the effectiveness of the RTI mode, the precise strategy matching and dynamic adjustment mechanisms for different learner characteristics are still immature and urgently need to be optimized and improved.

[0004] The integration of online and offline data still faces technical obstacles, and the information transmission and feedback mechanisms among teachers, students, and platforms lack efficient coordination, restricting the implementation effect of personalized teaching in the blended learning mode.

[0005] Although technologies such as machine learning and data mining have been applied to learner behavior analysis, how to deeply embed intelligent algorithms into the teaching decision-making and personalized support processes to achieve automated and precise intervention still needs further research.

[0006] The collection and storage of large-scale learner data pose severe challenges to privacy protection and information security. Existing technologies need to be strengthened in aspects such as data encryption, permission management, and secure transmission to ensure user information security. Summary of the Invention

[0007] In view of the problems existing in the prior art, the present invention provides a learning intervention method and system in a blended learning environment, using learner portraits to provide technical support in the learning intervention in the blended learning mode, in order to give full play to the value of learner portraits and achieve better blended learning effects.

[0008] The present invention is implemented as follows. A learning intervention method in a blended learning environment includes:

[0009] S1, constructing a portrait model;

[0010] S2, building an intervention framework based on learner portraits in a blended learning environment;

[0011] S3, implementing learning intervention.

[0012] Further, the process of constructing the portrait model specifically includes:

[0013] (1) Requirement analysis and goal determination: Requirement analysis can help set goals, review data, determine solutions, and prepare for project evaluation. The needs for learner portraits in a blended learning environment can be divided into three aspects: ① Learner analysis. By analyzing learners through portraits, their personality characteristics can be understood, and learning change trends can be grasped to meet the learning development needs. ② Learning services. Learning services based on portraits include learning resource recommendation services, personalized learning path customization, learning partner matching services, learning behavior supervision and early warning, etc. ③ Learning evaluation. Learner portraits can be used for personal evaluation of learners, and teachers' evaluation and analysis of the whole class or group. The goal of portrait construction is to objectively and clearly evaluate learners through portraits, and then provide precise learning intervention services for learners' learning.

[0014] (2) Data collection and processing: Data is the basis for constructing portraits and can determine the effectiveness of portraits. The learning data of the present invention comes from the online learning platform Xuexitong, the learning management system, students' homework and test papers, and learners' questionnaires. Therefore, the types of collected data are rich and diverse, and the data format needs to be unified through operations such as data cleaning, data standardization, and data conversion so that researchers can better store and utilize the data. The learning data in the research can be divided into four categories:

[0015] Information survey data includes learners' identity information, learning styles, and learning motivations.

[0016] Learning behavior data is a record of learners' behavior activities and participation in classroom and online environments.

[0017] Learning result data is a description of learners' phased achievements, test scores, etc.

[0018] Thinking ability data reflects students' higher-level cognitive levels and includes multiple elements such as problem-solving, thinking innovation, and critical decision-making. These data are the objects of learner portrait analysis.

[0019] (3) Portrait modeling and analysis: Portrait modeling and analysis is to mine information from learners' data and is the core and key in the portrait construction process. The dimension selection of the portrait model is mainly determined by the construction goal, and learners' data is labeled according to the dimensions.

[0020] (4) Portrait output and visualization: To give full play to the educational function of learner portraits, it is necessary to present learner characteristics in a visual way. In the present invention, individual learner portraits, group learner portraits, and overall learner portraits will be formed.

[0021] (5) Portrait application and evaluation: The construction of learner portraits ultimately serves learning intervention practices. Learners’ behaviors and outcomes will change over time, and portraits also need to be updated to achieve the original construction goals.

[0022] Furthermore, the intervention framework based on learner profiles in a blended learning environment specifically includes:

[0023] Provide the most appropriate learning intervention for each student based on the learner profile. The intervention framework is as follows:

[0024] Conduct learning situation analysis and diagnose learning problems based on learner portraits; implement interventions in the whole process of learning preparation, autonomous learning, consolidation and internalization, application transfer, and evaluation feedback in blended learning based on the principles of student subjectivity, accuracy, embeddedness, and effectiveness. Design overall intervention strategies based on learner portraits, including learning supervision and early warning, technical and resource guidance, and mind mapping; group intervention strategies include goal positioning and value guidance, teacher expectation transmission, task design and guidance, activity strategies, resource recommendations, and evaluation strategies; individual intervention strategies include individual reminder strategies, portrait feedback strategies, and self-reflection and adjustment strategies. Evaluate the effectiveness of intervention practices based on changes in self-portraits, differences between individual and group portraits, changes before and after the overall portrait, and comparison of the overall portraits of each class; distribute satisfaction questionnaires to students, and interview teachers to understand the emotional experience of teachers and students on intervention practices based on learner portraits.

[0025] Furthermore, the implementation of learning intervention specifically includes:

[0026] The implementation method of learning intervention based on learner portrait in a hybrid learning environment is described here in combination with the experiment of the present invention as follows:

[0027] (1) Understand the current effect of blended learning, analyze learners' learning situation, and determine the needs and goals of constructing a profile. The effect of blended learning is comprehensively judged through the academic performance data of the school's academic affairs office and the blended learning situation questionnaire; the analysis of learners' learning situation is based on the learners' overall academic performance and their learning situation at the beginning of the course, and is analyzed from the academic affairs office's grade data, platform learning data, and early (initial stage) classroom behavior data. If there are differences in the blended effects among learners or there is a certain distance between the overall effect and expectations, a learner profile can be constructed. The goal of determining the profile construction is to objectively and clearly evaluate the learners through the profile, and then provide accurate learning intervention services for the learners' learning to improve the effect of blended learning.

[0028] (2) Collect and process learners' data. To ensure the effectiveness of the data, it is necessary to determine the data sources and content according to the actual conditions and uses. The practice data of this invention mainly comes from the online learning platform and actual classroom teaching, including identity recognition data, classroom behavior data, online behavior data, questionnaire data, and learning result data. The time span of data collection is 4 months, and the specific data descriptions are as follows:

[0029] Identity recognition data: student ID number, gender. These data can be directly obtained from the university's Academic Affairs Office, where the student ID number serves as the student's identity identification code.

[0030] Classroom behavior data: The author records the learners' classroom behaviors in real time. The data includes learners' passive answering, active conversations, active questions, group discussions, classroom exercises, and report sharing.

[0031] Online behavior data: The students' online behaviors and interaction data from the Learning Pass network platform, including the number of course visits, the number of chapter learning times, the number of task points completed, the number of videos completed, the duration of watching teaching videos, the number of chapter test completions, the number of discussion posts, and the number of homework submissions.

[0032] Questionnaire data: Collect data on the learners' basic situation questionnaires, learning motivation scales, and learning style scales to provide a data basis for constructing learners' portraits; finally, collect learners' satisfaction data to analyze the intervention effects based on learners' portraits.

[0033] Learning result data: Include online learning scores, stage learning achievements, and final course scores. The online learning score is comprehensively calculated from the three items of the learners' video completion rate, unit test scores, and participation in discussion scores; the stage learning achievements include the number of usual homework submissions and the scores of each homework, as well as the classroom report ratings; the final course score is the learners' paper-based closed-book exam score.

[0034] The original collected data has poor differences in dimensions and value ranges and cannot be directly applied. Therefore, it is necessary to organize the initial data. The data processing methods in practice are as follows: Store the data in an SQL database for data processing and access. Use SPSS25 for data analysis, Python 3.9 for data processing and visualization of learners' portraits, and the classroom interaction coding system ITIAS for collecting and screening classroom behavior data, encoding and converting, and analyzing classroom interaction behaviors.

[0035] (3) Portrait modeling. Combine existing portrait models, expert evaluation methods, etc. to determine the dimensions of the portrait.

[0036] The present invention analyzes the existing learner profile models, which mainly include four dimensions: basic information, interest preferences, behavioral interactions, and learning outcomes. The practice targets college students and needs to examine higher-order thinking abilities. Combining the suggestions of blended learning instructors and profile research experts, the dimensions of the profile are set to six dimensions: basic information, learning preparation, learning style, learning behavior, learning outcomes, and thinking ability. The profile dimensions are more comprehensive, the analysis of the learning situation is more specific, and the learning intervention guidance provided for students is more targeted. To represent the characteristics of learners in detail, tags in the form of short texts are determined for each dimension. Basic information includes the student ID, gender, grade, and major of the learner; learning preparation includes knowledge base and learning motivation; learning style includes cognitive preference and learning time preference; learning behavior includes participation, activity, and concentration; learning outcomes include knowledge mastery and satisfaction; thinking ability includes problem-solving, thinking innovation, and critical decision-making. Since the computer cannot directly recognize, process, and analyze the text, it is necessary to determine the specific data content of the profile tags in combination with the collected data. Basic information includes student ID, gender, grade, major, and course class data; knowledge base includes the average grade of the previous academic year, the number of blended learning semesters, and grades; learning motivation includes the measurement data of the learning motivation scale; cognitive preference data is determined through the learning style scale data and corrected based on the learner's online learning behavior; the learning time preference uses the time zone with the most frequent learning; participation includes the number of learning resource views, video learning times, and classroom sign-in times; concentration is judged by the video quiz accuracy rate, homework completion rate, and scores; knowledge mastery includes the assessment of unit test and final test scores; satisfaction is the result data of the learner satisfaction questionnaire; problem-solving is through the behavior record of students solving problems in class, attitude analysis, and the answer situation of problem-solving questions in the test paper; thinking innovation is through the creative presentation of students in class demonstrations, the speculative ability and innovative viewpoints put forward in homework, and the answer situation of innovative thinking questions in the test; critical decision-making is through the relevance of information collected by students, the participation degree in group discussions, and the number of contribution results.

[0037] (4) Output and analysis of the learner profile. The output forms of the learner profile mainly include text tables and graphic charts. In the practice of the present invention, the graphic chart visualization method is adopted to present the learner profile, and bar charts, scatter plots, radar charts, etc. are comprehensively used. The practical tool is Python analysis and SPSS. In order to conduct in-depth analysis of the profile in the later stage, correlation analysis is carried out on each dimension.

[0038] (5) Implement interventions based on the learner profile. Conduct interventions throughout the entire process of blended learning. At the learning preparation stage, formulate learning goals for each group of students based on the group profile, position each small goal of each group from the perspectives of behavior, results, and capabilities, formulate and revise corresponding test questions and learning tasks to check whether students have achieved the small goals, gradually achieve the main goal, improve students' behavioral engagement, enhance their knowledge mastery, and develop their thinking abilities; overall, guide students to understand the course, understand the course value, correct their learning attitudes towards the course, inform learners of the usage methods of the learning platform and learning resources and learning methods, reduce technical barriers, and improve learning efficiency; for students with distinct personalized characteristics, it is also necessary to analyze their individual profiles, set appropriate learning goals, convey the goal requirements and teacher expectations to the learners to help them clarify their learning directions and enhance their learning motivation. At the autonomous learning stage, teachers upload and release online learning resources and knowledge quizzes on the platform; learners raise questions in combination with the autonomous learning process. For personalized questions, teachers answer them immediately; for common questions, they are jointly discussed in class. Supervise and warn the overall learners. Access the learners' learning resource browsing situation from the platform background, count the class average level, and give reminders when the individual student's video viewing differs greatly from the average level, and give warnings when the task deadline countdown reaches 3 days. Provide assistance when students encounter problems in terms of technology and learning resources. Students put forward their needs for learning resources in the community, and teaching assistants promptly push relevant materials to the students; open the mind map production module on the platform. The mind map module consists of two parts: students independently making mind maps and teachers making and pushing them. At the consolidation and internalization stage, teachers explain the common problems raised in online learning in class, check the effects of autonomous learning, provide individual guidance and tutoring to learners with poor situations, reproduce the knowledge structure, provide bridging tools such as scaffolds and mind maps to prompt students to construct connections between old and new knowledge, and promote students' learning participation and cognitive interaction through question discussion; according to the analysis results of the student profile, feedback individual learning reports to the learners; optimize classroom teaching design, provide real problem-solving cases, increase classroom interaction, and prepare conditions for learning transfer. At the application and transfer stage, focus on cultivating students' higher-order thinking abilities. Realistic scenario-based tasks can be added and students are guided to solve them. Design classroom learning questions or tasks, let students think independently first, and then collaborate and communicate with their partners. To ensure the high-quality completion of tasks, teachers provide task guidance and recommend learning resources to assist learners in efficiently carrying out task exploration. The learning activities at this stage include assignment submission, peer review, self-reflection, and self-regulation.The intervention involves evaluation strategies, portrait feedback, and self-regulation. The process is as follows: Teachers evaluate and grade the assignments during the learning process, and in-class presentations are scored by other students according to a scoring scale and commented on by the teaching teacher; students achieve self-evaluation through personal portrait feedback. The learner portrait helps learners understand themselves more clearly and comprehensively. Learners compare the portrait with themselves, examine their learning behaviors and achievements, and thus gain a deeper understanding of their strengths and weaknesses; by integrating the results of teacher, peer evaluation, and self-reflection, learners consciously adjust and optimize their learning strategies to improve learning efficiency and quality. This two-way feedback mechanism can promote the improvement of learners' evaluation ability and self-regulation ability, and stimulate learners' thirst for knowledge and enterprising spirit.

[0039] (6) Evaluation of the intervention effect based on the learner portrait. In the practice of this invention, to verify the application value of the learner portrait and the effectiveness of the intervention model, a satisfaction questionnaire was distributed to learners. The results showed that students were more satisfied with the learning method after the intervention; the grades of the experimental class were better than those of the control class; the results of the interviews with teaching teachers indicated that the use of the learner portrait in blended learning in colleges and universities could improve the cultivation of students' comprehensive qualities, but the workload would increase, and it was suitable for use in dual-teacher classrooms. In colleges and universities, the general way of conducting courses is for teachers to teach and teaching assistants to assist. Therefore, it is very suitable to carry out portrait-based learning intervention work in colleges and universities.

[0040] Furthermore, the Pearson coefficient is used in the output and analysis of the learner portrait, and the calculation formula is as follows:

[0041]

[0042] The analysis result is that there is no linear correlation between the basic information and each dimension, so it does not participate in the subsequent data analysis. In practice, the overall learner portrait, group portrait, and individual portrait are output.

[0043] Furthermore, the overall learner portrait analyzes the learners' data from each dimension and presents the overall situation of the class.

[0044] This invention divides learning preparation into two dimensions: knowledge foundation and learning motivation. The knowledge foundation reflects the initial knowledge level of learners entering learning, and the specific content is the average grade of the previous school year, blended learning experience, and grades. The average grade of the previous school year and blended learning grades are divided into four levels: excellent (90 points and above), good (80 - 90 points), medium (70 - 80 points), qualified (60 points and above); the blended learning experience is divided into four categories by the number of semesters. The learning motivation reflects the initial state and motivation of learners' learning, and the specific content is reflected through questionnaires, which can be adjusted according to the actual situation of students in practical applications.

[0045] Relevant data on learning motivation were collected by distributing a measurement scale for college students' learning motivation to students. The scale was revised from the compiled learning motivation scale to determine its applicability in China, and its reliability and validity have been verified during the compilation of the scale. The scale divides learning motivation into multiple dimensions, but in this invention, the focus is on the overall level of learners' learning motivation. The final score of the scale is the level of learning motivation.

[0046] Learning time preference was understood through learning platform data; the FSLSM learning style model is divided into four dimensions: information processing, information perception, information input, and information understanding, with two types in each dimension. Information processing includes active and reflective types; information perception includes perceptive and intuitive types; information input includes visual and verbal types; information understanding includes sequential and comprehensive types. The FLSM learning style scale has a total of 44 questions, with an average of 11 questions in each dimension. Each question has two options, A and B, representing the two types in each dimension. The number of A and B in each dimension of the scale was counted and recorded, and the difference in the number of the letters A and B was compared. The larger the number of letters, the stronger the degree of the preferred type. The difference was obtained by subtracting the smaller number from the larger number, and the difference value and letter were used to represent the learning style type of the student. Considering that the difference in the number of letters of the two types of some students is very small and there is no prominent learning style, an equilibrium type (difference <= 2) was added to each dimension of the learning style to better reflect the characteristics of such learners.

[0047] Learning behavior data include online behavior data and classroom behavior data. The learning behaviors in the study are divided into participation, activity, and concentration. In terms of participation, the online learning part includes learning resources, micro-courses, and application extension resources. The completion scores of course task points and the number of chapter learning times of each student were counted. In the offline learning part, the number of times students signed in for class was counted. In terms of activity, an online community was established for students by class, and the online interaction behavior data of all students were counted. The classroom behavior data were obtained through classroom observation records, mainly recording students' passive answers, active speeches, group discussions, classroom exercises, and report sharing, and the classroom activity behavior data of each student were counted. In terms of concentration, the initial video quizzes and homework data of students were recorded, and the homework and quizzes submitted after the deadline were not counted.

[0048] The learning results of the initial portrait are analyzed based on the learner's previous blended learning performance and the existing comprehensive online learning performance. The previous blended learning performance can reflect the learner's adaptability to the blended learning method. The existing comprehensive online learning performance can examine the learner's knowledge mastery level and learning behavior changes in the ongoing blended learning courses. By conducting a correlation analysis between behavior and performance, we can understand the impact of behavior on performance, and thus attempt to improve the learner's behavior to enhance learning performance. The blended learning performance can reflect the learner's existing blended learning level and provide a basis for learning intervention. The existing comprehensive learning performance includes course video viewing, chapter quizzes, chapter learning times, assignments, and classroom learning performance, which can reflect the learner's knowledge mastery level of the ongoing courses.

[0049] The initial thinking ability is analyzed based on the classroom performance and online problem-solving data of students in the initial stage. Problem-solving includes the correct rate of video questions, assignment grading, the number of correct answers in class, and group contribution rate. Thinking innovation is evaluated by teachers based on students' assignments for their critical thinking ability and the creative presentation score of classroom demonstrations, and the correct rate of innovative thinking questions in students' unit quizzes is statistically analyzed. Critical decision-making includes evaluating the relevance of information collection by students in homework submissions and classroom tasks, the participation rate in group discussions, and the number of contribution results. Group contribution comes from the evaluations of group peers and teachers.

[0050] The specific process of forming the group portrait is to reduce the dimension of the data after normalization processing, map it to a two-dimensional space, cluster using the K-means algorithm, and construct group portraits for different groups. In the features, the basic information does not participate in the analysis, and the learning style updates slowly. It is divided into static indicators and used as a separate classification criterion. The data of learning behavior, learning preparation, and learning results are selected for classification. To ensure the unified standard of the data, the data participating in the classification is processed by Min-Max normalization. According to the data processing results, the principal component analysis method is used for dimension reduction. In the practice of the present invention, the cumulative percentage of PC1 and PC2 explanations reaches 63.14%, which is representative to a certain extent and can be used for clustering calculation. Using K-means clustering requires determining the K value first, that is, the number of categories. The sum of squared residuals is calculated using Python, and the K value is set from 2 to 9. The optimal K value is determined by the elbow method. The elbow method determines the number of clusters according to the distortion degree of the clusters. For data with a certain degree of differentiation, a large change will occur when reaching a certain critical value, and the downward trend will gradually slow down. When K is 4, there is an obvious change in the trend. Therefore, K = 4 is determined, that is, the experimental class is divided into 4 groups. After clustering by the K-means algorithm, there are 32 people in this experimental class, divided into 4 groups. The number of people in these four groups is counted. Group 1 has 14 people, with learning styles of active, perceptive, and visual, medium behavior participation, medium knowledge mastery, medium thinking innovation, and medium performance in critical decision-making and problem-solving; Group 2 has 7 people, with learning styles defined as active, perceptive, and holistic, high behavior participation, high knowledge mastery, the best performance in learning behavior and results, more innovation displays, and good performance in problem-solving and critical decision-making; Group 3 has 5 people, with learning styles of active, intuitive, and visual, low learning behavior participation, low knowledge mastery, low levels of innovation display, problem-solving, and critical decision-making, and the worst overall performance among the groups; Group 4 has 6 people, with learning styles of active, visual, and sequential, medium-high learning behavior participation, but low knowledge mastery, and general levels of thinking innovation, problem-solving, and critical decision-making. Calculate the average values of each group on each dimension index to obtain the average values of the four groups on each index and the overall average level. The learning preparation and learning result levels of Group 1 are relatively high, but the behavior investment is low. Such learners have poor self-behavior restraint, resulting in the learning results not reaching the best effect. Therefore, Group 1 is defined as "developmental learners"; Group 2 has the best overall performance, but there are deficiencies in participation and interaction. Therefore, Group 2 is defined as "excellent learners"; Group 3 has the worst overall performance. Although it is relatively prominent in knowledge mastery, there are still deficiencies compared with other groups. Therefore, Group 3 is defined as "marginal learners"; Group 4 performs well in behavior, actively completes homework and exercises, but the scores are low, probably because the learners' blended learning ability is weak. Therefore, Group 4 is defined as "hardworking learners".By mastering the problems of learners through group positioning, the efficiency of intervention can be improved.

[0051] The individual portrait is relatively more specific and accurate, and the portrait feedback for learners also relies on the individual portrait. The individual portrait is fed back to students in the form of a learning report to promote students' self-awareness and evaluation, and play a role in early warning and adjustment. The learner with student number 2022*223 is an effortful learner, with high learning enthusiasm, strong motivation, relatively high behavioral engagement, and strong thinking innovation ability. However, their knowledge foundation, learning results, and problem-solving ability are poor, which does not match the expectations. Their learning style is active and perceptive, tending to actively participate in learning, discussion, and communication, liking group cooperative learning, with active thinking and innovation; having good comprehension of specific matters; having better comprehension of specific matters, being very patient with details, taking a longer time to learn new knowledge, and being good at memorizing facts and some ready-made work. In terms of learning behavior, they are high-engagement learners, with an engagement level exceeding the average, being able to actively participate in learning activities and complete learning tasks, but lacking in terms of interaction, and not being good at communicating and discussing with classmates. In terms of knowledge foundation, they are below the overall level, with a weak foundation and lack of knowledge mastery. They show new highlights in classroom presentations, with novel homework materials, but the matching degree between the materials and tasks is low, and the problem-solving level can be further improved. Suggestions for the students are as follows: actively interact with teachers, strengthen cooperation with contemplative classmates, and find their own learning methods; the knowledge foundation needs to be improved, and watch the relevant learning materials on the platform in a timely manner; set appropriate learning goals, make a learning plan, and reasonably allocate learning time and attention; communicate with teachers in a timely manner when having questions during the blended learning process.

[0052] Another object of the present invention is to provide a learning intervention system in a blended learning environment for implementing the learning intervention method in the blended learning environment, including:

[0053] The learning intervention method in the blended learning environment provided by the embodiments of the present invention includes:

[0054] A portrait model construction module, used for constructing a portrait model;

[0055] An intervention framework building module, used for building an intervention framework based on the learner portrait in the blended learning environment;

[0056] A learning intervention module, used for implementing learning intervention.

[0057] Another object of the present invention is to provide a computer device, which includes a memory and a processor. When a computer program stored in the memory is executed by the processor, the processor executes the steps of the learning intervention method in the blended learning environment.

[0058] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to execute the steps of the learning intervention method in the hybrid learning environment.

[0059] Another object of the present invention is to provide an information data processing terminal, which includes the learning intervention system in the hybrid learning environment.

[0060] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0061] In view of the lack of personalized intervention in traditional methods, the effect of hybrid learning is improved. Therefore, in the hybrid learning mode, learner portrait technology is introduced into the hybrid learning process to guide the design and implementation of learning intervention. The learning portrait model for hybrid learning includes dimensions such as student basic information, learning preparation, learning style, learning behavior, learning results and thinking ability, collects and processes learner classroom and online learning data, and presents learner overall, group and individual portraits in a visual form. The teacher reprocesses the portrait, generates a learning report sheet and feeds it back to the student, diagnoses learner problems through portraits, forms a precise and personalized learning intervention framework and strategy, and implements overall, group and individual learning intervention strategies in combination with portraits throughout the whole process of hybrid learning. Some learning data changes in real time with the learning process and intervention feedback, so the portrait will be dynamically updated, and the "plan-intervention-observation-reflection" intervention practice is also a cyclic iterative process to ensure the learner's individual experience and improve the learning efficiency and learning effect of hybrid learning. The present invention can be migrated and applied to educational scenarios such as hybrid learning, classroom teaching online education, vocational training, etc., which helps to improve learning quality and efficiency while meeting the personalized needs of learners.

[0062] Most intervention strategies are only targeted at the individual or overall level, lacking group stratification design, resulting in uneven resource allocation. The invention proposes a three-level intervention framework of "overall-group-individual". Design general strategies for common needs as a whole. Identify typical learning feature groups through cluster analysis and provide differentiated support. Recommend adaptive learning resources and learning companions based on individual portraits. Existing interventions mostly rely on single behavioral data and ignore the linkage effect of emotions and cognitive states. Combine multimodal data such as motivational sentiment analysis and cognitive development changes to design a dynamic intervention trigger mechanism. For example, meditation guidance is automatically pushed when learning anxiety is detected, and the difficulty of learning tasks is adjusted when cognitive overload is found.

[0063] Traditional evaluations mainly focus on academic achievements and neglect the quantitative assessment of thinking abilities and emotional development. The technical solution of the present invention establishes a multi-dimensional evaluation system covering academic performance, motivation and emotion, problem-solving, innovation ability, and critical decision-making ability, and conducts comprehensive evaluation by combining quantitative data with qualitative analysis.

[0064] In a blended learning environment, emphasis is placed on providing barrier-free learning experiences for all learners. Through technical means, it is ensured that learners with different abilities and needs can equally participate in learning activities. Diversified learning resources and activity forms are provided to meet the preferences and learning styles of different learners. Such adaptable design helps to reduce the sense of exclusion caused by technology or content preferences. Description of the Drawings

[0065] Figure 1 It is a flowchart of the learning intervention method in the blended learning environment provided by an embodiment of the present invention.

[0066] Figure 2 It is a statistical chart of the publication volume of research on the portraits of foreign learners provided by an embodiment of the present invention.

[0067] Figure 3 It is a flowchart of constructing the learner portrait provided by an embodiment of the present invention.

[0068] Figure 4 It is a framework diagram of learning intervention based on the learner portrait provided by an embodiment of the present invention.

[0069] Figure 5 It is a structural diagram of the learning intervention system in the blended learning environment provided by an embodiment of the present invention.

[0070] Figure 6 It is an effect diagram of the improvement of the abilities of each learning group after the intervention provided by an embodiment of the present invention; (a) excellent learners, (b) developing learners, (c) striving learners, (d) marginal learners.

[0071] Figure 7 It is an effect diagram of the improvement of individual learning results after the intervention provided by an embodiment of the present invention.

[0072] Figure 8 It is an effect diagram of the improvement of individual thinking abilities after the intervention provided by an embodiment of the present invention. Detailed Embodiment

[0073] In order to make the purpose, technical solution and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in combination with embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention and are not used to limit the present invention.

[0074] As Figure 1As shown in the figure, the learning intervention method in the hybrid learning environment provided by the embodiments of the present invention includes:

[0075] 1. Portrait model construction process.

[0076] (1) Requirement analysis and goal determination: Requirement analysis can help the present invention set goals, review data, determine solutions, and prepare for project evaluation. The needs for learner portraits in the hybrid learning environment can be divided into three aspects: ① Learner analysis. By analyzing learners through portraits, their personality characteristics can be understood, and the learning change trends can be grasped to meet the learning development needs. ② Learning services. Learning services based on portraits include learning resource recommendation services, personalized learning path customization, learning partner pairing services, learning behavior supervision and early warning, etc. ③ Learning evaluation. Learner portraits can be used for personal evaluation of learners, and teachers' evaluation and analysis of the whole class or group. The goal of portrait construction is to objectively and clearly evaluate learners through portraits, and then provide accurate learning intervention services for learners' learning.

[0077] (2) Data collection and processing: Data is the basis for constructing portraits and can determine the effectiveness of portraits. The learning data of the present invention comes from the online learning platform Xuexitong, the learning management system, students' homework and test papers, and learners' questionnaires. Therefore, the types of collected data are rich and diverse, and the data format needs to be unified through operations such as data cleaning, data standardization, and data conversion so that researchers can better store and utilize the data. The learning data in the research can be divided into four categories:

[0078] Information survey data includes learners' identity information, learning styles, and learning motivations.

[0079] Learning behavior data is a record of learners' behavior activities and participation in the classroom and online environments.

[0080] Learning result data is a description of learners' phased achievements, test scores, etc.

[0081] Thinking ability data is an embodiment of students' higher-level cognitive levels, including multiple elements such as problem-solving, thinking innovation, and critical decision-making. These data are the objects for learner portrait analysis.

[0082] (3) Portrait modeling and analysis: Portrait modeling and analysis is to mine information from learners' data and is the core and key in the portrait construction process. The dimension selection of the portrait model is mainly determined by the construction goal, and the learners' data is labeled according to the dimensions.

[0083] (4) Portrait output and visualization: In order to give full play to the educational function of learner portraits, learner characteristics need to be presented in a visualized manner. In the present invention, individual learner portraits, group learner portraits, and overall learner portraits are generated.

[0084] (5) Portrait application and evaluation: The construction of learner portraits ultimately serves learning intervention practices. Learners’ behaviors and outcomes will change over time, and portraits also need to be updated to achieve the original construction goals.

[0085] 2. Intervention framework based on learner profile in blended learning environment.

[0086] Provide the most appropriate learning intervention for each student based on the learner profile. The intervention framework is as follows:

[0087] Conduct learning situation analysis and diagnose learning problems based on learner portraits; implement interventions in the whole process of learning preparation, autonomous learning, consolidation and internalization, application transfer, and evaluation feedback in blended learning based on the principles of student subjectivity, accuracy, embeddedness, and effectiveness. Design overall intervention strategies based on learner portraits, including learning supervision and early warning, technical and resource guidance, and mind mapping; group intervention strategies include goal positioning and value guidance, teacher expectation transmission, task design and guidance, activity strategies, resource recommendations, and evaluation strategies; individual intervention strategies include individual reminder strategies, portrait feedback strategies, and self-reflection and adjustment strategies. Evaluate the effectiveness of intervention practices based on changes in self-portraits, differences between individual and group portraits, changes before and after the overall portrait, and comparison of the overall portraits of each class; distribute satisfaction questionnaires to students, and interview teachers to understand the emotional experience of teachers and students on intervention practices based on learner portraits.

[0088] 3. Implementation of learning interventions.

[0089] The implementation method of learning intervention based on learner portrait in a hybrid learning environment is described here in conjunction with the experiment of the present invention.

[0090] (1) Understand the current effect of blended learning, analyze learners' learning situation, and determine the needs and goals of constructing a profile. The effect of blended learning is comprehensively judged through the academic performance data of the school's academic affairs office and the blended learning situation questionnaire; the analysis of learners' learning situation is based on the learners' overall academic performance and their learning situation at the beginning of the course, and is analyzed from the academic affairs office's grade data, platform learning data, and early (initial stage) classroom behavior data. If there are differences in the blended effects among learners or there is a certain distance between the overall effect and expectations, a learner profile can be constructed. The goal of determining the profile construction is to objectively and clearly evaluate the learners through the profile, and then provide accurate learning intervention services for the learners' learning to improve the effect of blended learning.

[0091] (2)Collect and process learners' data. To ensure the effectiveness of the data, it is necessary to determine the data sources and content according to the actual conditions and uses. The practical data of this invention mainly comes from online learning platforms and actual classroom teaching, including identity recognition data, classroom behavior data, online behavior data, questionnaire data, and learning result data. The time span of data collection is 4 months, and the specific data descriptions are as follows:

[0092] Identity recognition data: student ID and gender, which can be directly obtained from the university's Academic Affairs Office. The student ID serves as the identity identification code for students.

[0093] Classroom behavior data: The author records the learners' classroom behaviors in real time. The data includes learners' passive answering, active conversations, active questions, group discussions, classroom exercises, and presentation sharing.

[0094] Online behavior data: The students' online behaviors and interaction data from the Learning Pass network platform, including the number of course visits, the number of chapter studies, the number of task points completed, the number of videos completed, the duration of watching teaching videos, the number of chapter tests completed, the number of discussion posts, and the number of homework submissions.

[0095] Questionnaire data: Collect data on learners' basic situation questionnaires, learning motivation scales, and learning style scales to provide a data basis for constructing learner portraits; finally, collect learners' satisfaction data to analyze the intervention effects based on learner portraits.

[0096] Learning result data: Include online learning scores, stage learning achievements, and final course scores. The comprehensive video completion rate, unit test scores, and discussion participation scores of learners are used as the online learning scores of learners; the stage learning achievements include the number of usual homework submissions and the scores of each homework, and the classroom presentation scores; the final course score is the learners' paper-based closed-book exam score.

[0097] The original collected data has poor differences in dimensions and value ranges and cannot be directly applied. Therefore, it is necessary to organize the initial data. The data processing methods in practice are as follows: Store the data in an SQL database for data processing and access. Use SPSS25 for data analysis, Python 3.9 for data processing and visualization of learner portraits, and the classroom interaction coding system ITIAS for collecting and screening classroom behavior data, coding and converting, and analyzing classroom interaction behaviors.

[0098] (3)Image modeling. Combine existing image models, expert evaluation methods, etc. to determine the dimensions of the image. In this invention, after analyzing existing learner image models, they mainly include four dimensions: basic information, interest preferences, behavioral interactions, and learning outcomes. The object of practice is college students, and higher-order thinking abilities need to be examined. Combining the suggestions of blended learning teachers and image research experts, the dimensions of the image are set to six dimensions: basic information, learning preparation, learning style, learning behavior, learning outcomes, and thinking ability. The image dimensions are more comprehensive, the analysis of the learning situation is more specific, and the learning intervention guidance provided for students is more targeted. To represent the characteristics of learners in detail, determine labels in the form of short texts for each dimension. Basic information includes the student ID, gender, grade, and major of the learner; learning preparation includes knowledge foundation and learning motivation; learning style includes cognitive preferences and learning time preferences; learning behavior includes participation, activity, and concentration; learning outcomes include knowledge mastery and satisfaction; thinking ability includes problem-solving, thinking innovation, and critical decision-making. Since the computer cannot directly recognize, process, and analyze text, it is necessary to determine the specific data content of the image labels in combination with the collected data. Basic information includes student ID, gender, grade, major, and course class data; knowledge foundation includes the average grade of the previous academic year, the number of blended learning semesters, and grades; learning motivation includes data measured by the learning motivation scale; cognitive preference data is determined through learning style scale data and corrected according to the learner's online learning behavior; learning time preference uses the time zone with the most frequent learning; participation includes the number of learning resource views, video learning times, and classroom sign-in times; concentration is judged by the video quiz accuracy rate, homework completion rate, and scores; knowledge mastery includes the evaluation of unit test and final test scores; satisfaction is the result data of the learner satisfaction questionnaire; problem-solving is based on the behavior records of students solving problems in class, attitude analysis, and the answering situation of problem-solving questions in the test paper; thinking innovation is based on the creative presentations of students in class demonstrations, the speculative ability and innovative viewpoints put forward in homework, and the answering situation of innovative thinking questions in the test; critical decision-making is based on the relevance of information collected by students, the participation in group discussions, and the number of contribution results.

[0099] (4)Output and analysis of the learner image. The output forms of the learner image mainly include text tables and graphics. In the practice of this invention, the learner image is presented in a visual way of graphics, comprehensively using bar charts, scatter plots, and radar charts, etc. The practical tool is Python analysis and SPSS. To conduct in-depth analysis of the image in the later stage, correlation analysis is carried out on each dimension. In practice, the Pearson coefficient is used, and the calculation formula is as follows:

[0100]

[0101] The analysis result shows that there is no linear correlation between the basic information and each dimension, so it does not participate in the subsequent data analysis. In practice, the overall portrait of learners, group portrait, and individual portrait are output.

[0102] The overall portrait of learners analyzes the data of learners from each dimension and presents the overall situation of the class.

[0103] 1) Dimension of learning preparation;

[0104] In the present invention, learning preparation is divided into two dimensions: knowledge base and learning motivation. The knowledge base reflects the initial knowledge level of learners entering learning, and the specific content is the average grade of the previous academic year, mixed learning experience, and grades. The average grade of the previous academic year and mixed learning grades are divided into four levels: excellent (90 points and above), good (80 - 90 points), medium (70 - 80 points), qualified (60 points and above); the mixed learning experience is divided into four categories by the number of semesters. The learning motivation reflects the initial state and motivation of learners' learning, and the specific content is reflected through questionnaires, and can be adjusted according to the actual situation of students in practical applications.

[0105] Relevant data on learning motivation are collected by distributing a measurement scale of college students' learning motivation to students. The scale is revised by the compiled learning motivation scale to determine its applicability in China, and its reliability and validity have been verified when compiling the scale. The scale divides learning motivation into multiple dimensions, but in the present invention, the focus is on the overall level of learners' learning motivation. The final score of the scale is the level of learning motivation.

[0106] 2) Dimension of learning style;

[0107] The preference for learning time is understood through learning platform data; the FSLSM learning style model is divided into four dimensions: information processing, information perception, information input, and information understanding, and each dimension contains two types. Information processing includes active type and reflective type; information perception includes perceptive type and intuitive type; information input includes visual type and verbal type; information understanding includes sequential type and comprehensive type. The FSLSM learning style scale has a total of 44 questions, with an average of 11 questions for each dimension. Each question has two options, A and B, representing the two types of each dimension respectively. Count and record the number of A and B in each dimension of the scale, compare the difference in the number of letters A and B. The more the number of letters, the stronger the degree of the preferred type. Subtract the smaller number from the larger number to get the difference value, and use the difference value and letters to represent the learning style type of students. Considering that the difference in the number of letters of the two types of some students is very small and there is no prominent learning style, an equilibrium type (difference value <= 2) is added to each dimension of the learning style to better reflect the characteristics of such learners.

[0108] 3) Dimension of learning behavior;

[0109] Learning behavior data includes online behavior data and classroom behavior data. The learning behaviors in the study are divided into participation, activity, and concentration. In terms of participation, the online learning part includes learning resources, micro-lessons, and application extension resources. The completion scores of course task points and the scores of chapter learning times of each student are statistically analyzed. In the offline learning part, the number of times students sign in for class is statistically analyzed. In terms of activity, an online community is established for students by class, and the online interaction behavior data of all students is statistically analyzed. The classroom behavior data is obtained through classroom observation records, mainly recording students' passive answers, active speeches, group discussions, classroom exercises, and report sharing, and statistically analyzing the classroom activity behavior data of each student. In terms of concentration, the initial video quizzes and homework data of students are recorded, and the homework and quizzes submitted after the time limit are not statistically analyzed.

[0110] 4) Learning outcome dimension;

[0111] The learning outcomes of the initial portrait are analyzed based on the learners' previous blended learning achievements and existing online learning comprehensive scores. The previous blended learning achievements can reflect the learners' adaptability to the blended learning method. The existing online learning comprehensive scores can examine the learners' knowledge mastery level and learning behavior changes in the current blended learning courses. By conducting a correlation analysis between behaviors and scores, the impact of behaviors on scores can be understood, and thus attempts can be made to improve learners' behaviors to enhance learning performance. The blended learning achievements can reflect the learners' existing blended learning level and provide a basis for learning intervention. The existing learning comprehensive scores include course video views, chapter quizzes, chapter learning times, homework, and classroom learning performance, which can reflect the learners' knowledge mastery level of the current courses.

[0112] 5) Thinking ability dimension;

[0113] The initial thinking ability is analyzed based on the classroom performance and online question answering data of students at the initial stage. Problem-solving includes the correct rate of video questions, homework grading, the number of correct answers in class, and the group contribution rate. Thinking innovation is that teachers evaluate the students' speculative ability based on their homework and the creative presentation score of classroom demonstrations, and statistically analyze the correct rate of innovative thinking questions in the students' unit quizzes. Critical decision-making includes evaluating the relevance of information collection by students in homework submissions and classroom tasks, the participation rate in group discussions, and the number of contribution results. The group contribution comes from the evaluations of group peers and teachers.

[0114] The specific process of forming the group portrait is to reduce the dimension of the data after normalization processing, map it to a two-dimensional space, use the K-means algorithm for clustering, and construct group portraits for different groups. Among the features, the basic information does not participate in the analysis, and the learning style updates slowly. It is divided into static indicators and used as a separate classification criterion. The data of learning behavior, learning preparation, and learning results are selected for classification. To ensure the unified standard of the data, the data participating in the classification is processed by Min-Max normalization here. According to the data processing results, the principal component analysis method is used for dimension reduction. In the practice of the present invention, the cumulative percentage of PC1 and PC2 explanations reaches 63.14%, which has a certain representativeness and can be used for clustering calculation. Using K-means clustering requires determining the K value first, that is, the number of categories. Use Python to calculate the sum of squared residuals, set the K value from 2 to 9, and determine the optimal K value through the elbow method. The elbow method determines the number of clusters according to the distortion degree of the clusters. For data with a certain degree of discrimination, a large change will occur when reaching a certain critical value, and the downward trend will gradually slow down. When K is 4, the trend changes significantly, so K = 4 is determined, that is, the experimental class is divided into 4 groups. After clustering by the K-means algorithm, there are 32 people in this experimental class, divided into 4 groups. The number of people in these four groups is counted. Group 1 has 14 people, with learning styles of active, perceptive, and visual, medium behavior participation, medium knowledge mastery, medium thinking innovation, and medium performance in critical decision-making and problem-solving; Group 2 has 7 people, with learning styles defined as active, perceptive, and holistic, high behavior participation, high knowledge mastery, the best performance in learning behavior and results, more innovation displays, and good performance in problem-solving and critical decision-making; Group 3 has 5 people, with learning styles of active, intuitive, and visual, low learning behavior participation, low knowledge mastery, low levels of innovation display, problem-solving, and critical decision-making, and the overall performance is the worst among the groups; Group 4 has 6 people, with learning styles of active, visual, and sequential, medium-high learning behavior participation, but low knowledge mastery, and general levels of thinking innovation, problem-solving, and critical decision-making. Calculate the average values of each group on each dimension index to obtain the average values of the four groups on each index and the overall average level. The learning preparation and learning result levels of Group 1 are relatively high, but the behavior investment is low. Such learners have poor self-discipline in their behaviors, resulting in the learning results not reaching the best effect. Therefore, Group 1 is defined as "developmental learners"; Group 2 has the best overall performance, but there are deficiencies in participation and interaction. Therefore, Group 2 is defined as "excellent learners"; Group 3 has the worst overall performance. Although it is relatively prominent in knowledge mastery, there are still deficiencies compared with other groups. Therefore, Group 3 is defined as "marginal learners"; Group 4 performs well in behavior, actively completes homework and exercises, but has a low score, possibly because the learners' blended learning ability is weak. Therefore, Group 4 is defined as "hardworking learners".Master the problems of learners through group positioning of learners to improve the efficiency of intervention.

[0115] The individual portrait is relatively more specific and accurate, and the portrait feedback for learners also relies on the individual portrait. The individual portrait is feedback to students in the form of a learning report to promote students' self-awareness and evaluation, and play a role in early warning and adjustment. The learner with student number 2022*223 is an effortful learner, with positive learning attitude, strong motivation, high behavioral engagement, and relatively strong thinking innovation ability. However, their knowledge foundation, learning outcomes, and problem-solving ability are poor, which does not match the expectations. Their learning style is active and perceptive, tending to actively participate in learning, discussion and communication, liking group cooperative learning, with active thinking and innovation; having good comprehension of specific matters; having better comprehension of specific matters, being patient with details, taking a longer time to learn new knowledge, and being good at memorizing facts and some ready-made work. In terms of learning behavior, they are high-engagement learners, with engagement exceeding the average level, being able to actively participate in learning activities and complete learning tasks, but lacking in terms of interaction, and not being good at communicating and discussing with classmates. In terms of knowledge foundation, they are below the overall level, with weak foundation and lack of knowledge mastery. They show new highlights in classroom presentations, with novel homework materials, but the matching degree between the materials and tasks is low, and the problem-solving level can be further improved. Suggestions for the students are as follows: actively interact with teachers, strengthen cooperation with contemplative classmates, and find out their own learning methods; the knowledge foundation needs to be improved, and watch the relevant learning materials on the platform in a timely manner; set appropriate learning goals, make a learning plan, and reasonably allocate learning time and attention; communicate with teachers in a timely manner when having questions during the blended learning process.

[0116] Implement interventions based on learner profiles. Conduct interventions throughout the entire process of blended learning. At the learning preparation stage, formulate learning goals for each group of students based on the group profile, position each sub-goal of each group from the perspectives of behavior, results, and capabilities, formulate and revise corresponding test questions and learning tasks to check whether students have achieved the sub-goals, gradually achieve the main goal, improve students' behavioral engagement, enhance their knowledge mastery, and develop their thinking abilities; overall, guide students to understand the course, understand the value of the course, correct their learning attitudes towards the course, inform learners of the usage methods of the learning platform and learning resources and learning methods, reduce technical barriers, and improve learning efficiency; for students with distinct individual characteristics, it is also necessary to analyze their individual profiles, set appropriate learning goals, convey the goal requirements and teachers' expectations to the learners to help them clarify their learning directions and enhance their learning motivation. At the autonomous learning stage, teachers upload and release online learning resources and knowledge quizzes on the platform; learners raise questions in combination with the autonomous learning process. For individual problems, teachers answer them immediately; for common problems, they are discussed together in class. Supervise and give early warnings to all learners, access the learning resource browsing situation of students from the platform background, and calculate the class average level. Give reminders when the difference between an individual student's video viewing and the average level is large, and give alerts when there are 3 days left until the task deadline. Provide assistance when students encounter problems in terms of technology and learning resources. Students put forward their needs for learning resources in the community, and teaching assistants promptly push relevant materials to students; open the mind map making module on the platform, which consists of two parts: students making mind maps independently and teachers making and pushing them. At the consolidation and internalization stage, teachers explain the common problems raised in online learning in class, check the effects of autonomous learning, provide individual guidance and tutoring to learners with poor performance, reproduce the knowledge structure, provide bridging tools such as scaffolds and mind maps, and prompt students to construct connections between old and new knowledge. Promote students' learning participation and cognitive interaction through question-and-discussion methods; according to the analysis results of the student profile, provide learners with personal learning reports; optimize classroom teaching design, provide real problem-solving cases, increase classroom interaction, and prepare conditions for learning transfer. At the application and transfer stage, focus on cultivating students' higher-order thinking abilities. Realistic scenario-based tasks can be added and students are guided to solve them. Design classroom learning questions or tasks, let students think independently first, and then collaborate and communicate with their partners. To ensure the high-quality completion of tasks, teachers provide task guidance and recommend learning resources to assist learners in efficiently carrying out task exploration. The learning activities at this stage include assignment submission, peer review, self-reflection, and self-regulation.The intervention involves evaluation strategies, portrait feedback, and self-regulation. The process is as follows: Teachers evaluate and grade the assignments during the learning process, and in-class presentations are scored by other students according to a scoring scale and commented on by the teaching teacher; students achieve self-evaluation through personal portrait feedback. The learner portrait helps learners understand themselves more clearly and comprehensively. Learners compare the portrait and examine their learning behaviors and achievements, so as to understand their advantages and disadvantages more deeply; by integrating the results of teacher, peer evaluation, and self-reflection, learners consciously adjust and optimize their learning strategies to improve learning efficiency and quality. This two-way feedback mechanism can promote the improvement of learners' evaluation ability and self-regulation ability, and stimulate learners' thirst for knowledge and enterprising spirit.

[0117] As Figure 5 shown, the learning intervention system in the hybrid learning environment provided by the embodiment of the present invention includes:

[0118] The learning intervention method in the hybrid learning environment provided by the embodiment of the present invention includes:

[0119] A portrait model construction module for constructing a portrait model;

[0120] An intervention framework building module for building an intervention framework based on the learner portrait in the hybrid learning environment;

[0121] A learning intervention module for implementing learning intervention.

[0122] The application embodiment of the present invention provides a computer device. The computer device includes a memory and a processor. When the computer program stored in the memory is executed by the processor, the processor executes the steps of the learning intervention method in the hybrid learning environment.

[0123] The application embodiment of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by the processor, the processor executes the steps of the learning intervention method in the hybrid learning environment.

[0124] The application embodiment of the present invention provides an information data processing terminal, and the information data processing terminal includes a learning intervention system in the hybrid learning environment.

[0125] The specific application field of the present invention is to optimize the learning process in the hybrid learning environment, conduct personalized learning intervention according to the specific situation of learners, and improve learning efficiency and learning effectiveness.

[0126] In the practice of this invention, to verify the application value of the learner portrait and the effectiveness of the intervention model, a satisfaction questionnaire was distributed to learners. The results showed that students were more satisfied with the learning method after the intervention; the scores of the experimental class were better than those of the control class; the interview results of the teaching teachers showed that the use of the learner portrait in the mixed learning in colleges and universities could improve the comprehensive quality cultivation of students, but the workload would increase, and it was suitable for use in dual-teacher classrooms. Generally, in colleges and universities, the courses are carried out in the way of teachers teaching and teaching assistants assisting. Therefore, it is very suitable to carry out the learning intervention work based on the portrait in colleges and universities.

[0127] In the process of the portrait construction process of this invention, the needs analysis and goal determination are added. This stage is the first step of the portrait construction process, which can provide clear guidance for the subsequent links, reduce the redundant work of data collection, improve the effectiveness of data modeling, and establish standards for evaluating and modifying the portrait. The dimension of the portrait adds a thinking ability module to adapt to the new standards of the quality of talents in the knowledge era.

[0128] The intervention application value based on the learner portrait has been verified in practice. In practice, it was determined that after the intervention, in the first experiment, the average score of the experimental class was 87.83, and the average score of the control class was 79.93, with a difference of 7.9 points; the standard deviation of the scores of the experimental class was 7.09, and the standard deviation of the scores of the control class was 18.346; through the independent sample test, the significance = 0.005 < 0.05, not assuming equal variances, and the t-test sig (two-tailed) = 0.033 < 0.05. It can be seen that there is a significant difference between the experimental class and the control class, the overall scores of the experimental class are better than those of the control class, and the development of the students' scores is more balanced. In the second experiment, the average score of the experimental class was 87.34, and the average score of the control class was 82.13, with a difference of 5.21; the standard deviation of the scores of the experimental class was 6.667, and the standard deviation of the scores of the control class was 11.390; through the independent sample test, the significance = 0.106 > 0.05, assuming equal variances, and the t-test sig (two-tailed) = 0.035 < 0.05. It can be seen that there is a significant difference in the scores of the experimental class and the control class in Experiment 2 after the intervention, the overall average level of the experimental class is better than the average level of the control class, and the student level of the experimental class is more balanced. Through these two groups of experiments, it can be concluded that implementing the intervention using the learner portrait in the mixed learning environment can improve the students' learning scores and reduce the score differences among individual students.

[0129] Table 1 Statistical Table of Experiment 1

[0130]

[0131] Table 2 Independent Sample Test of Experiment 1

[0132]

[0133] Table 3 Statistical Table of Experiment 2

[0134]

[0135] Table 4 Independent Samples Test for Experiment 2

[0136]

[0137] From the data analysis, the situations of learners before and after intervention can be compared as a whole. To further intuitively understand the changes, by outputting the group portrait and individual portrait after intervention, a better comparison can be made.

[0138] (1) Group portrait;

[0139] Before comparison, the data of each dimension of each group are normalized, and the average value of each data item is calculated, as shown in Table 5. The data before and after intervention are plotted in a radar chart, as Figure 6 shown. Excellent learners have improvements in all dimensions. Among them, the biggest change is in learning behavior, and there are small improvements in learning results and thinking ability; Developmental learners have obvious changes in behavioral participation. Although there is progress in activity level, there is still a large room for improvement, and the changes in thinking ability and learning results are small, and they can be further improved; Effortful learners actively participate in learning and have improvements in learning behavior, thinking ability and learning results; Among the four types of learners, the changes of marginal learners are the largest. Although their initial level is low, they have significant improvements in learning behavior and thinking ability, indicating that these students are mentally active and still need to be concerned about in the later learning to promote better progress.

[0140] Table 5 Group Portrait Data

[0141]

[0142] (2) Individual portrait;

[0143] The initial portrait dimensions of individual learning results include knowledge base, learning motivation and knowledge mastery. The later portrait needs to be updated by deleting the knowledge base and including learning motivation, classroom observation scores and test scores. Compare the individual learning results before and after intervention and the overall learning result level after intervention. The comparison results are as Figure 7 shown. After individual intervention, there are great improvements in knowledge mastery, classroom observation and learning motivation compared with before intervention. Slightly exceeding the overall class level in knowledge mastery, the learning enthusiasm of learners has been greatly improved, and the increase in self-efficacy has also made the learning motivation exceed the overall level from being lower than the overall level at the beginning. The classroom observation has changed from being serious and focused in class before intervention but not actively participating in interactions to actively participating in class, actively answering and discussing.

[0144] The individual thinking abilities before and after the intervention include problem-solving, thinking innovation, and critical decision-making. The thinking abilities before and after the intervention are compared and analyzed with the overall level after the intervention, and the results are as Figure 8 shown. Before the intervention, there was a significant difference between this learner's critical decision-making and the overall level, and there was a slight deficiency in thinking innovation. After the intervention, the individual exceeded the overall level in problem-solving, thinking innovation, and critical decision-making, showing great progress compared to before the intervention. It can be seen that after multiple interventions, the final thinking ability of this learner has been greatly improved.

[0145] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those of ordinary skill in the art can understand that the above-mentioned devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code is provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and their modules of the present invention can be implemented by hardware circuits of programmable hardware devices such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips and transistors, or field programmable gate arrays and programmable logic devices, can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above hardware circuits and software, such as firmware.

[0146] The above is only the specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention by those skilled in the art within the technical scope disclosed by the present invention shall be covered by the protection scope of the present invention.

Claims

1. A learning intervention method in a blended learning environment, characterized in that, The following steps are involved: (1) Collecting data from online learning platforms, student affairs systems, teaching systems, student competition systems, library systems, seat reservation systems, and questionnaires, and comprehensively integrating offline and online data, making the learner portrait in the hybrid learning environment richer than the learner portrait based on the online learning platform; (2) cleaning, standardizing and formatting the data and storing them in a database; (3) applying learner portraits to a hybrid learning environment, breaking through the traditional classroom and online learning fields, and constructing a multi-dimensional learner portrait model based on the processed data, wherein the model includes at least six dimensions: basic information, learning preparation, learning style, learning behavior, learning outcomes, and thinking ability; (4) in a hybrid learning environment, building a learning intervention framework based on the multidimensional learner portrait model, wherein the framework includes overall intervention, group intervention, and individual intervention levels; (5) Implementing intervention through the learning intervention framework, re-collecting and processing the new data generated by online and offline teaching sessions, and updating the portrait model; The intervention framework based on learner profiles in a blended learning environment specifically includes: Provide the most appropriate learning intervention for each student based on the learner profile. The intervention framework is as follows: Conduct learning situation analysis based on learner portraits and diagnose learning problems; implement interventions in the entire process of learning preparation, autonomous learning, consolidation and internalization, application transfer, and evaluation feedback in blended learning based on the principles of student subjectivity, accuracy, embeddedness, and effectiveness; The overall intervention strategy designed in combination with learner portraits includes learning supervision and early warning, technical and resource guidance, and mind mapping; group intervention strategies include goal positioning and value guidance, teacher expectation transmission, task design and guidance, activity strategies, resource recommendation, and evaluation strategies; individual intervention strategies include individual reminder strategies, portrait feedback strategies, and self-reflection and adjustment strategies; the effectiveness of intervention practice is evaluated by combining changes in one's own portraits, differences between individual and group portraits, changes in the overall portrait before and after, and comparison of the overall portraits of each class; satisfaction questionnaires are distributed to students, and teacher interviews are conducted to understand the emotional experience of teachers and students towards intervention practices based on learner portraits.

2. The learning intervention method in a blended learning environment according to claim 1, wherein The data includes the following four types: (a) Identity and Survey Data; (b) learning behavior data; (c) learning outcome data; (d) Thinking ability data.

3. The learning intervention method in a blended learning environment according to claim 2, wherein The thinking ability data includes information representing problem solving, innovative thinking and critical decision-making, and the information comes from students' online answer records, homework scores, classroom behavior records or peer evaluations.

4. The learning intervention method in a blended learning environment according to claim 1, characterized in that, When building a learning intervention framework, a number of intervention strategy plans at the group level are formed by clustering the learning preparation, learning behavior and learning outcome data in the portrait model.

5. The learning intervention method in a blended learning environment according to claim 1, characterized in that, When implementing intervention, by comparing the personal portrait data with the group portrait or the preset threshold, when the difference exceeds the preset range, a prompt message is triggered to the learner or teacher.

6. The learning intervention method in a blended learning environment according to claim 1, wherein The steps for updating the portrait model include: re-collecting and processing the newly generated learning behavior data and learning result data online and offline, correcting or recalculating the various dimensional labels of the model, and replacing the old version of the portrait model in the database.

7. A learning intervention system in a blended learning environment for implementing the learning intervention method in the blended learning environment as described in any one of claims 1 to 6, characterized in that, including: a portrait model construction module for constructing a portrait model; an intervention framework building module for building an intervention framework based on the learner portrait in a blended learning environment; a learning intervention module for implementing learning interventions.

8. A computer device, comprising a memory and a processor, the memory storing a computer program, and when the computer program is executed by the processor, the processor executes the steps of the learning intervention method in a blended learning environment according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, and when the computer program is executed by the processor, the processor executes the steps of the learning intervention method in a blended learning environment according to any one of claims 1 to 6.

10. An information data processing terminal, the information data processing terminal including the learning intervention system in a blended learning environment according to claim 7.

Images