Whole-cycle intelligent education evaluation system based on knowledge graph

Abstract

The invention provides a full-cycle intelligent education evaluation system based on a knowledge graph, and relates to the technical field of data processing, and the system is used for carrying out the correlation analysis of a learning behavior data set and a knowledge point data set of a student, calculating the mastering degree of the student on each knowledge point, recognizing the relation between the student and each knowledge point, and carrying out the calculation of the mastering degree. The method comprises the following steps: acquiring knowledge points of students, mapping the knowledge points into a graph structure to obtain a staged knowledge graph, updating the staged knowledge graph, adjusting the relationship between the students and each knowledge point to obtain a dynamic knowledge graph, extracting the real-time mastery degree of each knowledge point according to the dynamic knowledge graph, and periodically analyzing the knowledge points in combination with time dimensions. And obtaining periodic evaluation data, analyzing the mastering degree, the progress trend and the weak knowledge points of the students at each knowledge point, and generating an individual evaluation data set. According to the invention, the knowledge graph can be constructed to periodically evaluate the learning condition of the student.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a full-cycle intelligent education evaluation system based on knowledge graph. Background Art

[0002] Existing education assessment systems usually rely on traditional methods such as test scores, homework scores and teachers' subjective assessments to evaluate students. These assessment methods are often static and lack continuous tracking and in-depth analysis of students' learning process. They may not accurately reflect students' knowledge mastery and learning potential. In some education assessment systems, although students' learning data can be collected, it may not be possible to effectively integrate these data, resulting in the inability to achieve accurate and personalized education assessment.

[0003] For example, some education assessment systems generate assessment reports regularly, but the assessment cycle is long and often cannot provide real-time feedback on students’ learning status, causing delays in important feedback information in the learning process. This information delay may affect teachers and students’ timely adjustment of their learning plans, resulting in reduced learning outcomes. Summary of the invention

[0004] The purpose of the present invention is to provide a full-cycle intelligent education evaluation system based on knowledge graph, aiming to solve the problems mentioned in the background technology.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] A full-cycle intelligent education evaluation system based on knowledge graph, the system comprising:

[0007] The association module is used to perform association analysis on the student's learning behavior data set and the knowledge point data set, calculate the student's mastery of each knowledge point, and obtain the learning progress data set;

[0008] The first graph module is used to identify the relationship between students and each knowledge point according to the learning progress data set, obtain knowledge relationship data, and map it into a graph structure to obtain a staged knowledge graph;

[0009] The second graph module is used to obtain the students' newly added learning data set, and update the phased knowledge graph based on it, adjust the relationship between the students and each knowledge point, and obtain a dynamic knowledge graph;

[0010] The cycle analysis module is used to extract the real-time mastery of each knowledge point based on the dynamic knowledge graph, obtain the real-time learning progress data set, and conduct periodic analysis on it in combination with the time dimension to obtain periodic evaluation data;

[0011] The evaluation module is used to analyze students’ mastery of each knowledge point, progress trends, and weak knowledge points based on periodic evaluation data, and generate individual evaluation data sets;

[0012] The feedback module is used to send individual evaluation data sets to the student and teacher terminals synchronously to adjust students' learning behaviors.

[0013] Furthermore, the association module includes:

[0014] A learning behavior data acquisition unit is used to acquire the learning time, homework scores and test scores of students in different learning cycles to obtain a learning behavior data set;

[0015] The association analysis unit is used to extract the homework scores and test scores of students related to different knowledge points in different learning cycles according to the learning behavior data set, and calculate the average homework scores and average test scores in different learning cycles to obtain the knowledge point related data set;

[0016] The mastery degree calculation unit is used to integrate the learning time, average homework scores and average test scores related to different knowledge points in different learning cycles of students according to the knowledge point related data sets, calculate the mastery degree of students on different knowledge points, and obtain the learning progress data set.

[0017] Furthermore, the first graph module includes:

[0018] The knowledge point learning data set unit is used to extract the student's access records to different knowledge points, the number of learning tasks and the accumulated learning time in the learning cycle according to the learning progress data set to obtain the knowledge point learning data set;

[0019] A coverage index data set unit is used to calculate the ratio between the number of knowledge points that each student has learned in different courses and the total number of knowledge points in the course according to the knowledge point learning data set, so as to obtain a coverage index data set;

[0020] The completion depth indicator data set unit is used to count the number of learning tasks and cumulative learning time of each student in different courses according to the knowledge point learning data set to obtain the completion depth indicator data set;

[0021] The learning relationship analysis unit is used to determine the learning relationship between students and knowledge points in different courses based on the coverage indicator data set and the completion depth indicator data set to obtain a phased knowledge graph.

[0022] Furthermore, the learning relationship analysis unit includes:

[0023] The node data unit is used to define the nodes of the graph structure according to the learning relationship data, define the student identifier as a student node, define the knowledge point as a knowledge point node, and obtain the node data;

[0024] The edge data unit is used to construct the edges of the graph structure according to the learning relationship data, integrate the coverage index with the completion depth index, calculate the edge weight values ​​between students and different knowledge points, and obtain edge data;

[0025] The graph construction unit is used to connect student nodes with different knowledge point nodes according to node data and edge data, and combine them to form an undirected graph structure to obtain a phased knowledge graph.

[0026] Furthermore, the second graph module includes:

[0027] A new learning data set unit is added to obtain the learning time, homework scores and test scores in the current learning cycle, and merge them according to knowledge points to obtain a new learning data set;

[0028] The standard data set updating unit is used to update the coverage index and completion depth index between the learning and different knowledge points according to the newly added learning data set, so as to obtain the updated standard data set;

[0029] The learning relationship change data unit is used to calculate the difference between the updated standard data set and the staged knowledge graph to obtain the learning relationship change data;

[0030] The edge increment data set unit is used to adjust the edge weight value in the staged knowledge graph according to the learning relationship change data, update the connection strength between students and different knowledge points, and obtain the edge increment data set;

[0031] The graph update unit is used to construct new graph structure nodes and edges based on the edge incremental data set, and to replace the structure of the phased knowledge graph to obtain a dynamic knowledge graph.

[0032] Furthermore, the graph updating unit includes:

[0033] The edge data updating unit is used to analyze the node data and edge data of the phased knowledge graph according to the edge incremental data set, identify the edge weight value that needs to be updated, and obtain the updated edge data;

[0034] A new node data unit is added, which is used to identify the relationship between the newly added students and knowledge points according to the edge incremental data set, obtain the newly added relationship data, and construct new graph structure nodes and edges according to it to obtain the newly added node data;

[0035] The dynamic knowledge graph unit is used to add new nodes and edges according to the newly added node data and updated edge data, and update the original edge weight values ​​to generate a new graph structure and obtain a dynamic knowledge graph.

[0036] Furthermore, the cycle analysis module includes:

[0037] The real-time mastery degree data unit is used to extract the latest learning time, homework scores and test scores of different knowledge points according to the dynamic knowledge graph, determine the real-time mastery degree of students in different knowledge points, and obtain real-time mastery degree data;

[0038] The mastery change trend data unit is used to compare the mastery degree at the real-time time point with the mastery degree at the previous time point according to the real-time mastery degree data, calculate the mastery change trend of different knowledge points, and obtain the mastery change trend data;

[0039] The real-time learning progress data set unit is used to determine the student's learning progress in the current cycle by merging the real-time mastery degree data and the mastery change trend data to obtain the real-time learning progress data set.

[0040] Furthermore, the cycle analysis module also includes:

[0041] The periodic analysis unit is used to perform periodic analysis on the learning progress of different knowledge points in the time dimension according to the real-time learning progress data set, and to accumulate and compare the learning progress data of multiple time nodes to obtain periodic analysis results;

[0042] The periodic evaluation unit is used to calculate the change rate of the mastery degree and the change rate of the learning progress of different knowledge points according to the periodic analysis results to obtain the periodic evaluation data.

[0043] Furthermore, the evaluation module includes:

[0044] The student ability unit is used to extract the students’ mastery of different knowledge points based on the periodic assessment data and obtain the students’ ability data;

[0045] The progress trend unit is used to compare the students' mastery levels at different time points based on their ability data, calculate the learning progress rate of different knowledge points, and obtain the progress trend data;

[0046] The weak knowledge point unit is used to calculate the deviation of students' mastery level on different knowledge points based on the progress trend data, identify students' weak knowledge points, and obtain weak knowledge point data.

[0047] Furthermore, the weak knowledge point unit includes:

[0048] The mastery degree change data unit is used to extract the change rate of students' mastery degree of different knowledge points at different time points according to the progress trend data, and obtain the mastery degree change data;

[0049] The mastery degree deviation data unit is used to calculate the historical average mastery change rate of different knowledge points according to the mastery degree change data, and calculate the mastery degree deviation between the student's real-time mastery degree change rate on different knowledge points and the historical average mastery change rate according to the mastery degree deviation data to obtain the mastery degree deviation data;

[0050] The weak knowledge point judgment unit is used to identify the students' weak knowledge points based on the mastery degree deviation data. When the mastery degree deviation of a certain knowledge point exceeds a preset deviation threshold, the knowledge point is the student's weak knowledge point, and a weak knowledge point data set is obtained.

[0051] The above solution of the present invention includes at least the following beneficial effects:

[0052] The present invention calculates the students' mastery of each knowledge point by performing correlation analysis on the students' learning behavior data set and the knowledge point data set, thereby achieving quantitative evaluation of the students' knowledge mastery and providing accurate data support for subsequent educational evaluation. By correlating the learning time, homework grades and test scores with the knowledge points, the present invention can comprehensively capture the students' performance in the learning process and reveal the differences in students' mastery of different knowledge points, helping the education system to identify students' weak links in certain knowledge points, thereby providing a basis for personalized learning, avoiding the one-sidedness of traditional evaluation that relies solely on exams and homework, and providing more accurate and comprehensive evaluation results.

[0053] The present invention uses a learning progress data set to identify the relationship between students and each knowledge point, and maps these relationships into a graph structure to generate a phased knowledge graph. Through the construction of the graph structure, the dynamic relationship between students and knowledge points can be clearly displayed, which not only reflects the students' current learning status, but also reveals the correlation between knowledge points. Through this visual method, teachers can intuitively understand the students' learning progress at different stages and identify blind spots or loopholes in students' knowledge mastery. The phased knowledge graph provides a scientific basis for subsequent dynamic adjustment and feedback.

[0054] The present invention acquires students' newly added learning data sets, updates periodic knowledge graphs, and constructs dynamic knowledge graphs. The real-time updating capability enables the system to reflect immediate changes in students' learning process. Unlike traditional static assessment systems, dynamic knowledge graphs can track students' changing mastery of knowledge during the learning process, thereby providing more accurate and timely assessment results. This dynamism is critical for guiding students' personalized learning, and can adjust teaching plans in a timely manner to ensure that necessary help and support can be provided at the right time.

[0055] The present invention performs periodic analysis through dynamic knowledge graphs, extracts the real-time mastery of each knowledge point and generates periodic evaluation data. Periodic analysis can combine students' learning progress with the time dimension for dynamic tracking. By comparing and analyzing students' learning progress at different time points, the periodic analysis module can reveal the changing trend of students' mastery. This analysis can not only understand students' growth trajectory in the long-term learning process, but also provide a basis for the formulation of personalized learning plans, ensuring timely feedback and adjustments during the learning process.

[0056] The present invention analyzes students' mastery of each knowledge point, progress trends and weak knowledge points through periodic evaluation data, and generates individual evaluation data sets. It can analyze students' learning progress from multiple dimensions, including mastery, progress speed and weak links, thereby achieving a comprehensive evaluation of students. This evaluation can not only accurately identify students' strengths and weaknesses in different knowledge points, but also discover students' weak areas, and further guide the focus of learning. Through the generation of evaluation data, positive feedback can be formed in the learning process to improve learning effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a flowchart of a full-cycle intelligent education evaluation system based on knowledge graph provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0059] like Figure 1 As shown, an embodiment of the present invention proposes a full-cycle intelligent education evaluation system based on a knowledge graph, and the system includes:

[0060] The association module is used to perform association analysis on the student's learning behavior data set and the knowledge point data set, calculate the student's mastery of each knowledge point, and obtain the learning progress data set;

[0061] The first graph module is used to identify the relationship between students and each knowledge point according to the learning progress data set, obtain knowledge relationship data, and map it into a graph structure to obtain a staged knowledge graph;

[0062] The second graph module is used to obtain the students' newly added learning data set, and update the phased knowledge graph based on it, adjust the relationship between the students and each knowledge point, and obtain a dynamic knowledge graph;

[0063] The cycle analysis module is used to extract the real-time mastery of each knowledge point based on the dynamic knowledge graph, obtain the real-time learning progress data set, and conduct periodic analysis on it in combination with the time dimension to obtain periodic evaluation data;

[0064] The evaluation module is used to analyze students’ mastery of each knowledge point, progress trends, and weak knowledge points based on periodic evaluation data, and generate individual evaluation data sets;

[0065] The feedback module is used to send individual evaluation data sets to the student and teacher terminals synchronously to adjust students' learning behaviors.

[0066] In an embodiment of the present invention, an association module is used to perform association analysis on a student's learning behavior data set and a knowledge point data set, calculate the student's mastery of each knowledge point, and obtain a learning progress data set. By quantifying the student's mastery of each knowledge point, an accurate assessment of the learning progress is achieved; a first graph module is used to identify the relationship between the student and each knowledge point based on the learning progress data set, obtain knowledge relationship data, and map it into a graph structure to obtain a phased knowledge graph. Through the construction of the graph structure, the first graph module provides a visual presentation method for the student's learning process; a second graph module is used to obtain a new learning data set for the student, and based on it, the graph updates the phased knowledge graph, adjusts the relationship between the student and each knowledge point, and obtains a dynamic knowledge graph. By continuously updating the student's learning data, the system can adjust the relationship between the student and the knowledge point in real time. , making educational evaluation more timely and accurate; the periodic analysis module is used to extract the real-time mastery of each knowledge point according to the dynamic knowledge graph, obtain the real-time learning progress data set, and conduct periodic analysis on it in combination with the time dimension to obtain periodic evaluation data. Through periodic analysis, the system can identify the progress of students in the long-term learning process; the evaluation module is used to analyze the students' mastery of each knowledge point, progress trend and weak knowledge points according to the periodic evaluation data, and generate individual evaluation data sets. Through comprehensive analysis of the students' mastery of each knowledge point, progress trend and weak links, the system can provide teachers with specific teaching directions; the feedback module is used to synchronously send the individual evaluation data set to the student end and the teacher end, adjust the students' learning behavior, and realize real-time, two-way information flow, helping students and teachers to obtain learning data in a timely manner and respond quickly to problems in learning.

[0067] The feedback module is used to synchronously send individual evaluation data sets to the student and teacher terminals to adjust students' learning behaviors, including:

[0068] The individual assessment data set is generated based on the system's periodic assessment data and the students' learning behavior data. It includes each student's mastery of each knowledge point, progress trends, and identified weak knowledge points. The system will analyze the output data of the assessment module and integrate various indicators such as study time, homework grades, test scores, mastery level, etc. to generate an individual assessment data that comprehensively reflects the student's learning situation. These data include not only the students' current mastery level, but also the deficiencies and weak links in the learning process.

[0069] Next, the system will format these individual assessment data to make them suitable for the display needs of the student and teacher sides. The formatting process includes converting the data into an easy-to-understand report format, which may include charts, trend curves, percentages, and specific suggestions. In order to ensure that the data can be accurately transmitted to students and teachers, the system will simplify and detail the data to varying degrees based on the needs of different users, such as students or teachers. For example, for students, feedback information may be more intuitive and easy to understand, highlighting students' weaknesses and recommended learning methods; for teachers, feedback content includes more detailed statistical data and possible teaching improvement suggestions to help teachers understand the learning situation of student groups and individuals.

[0070] After the data is formatted, the feedback module will synchronously send the individual assessment data set to the student and teacher ends via the Internet. In order to ensure the real-time and accuracy of data transmission, the system adopts a stable data transmission protocol and encryption method to ensure the security of the data. After the transmission is completed, students can view their personal learning assessment reports through the student-side application, and teachers can obtain the learning progress of student groups or individuals through the teacher-side interface. This two-way synchronous feedback mechanism ensures that information can be delivered to students and teachers in a timely manner, making it easier for both parties to adjust their learning and teaching strategies based on feedback.

[0071] After receiving feedback, students can adjust their learning behaviors according to the suggestions in the feedback. For example, students can do additional learning or review on weak knowledge points, or evaluate whether their learning methods are effective based on progress trend data. This self-adjustment and improvement can promote the improvement of students' independent learning ability, thereby making the learning process more efficient.

[0072] For teachers, the feedback information received provides strong support for their teaching adjustments. Teachers can understand students' individual needs based on feedback data, especially for those students with low mastery. Teachers can provide additional tutoring or adjust teaching content in a timely manner to ensure that students do not fall behind the course progress. At the same time, teachers can adjust teaching methods and learning activities based on students' progress trend information, so as to provide more flexible and effective support for students with different learning progress.

[0073] In a preferred embodiment of the present invention, the association module includes:

[0074] A learning behavior data acquisition unit is used to acquire the learning time, homework scores and test scores of students in different learning cycles to obtain a learning behavior data set;

[0075] The association analysis unit is used to extract the homework scores and test scores of students related to different knowledge points in different learning cycles according to the learning behavior data set, and calculate the average homework scores and average test scores in different learning cycles to obtain the knowledge point related data set;

[0076] The mastery degree calculation unit is used to integrate the learning time, average homework scores and average test scores related to different knowledge points in different learning cycles of students according to the knowledge point related data sets, calculate the mastery degree of students on different knowledge points, and obtain the learning progress data set.

[0077] In an embodiment of the present invention, a learning behavior data acquisition unit is used to acquire the learning time, homework scores and test scores of students in different learning cycles to obtain a learning behavior data set, thereby realizing multi-channel and multi-source data collection and ensuring the timeliness and accuracy of the data; an association analysis unit is used to extract the homework scores and test scores of students related to different knowledge points in different learning cycles according to the learning behavior data set, and calculate the average homework scores and average test scores in different learning cycles to obtain a knowledge point related data set, which can provide a more accurate learning performance evaluation through data integration and association, thereby avoiding the drawback of traditional methods that rely only on a single score; a mastery degree calculation unit is used to integrate the learning time, average homework scores and average test scores of students related to different knowledge points in different learning cycles according to the knowledge point related data set, calculate the mastery degree of students on different knowledge points, and obtain a learning progress data set, which can quantify the mastery degree of students on each knowledge point, thereby providing a scientific basis for educational decision-making.

[0078] The calculation formula of the mastery degree is: ,

[0079] in, For students in The degree of mastery of each knowledge point, is the index of the knowledge point, is the total number of learning cycles, is the index of the learning cycle, For the The duration of a learning cycle, For the The average homework score for each learning cycle, For the The average test scores of the study period, is the coefficient, is a constant.

[0080] in, It reflects the effect of learning time on mastery, and uses a logarithmic function to avoid over-amplifying the effect of learning time on mastery. This part reflects the effect of the number of learning tasks on the mastery level. This part uses a square form, indicating that the effect of the number of tasks on the mastery level is not linear. The more tasks there are, the greater the effect. In particular, when the number of tasks increases, the effect will be more significant. It is introduced by means of a sine function, indicating that changes in grades will have a nonlinear effect on mastery. The form of this function ensures that lower grades have a smaller impact, while higher grades will significantly increase mastery. It is the learning duration of each learning cycle, which means that different knowledge points require different learning times to master. When a knowledge point is difficult to master, the improvement in the mastery level will be inhibited, because learning the knowledge point requires more learning input to effectively master it.

[0081] in, It is the coefficient that controls the effect of study time on students' mastery. This coefficient determines the degree of influence of study time on mastery. In particular, in the logarithmic function, it determines the marginal effect of study time on mastery. For example, in the case of self-study courses, students' study time usually does not directly linearly affect their mastery. The effect of study time on mastery is gradual. Therefore, a smaller coefficient is used. The value is 0.1; in a shorter learning cycle, especially for task-driven learning activities, if the course content is dense and the time is short, the increase in learning time will significantly improve students' mastery, so the coefficient is larger. The value is 0.5; in an intensive training environment, such as a concentrated skill training course, in this environment, the length of study significantly affects the degree of mastery, so a larger coefficient is needed. The value is 1.

[0082] in, is the coefficient that controls the effect of the number of learning tasks on students' mastery. Since the square of the number of tasks is used for adjustment, its value determines the nonlinear effect of the number of tasks on mastery. For example, in most regular learning scenarios, the number of tasks is moderate and the workload is not extremely increased, which makes the effect of the number of tasks on students' mastery relatively mild. The value is 0.2; in a high-intensity learning environment, students need to complete a large amount of homework or exercises, such as in programming learning, language learning and other courses. The number of tasks is crucial to improving students' mastery. The value is 0.5; courses with a small number of tasks, such as lecture-style courses or content-intensive courses, in which the number of tasks has a smaller impact on mastery. The value is 0.05.

[0083] in, It is the coefficient that controls the overall increase of the numerator. This coefficient controls the amplification effect of all influencing factors, such as study time, number of tasks, and test scores. For example, in the standard learning mode, study time, number of tasks, and test scores contribute to the students' mastery in proportion. At this time, the contribution of each factor of the students is relatively balanced. The value is 1; in scenarios where students' efforts and learning outcomes need to be improved, especially in concentrated learning courses, when students' efforts and progress need to be significantly reflected, it will be set to a higher value. A value of 2: a more relaxed learning environment where students' learning progress will not be overly magnified by each factor. The value is 0.5.

[0084] in, is a minimum constant used to avoid the denominator being zero.

[0085] In a preferred embodiment of the present invention, the first atlas module includes:

[0086] The knowledge point learning data set unit is used to extract the student's access records to different knowledge points, the number of learning tasks and the accumulated learning time in the learning cycle according to the learning progress data set to obtain the knowledge point learning data set;

[0087] A coverage index data set unit is used to calculate the ratio between the number of knowledge points that each student has learned in different courses and the total number of knowledge points in the course according to the knowledge point learning data set, so as to obtain a coverage index data set;

[0088] The completion depth indicator data set unit is used to count the number of learning tasks and cumulative learning time of each student in different courses according to the knowledge point learning data set to obtain the completion depth indicator data set;

[0089] The learning relationship analysis unit is used to determine the learning relationship between students and knowledge points in different courses based on the coverage indicator data set and the completion depth indicator data set to obtain a phased knowledge graph.

[0090] In an embodiment of the present invention, a knowledge point learning data set unit is used to extract the student's access records, the number of learning tasks and the cumulative learning time for different knowledge points in the learning cycle according to the learning progress data set, and obtain the knowledge point learning data set, which ensures the detailed record of each student's specific behavior in the learning process, and can accurately track the degree of participation and investment time of each student in the knowledge point learning process; the coverage degree indicator data set unit is used to calculate the ratio between the number of knowledge points learned by each student in different courses and the total number of knowledge points in the course according to the knowledge point learning data set, and obtain the coverage degree indicator data set, which can intuitively reflect the learning breadth of students in different courses; the completion depth indicator data set unit is used to count the number of learning tasks and the cumulative learning time of each student in different courses according to the knowledge point learning data set, and obtain the completion depth indicator data set, which can show the students' actual mastery and depth of knowledge points; the learning relationship analysis unit is used to determine the learning relationship between students and knowledge points in different courses according to the coverage degree indicator data set and the completion depth indicator data set, and obtain a phased knowledge graph, which can accurately depict the students' mastery of each knowledge point and provide important relationship data for subsequent graph construction.

[0091] In a preferred embodiment of the present invention, the learning relationship analysis unit includes:

[0092] The node data unit is used to define the nodes of the graph structure according to the learning relationship data, define the student identifier as a student node, define the knowledge point as a knowledge point node, and obtain the node data;

[0093] The edge data unit is used to construct the edges of the graph structure according to the learning relationship data, integrate the coverage index with the completion depth index, calculate the edge weight values ​​between students and different knowledge points, and obtain edge data;

[0094] The graph construction unit is used to connect student nodes with different knowledge point nodes according to node data and edge data, and combine them to form an undirected graph structure to obtain a phased knowledge graph.

[0095] In an embodiment of the present invention, a node data unit is used to define nodes of a graph structure according to learning relationship data, define a student identifier as a student node, and define a knowledge point as a knowledge point node to obtain node data, thereby providing high-quality, structured node data for subsequent graph construction of the system and ensuring the accuracy and completeness of the knowledge graph; an edge data unit is used to construct edges of a graph structure according to learning relationship data, fuse coverage indicators with completion depth indicators, calculate edge weight values ​​between students and different knowledge points, and obtain edge data, which can express complex learning relationships in the form of a graph, thereby providing support for subsequent dynamic adjustments and personalized learning recommendations; a graph construction unit is used to connect student nodes with different knowledge point nodes according to node data and edge data, and combine them to form an undirected graph structure to obtain a phased knowledge graph, which can clearly display the learning association between students and each knowledge point, and can be dynamically adjusted as students' learning behaviors change.

[0096] The edge weight value is calculated as follows: ,

[0097] in, For students and The edge weights between knowledge points are is the index of the knowledge point, For students in The coverage index of each knowledge point is For students in The cumulative learning time on each knowledge point is For students in The number of learning tasks completed on each knowledge point, is the coefficient.

[0098] in, The degree of students' coverage of knowledge points will be nonlinearly adjusted by the length of study time. As the length of study time increases, the edge weight value will show a certain increase effect. This relationship reflects the important influence of learning depth on the degree of mastery. It combines the number of tasks and study time of students on a certain knowledge point to form a comprehensive evaluation of the knowledge point. The combination of the number of learning tasks and the study time can better reflect the students' mastery of knowledge in some cases, especially the differences in the time and energy invested by students. The nonlinear relationship between the number of learning tasks and the coverage of knowledge points is taken into account. Through the exponential decay function, the influence of the number of learning tasks and the coverage of knowledge points will gradually decrease to avoid excessively increasing the impact of the number of tasks and the coverage of knowledge points on the edge weights, thereby maintaining the rationality of the edge weight value.

[0099] in, is the weight coefficient, which determines the contribution ratio of each item in the formula to the final edge weight value. Normally, the sum of the three weight coefficients should be 1 to ensure that the contribution of all items to the final result is balanced and reasonable. The system makes a reasonable division of the impact of different factors on students' mastery of knowledge points. The depth of learning reflects the degree of students' investment in knowledge points, which is usually determined by factors such as learning time and task completion. The greater the depth, the deeper the students' mastery of the knowledge points. The influence of learning depth on the mastery of knowledge points should be very important, and usually accounts for a large proportion. The value is 0.5; The number of learning tasks reflects the efforts and practices that students have made to master the knowledge points. The number of tasks completed affects students' mastery of knowledge points to a certain extent, but compared with the depth of learning, the number of tasks may not have as direct an impact on students' mastery as the depth of learning. The value is 0.3; It takes into account the nonlinear relationship between the number of learning tasks and the coverage of knowledge points. As the number of tasks students learn gradually increases, this term adjusts the complex relationship between the number of tasks and the coverage. Although this term is nonlinear, its impact on the final evaluation is usually small. The value is 0.2.

[0100] in, To control the impact of learning depth on students’ mastery of knowledge points, since learning depth is an important factor affecting students’ mastery, its value is set to 1.5. A larger value means that learning depth has a stronger impact on edge weight values; It controls the effect of the square term of learning depth on the number of tasks. Its value is set to 0.8, indicating that the effect of the square of learning depth on task completion is nonlinear. A smaller adjustment coefficient will make the effect of this term milder. In order to control the impact of learning depth on knowledge point coverage, its value is set to 0.6, which means that the impact of learning depth in controlling knowledge point coverage is relatively moderate and will not over-amplify its effect; In order to control the impact of the number of learning tasks on the edge weight, since the number of tasks usually increases as students progress in their learning, its value is set to 0.4, which can effectively regulate the impact of the number of tasks on the edge weight so that its effect is not too prominent.

[0101] Among them, the graph construction unit is used to connect the student nodes with different knowledge point nodes according to the node data and edge data, and combine them to form an undirected graph structure to obtain a phased knowledge graph, which specifically includes:

[0102] First, we need to extract the basic elements used to construct the graph structure from the node data and edge data. The node data includes student nodes and knowledge point nodes. The student node represents each student, and the knowledge point node represents each knowledge point related to the student's learning. Each node has specific attributes associated with it. For example, the student node may contain information such as the student ID and learning status, while the knowledge point node contains information such as the knowledge point ID and related learning tasks. The node data unit will first create these nodes based on the association data between students and knowledge points.

[0103] Next, the edge data unit is responsible for establishing a connection between each pair of student nodes and knowledge point nodes. The edge represents the learning relationship between students and knowledge points. Specifically, the weight of each edge represents the student's mastery of a certain knowledge point. The weight value of this edge is calculated using the previous formula. It combines multiple factors such as the number of learning tasks, study time, homework grades, and test scores of students on a certain knowledge point. Based on these calculation results, the edge data unit assigns a weight value to the edge between each student node and the knowledge point node. The size of the weight value reflects the student's mastery of the knowledge point. The larger the weight value, the higher the mastery, and vice versa.

[0104] After obtaining the node data and edge data, the graph construction unit is responsible for organizing these nodes and edges to form a graph structure. The specific implementation method may include using graph theory algorithms, such as depth-first search or other graph construction techniques. According to the relationship between student nodes and knowledge point nodes, all student nodes are connected to the corresponding knowledge point nodes through edges to ensure that each edge can correctly connect the corresponding nodes.

[0105] The type of graph is an undirected graph, that is, the edges have no direction. An undirected graph means that the edges have no starting point and end point, but are bidirectional, indicating that there is an equal relationship between students and knowledge points. For example, if a student has learned a certain knowledge point, then the knowledge point and the student will be connected by an edge, indicating that the student has mastered the knowledge point, and this relationship is bidirectional. The advantage of the undirected graph structure in this scenario is that it can conduct a comprehensive analysis of the interaction between students and knowledge points, rather than simply relying on information in a certain direction.

[0106] During the construction process, the graph construction unit also needs to process the hierarchical relationship between nodes and the topological structure of the graph. For example, the edges between student nodes and knowledge point nodes can be sorted by weight value, so as to adjust the structure of the graph according to the students' mastery level, so that the students' mastery of each knowledge point can be more intuitively evaluated in subsequent analysis. In this way, the staged knowledge graph can clearly present the relationship between students and each knowledge point, and reflect the students' learning progress on different knowledge points. In addition, in order to support subsequent dynamic updates and graph adjustments, a unique identifier is usually assigned to each node and edge to ensure the scalability and flexibility of the graph structure.

[0107] In a preferred embodiment of the present invention, the second atlas module includes:

[0108] A new learning data set unit is added to obtain the learning time, homework scores and test scores in the current learning cycle, and merge them according to knowledge points to obtain a new learning data set;

[0109] The standard data set updating unit is used to update the coverage index and completion depth index between the learning and different knowledge points according to the newly added learning data set, so as to obtain the updated standard data set;

[0110] The learning relationship change data unit is used to calculate the difference between the updated standard data set and the staged knowledge graph to obtain the learning relationship change data;

[0111] The edge increment data set unit is used to adjust the edge weight value in the staged knowledge graph according to the learning relationship change data, update the connection strength between students and different knowledge points, and obtain the edge increment data set;

[0112] The graph update unit is used to construct new graph structure nodes and edges based on the edge incremental data set, and to replace the structure of the phased knowledge graph to obtain a dynamic knowledge graph.

[0113] In an embodiment of the present invention, a new learning data set unit is added to obtain the learning time, homework scores and test scores in the current learning cycle, and merge them according to knowledge points to obtain a new learning data set, which provides the latest learning information for graph updating; an updated standard data set unit is used to update the coverage index and completion depth index between learning and different knowledge points according to the new learning data set to obtain an updated standard data set. By continuously updating the standard data set, it is ensured that the relationship between each student and each knowledge point in the graph can be consistent with the student's actual learning progress; a learning relationship change data unit is used to calculate its relationship with the staged knowledge graph according to the updated standard data set. The difference between the edge graph and the knowledge graph is used to obtain the learning relationship change data, which can dynamically capture the changes in the students' learning process; the edge incremental data set unit is used to adjust the edge weight value in the stage-by-stage knowledge graph according to the learning relationship change data, update the connection strength between students and different knowledge points, and obtain the edge incremental data set, which can accurately adjust the connection strength between students and knowledge points to ensure the dynamic nature of the knowledge graph; the graph update unit is used to construct new graph structure nodes and edges according to the edge incremental data set, and replace the structure of the stage-by-stage knowledge graph to obtain a dynamic knowledge graph, which can update the entire knowledge graph structure after each learning cycle so that it can accurately reflect the students' latest learning progress.

[0114] Among them, the edge incremental data set unit is used to adjust the edge weight value in the staged knowledge graph according to the learning relationship change data, update the connection strength between students and different knowledge points, and obtain the edge incremental data set, which specifically includes:

[0115] First, the system needs to obtain learning relationship change data, which records the changes in the relationship between students and different knowledge points in the current learning cycle. These changes usually come from fluctuations in students' grades in homework and exams, increases or decreases in study time, and changes in other indicators related to knowledge point mastery. By comparing the data of the previous cycle with the current cycle, the system can accurately identify the student's progress or regression in each knowledge point.

[0116] Next, the edge increment dataset unit calculates the edge increment of each knowledge point by comparing the students' learning data in different periods. The increment reflects the students' learning progress on each knowledge point, which is usually achieved by adjusting the edge weights of the connections between students and knowledge points. Specifically, the edge increment is based on multi-dimensional factors such as the students' learning time, homework grades, and test scores. The changes in the strength of the connection between each student and each knowledge point are calculated. According to these changes, the system can gradually adjust the weight value of each edge in the graph to accurately describe the students' mastery of each knowledge point. Through this process, the edge increment dataset unit not only enables the knowledge graph to reflect the dynamic changes in students' learning status, but also ensures that the graph can adapt to the gradually changing degree of mastery during the students' learning process, thereby improving the system's ability to reflect students' personalized learning process.

[0117] In a preferred embodiment of the present invention, the atlas updating unit comprises:

[0118] The edge data updating unit is used to analyze the node data and edge data of the phased knowledge graph according to the edge incremental data set, identify the edge weight value that needs to be updated, and obtain the updated edge data;

[0119] A new node data unit is added, which is used to identify the relationship between the newly added students and knowledge points according to the edge incremental data set, obtain the newly added relationship data, and construct new graph structure nodes and edges according to it to obtain the newly added node data;

[0120] The dynamic knowledge graph unit is used to add new nodes and edges according to the newly added node data and updated edge data, and update the original edge weight values ​​to generate a new graph structure and obtain a dynamic knowledge graph.

[0121] In an embodiment of the present invention, an edge data updating unit is used to analyze the node data and edge data of a staged knowledge graph according to an edge incremental data set, identify the edge weight values ​​that need to be updated, and obtain updated edge data, which can accurately update the connection relationship in the graph so that the weight of each edge is more in line with the actual learning situation of the students; a new node data unit is used to identify the relationship between the newly added students and the knowledge points according to the edge incremental data set, obtain the newly added relationship data, and construct new graph structure nodes and edges therefrom to obtain the newly added node data, which can flexibly add new connections between students and knowledge points in the graph so that the graph can comprehensively reflect the students' latest learning achievements and knowledge mastery status; a dynamic knowledge graph unit is used to add new nodes and edges according to the newly added node data and the updated edge data, and update the original edge weight values ​​to generate a new graph structure to obtain a dynamic knowledge graph, which can ensure the system's real-time tracking of students' learning status.

[0122] The dynamic knowledge graph unit is used to add new nodes and edges according to the newly added node data and updated edge data, and update the original edge weight values ​​to generate a new graph structure and obtain a dynamic knowledge graph, which specifically includes:

[0123] First, the system will integrate the output results from the new node data unit and the updated edge data unit. The new node data unit provides the relationship data between the new students and knowledge points, which may include new student nodes, knowledge point nodes and new relationships between them, while the updated edge data unit provides update information on the existing edge weights based on the new learning data. This information reflects the changes in the students' mastery of certain knowledge points. After integrating these two data sets, the system can comprehensively check and update the nodes and edges in the entire graph structure to ensure that the new data is compatible and consistent with the existing data.

[0124] Next, the system needs to identify and analyze the needs for new nodes and edges. For new node data, the system will determine which nodes are new by comparing them with existing nodes. These may be new students or knowledge points, or some new learning behaviors. For each newly identified node, the system will create a new graph node for it and add it to the graph structure. For edge updates, the system will analyze the changing relationship between students and knowledge points based on the new learning data. For example, if a student's learning effect on a certain knowledge point increases significantly, the system will update the weight value of the edge between the student and the knowledge point based on the new learning record, thereby reflecting the student's actual mastery of the knowledge point.

[0125] After completing the identification of nodes and edges, the system will integrate the newly added nodes and updated edges and update the graph structure. Specifically, the newly added nodes will be connected to the existing nodes based on their relationship data, and the updated edge weight values ​​will adjust the existing node relationships based on the newly added learning data. This operation is completed through the graph structure construction algorithm. The algorithm will automatically adjust the layout of the graph according to the data relationship between the nodes and the edges, so that the new data can be effectively integrated into the entire graph structure. In this process, the system will not only add new nodes, but also ensure that the relationship between the old nodes is accurately reflected under the new learning data.

[0126] Finally, the system will generate a new dynamic knowledge graph. Each node of the graph represents a student or knowledge point. The edges between the nodes represent the learning relationship between the students and the knowledge points. The weight of the edges indicates the students' mastery of the knowledge points. In this way, the dynamic knowledge graph can update the students' learning status in real time and reflect the progress of each student in mastering each knowledge point. The generated dynamic knowledge graph is not only real-time and accurate, but also can evaluate students' learning progress in real time through continuous learning data updates and provide personalized educational guidance.

[0127] In a preferred embodiment of the present invention, the cycle analysis module includes:

[0128] The real-time mastery degree data unit is used to extract the latest learning time, homework scores and test scores of different knowledge points according to the dynamic knowledge graph, determine the real-time mastery degree of students in different knowledge points, and obtain real-time mastery degree data;

[0129] The mastery change trend data unit is used to compare the mastery degree at the real-time time point with the mastery degree at the previous time point according to the real-time mastery degree data, calculate the mastery change trend of different knowledge points, and obtain the mastery change trend data;

[0130] The real-time learning progress data set unit is used to determine the student's learning progress in the current cycle by merging the real-time mastery degree data and the mastery change trend data to obtain the real-time learning progress data set.

[0131] In an embodiment of the present invention, a real-time mastery degree data unit is used to extract the latest learning time, homework scores and test scores of different knowledge points according to a dynamic knowledge graph, determine the real-time mastery degree of students at different knowledge points, and obtain real-time mastery degree data, which can accurately reflect the students' mastery of each knowledge point in the current learning cycle; a mastery change trend data unit is used to compare the mastery degree of the real-time time point with the mastery degree of the previous time point according to the real-time mastery degree data, calculate the mastery change trend of different knowledge points, and obtain mastery change trend data to help identify whether the students are making progress or regressing on certain knowledge points; a real-time learning progress data set unit is used to determine the students' learning progress in the current cycle by merging the real-time mastery degree data and the mastery change trend data, and obtain a real-time learning progress data set, which can comprehensively consider the students' real-time mastery and learning change trends, and provide accurate learning progress evaluation.

[0132] In a preferred embodiment of the present invention, the period analysis module further includes:

[0133] The periodic analysis unit is used to perform periodic analysis on the learning progress of different knowledge points in the time dimension according to the real-time learning progress data set, and to accumulate and compare the learning progress data of multiple time nodes to obtain periodic analysis results;

[0134] The periodic evaluation unit is used to calculate the change rate of the mastery degree and the change rate of the learning progress of different knowledge points according to the periodic analysis results to obtain the periodic evaluation data.

[0135] In the embodiment of the present invention, the periodic analysis unit is used to perform periodic analysis on the learning progress of different knowledge points in the time dimension according to the real-time learning progress data set, and to accumulate and compare the learning progress data of multiple time nodes to obtain periodic analysis results, so as to provide cross-time dimension learning data analysis, and to identify the learning progress trend of students by comparing the learning progress of different periods;

[0136] The periodic evaluation unit is used to calculate the change rate of mastery of different knowledge points and the change rate of learning progress based on the periodic analysis results, and obtain periodic evaluation data, which can accurately analyze the changes in students' learning progress on each knowledge point.

[0137] Among them, the periodic analysis unit is used to perform periodic analysis on the learning progress of different knowledge points in the time dimension according to the real-time learning progress data set, and to accumulate and compare the learning progress data of multiple time nodes to obtain periodic analysis results, which specifically include:

[0138] First, extract real-time learning progress datasets from the dynamic knowledge graph to ensure that these data can cover different time nodes. These learning progress datasets include students’ study time, homework grades, test scores, etc. in each learning cycle. These data reflect students’ mastery of various knowledge points in a certain period of time.

[0139] During this process, these multi-dimensional learning progress data are sorted out, and the learning data of each time node are grouped according to knowledge points to ensure that the learning progress of each knowledge point at different time nodes can be clearly displayed. Through cumulative comparison, the system compares the learning progress of different time nodes to find out the changes in students' mastery of each knowledge point. For example, whether students can better master a certain knowledge point at a certain stage, or whether the mastery of certain knowledge points is stagnant.

[0140] Through this kind of data accumulation and comparative analysis in the time dimension, it is possible to reveal students' learning trends in long-term learning, including the speed of learning progress, the time periods of learning stagnation, and the students' progress on different knowledge points. Ultimately, periodic analysis results will be generated. These results provide basic data support for subsequent evaluations. The periodic analysis results can not only understand the students' overall learning status, but also identify the students' learning trajectories, providing a basis for the next step of learning adjustments.

[0141] The periodic evaluation unit is used to calculate the change rate of the mastery degree and the change rate of the learning progress of different knowledge points according to the periodic analysis results, and obtain periodic evaluation data, which specifically includes:

[0142] First, the system receives periodic analysis results, which include students' mastery of each knowledge point at multiple time points. These analysis results will be used to calculate the change rate of mastery of each knowledge point. The calculation of the mastery change rate is based on the comparison of students' learning data at different time points. By calculating these change rates, it is possible to quantitatively reveal the improvement or decline in students' learning progress on each knowledge point. For example, if a student's homework score on a certain knowledge point improves significantly within a certain period, the change rate of mastery of that knowledge point will also increase accordingly. Otherwise, it may show a downward trend.

[0143] In addition to calculating the rate of change of mastery level, the rate of change of students' learning progress will also be calculated. The basis for calculating the rate of change of learning progress is the students' learning time and performance data at different time nodes. By comparing the changes in students' learning time and performance in different cycles, the rate of change of learning progress is calculated. This process can reveal students' learning investment and actual progress in each knowledge point at different time nodes, and further understand students' learning motivation and learning effect.

[0144] After the calculation is completed, these change rate data will be integrated to generate periodic evaluation data. These data not only include the change rate of the mastery of each knowledge point, but also include the changes in students' progress on each knowledge point. This information will form a comprehensive periodic evaluation report to provide detailed learning feedback for teachers and students. Teachers can adjust teaching content or teaching methods based on these periodic evaluation data to provide more support for students' weak links; students can adjust their learning strategies based on the evaluation data and focus on those knowledge points that need to be strengthened, thereby improving learning efficiency.

[0145] In a preferred embodiment of the present invention, the evaluation module comprises:

[0146] The student ability unit is used to extract the students’ mastery of different knowledge points based on the periodic assessment data and obtain the students’ ability data;

[0147] The progress trend unit is used to compare the students' mastery levels at different time points based on their ability data, calculate the learning progress rate of different knowledge points, and obtain the progress trend data;

[0148] The weak knowledge point unit is used to calculate the deviation of students' mastery level on different knowledge points based on the progress trend data, identify students' weak knowledge points, and obtain weak knowledge point data.

[0149] In an embodiment of the present invention, a student ability unit is used to extract the students' mastery of different knowledge points based on periodic evaluation data to obtain student ability data, which can provide students with more accurate ability evaluation data, specifically manifested as a comprehensive quantification of the students' knowledge mastery; a progress trend unit is used to compare the students' mastery of different time points based on the student ability data, calculate the learning progress rate of different knowledge points, and obtain progress trend data, which can find out the knowledge points where students progress faster or slower in the learning process, so as to adjust the teaching strategy in a targeted manner; a weak knowledge point unit is used to calculate the students' mastery deviations on different knowledge points based on the progress trend data, identify the students' weak knowledge points, and obtain weak knowledge point data. By comparing the real-time mastery change rate with the historical average data, the system can accurately find the students' deficiencies in certain knowledge points.

[0150] Among them, the progress trend unit is used to compare the students' mastery levels at different time points based on the students' ability data, calculate the learning progress rate of different knowledge points, and obtain the progress trend data, which specifically includes:

[0151] First, the system extracts students' ability data at different time points. These data usually include students' academic performance, homework performance, test scores, and study time within a specific period of time. The system will obtain students' mastery of different periods based on the students' ability data in the periodic evaluation module. At each time point, the system will conduct a quantitative assessment of students' mastery of each knowledge point. These mastery levels are usually calculated based on comprehensive factors such as students' homework performance, test scores, and study time.

[0152] Next, the mastery levels at different time nodes are compared. This comparison is based on the time dimension, which means that the system not only considers the student performance at a single time point, but also compares the student's changes at multiple time nodes horizontally. Specifically, the system calculates the changes in the mastery level of each knowledge point between two time nodes. By comparing the student's performance at different learning stages, the progress trend unit can capture the student's changes in mastery of a certain knowledge point, and then identify the student's progress during these time periods.

[0153] Then, the learning progress rate of each knowledge point is calculated. The learning progress rate is quantified by the rate of change of the mastery level, that is, calculating the increase or decrease in the student's mastery of a certain knowledge point within a certain period of time. In specific implementation, the system will calculate the change in the mastery level and divide it by the corresponding time period to obtain the progress rate of each knowledge point. The progress rate can reflect the student's progress speed in the learning process and help teachers and students better understand the student's learning status on each knowledge point.

[0154] Finally, based on the calculated progress rate, progress trend data are generated. These data can not only reflect the students' mastery of the knowledge at a certain time point, but also show the students' progress trajectory of each knowledge point in the learning process. Progress trend data can identify whether students are making faster or slower progress on certain knowledge points, so as to make timely adjustments to the teaching plan. In addition, students can also clearly see their own learning progress through progress trend data.

[0155] In a preferred embodiment of the present invention, the weak knowledge point unit includes:

[0156] The mastery degree change data unit is used to extract the change rate of students' mastery degree of different knowledge points at different time points according to the progress trend data, and obtain the mastery degree change data;

[0157] The mastery degree deviation data unit is used to calculate the historical average mastery change rate of different knowledge points according to the mastery degree change data, and calculate the mastery degree deviation between the real-time mastery degree change rate of the students on different knowledge points and the historical average mastery change rate according to the mastery degree deviation data to obtain the mastery degree deviation data;

[0158] The weak knowledge point judgment unit is used to identify the students' weak knowledge points based on the mastery degree deviation data. When the mastery degree deviation of a certain knowledge point exceeds a preset deviation threshold, the knowledge point is the student's weak knowledge point, and a weak knowledge point data set is obtained.

[0159] In an embodiment of the present invention, a mastery degree change data unit is used to extract the mastery degree change rate of students on different knowledge points at different time nodes according to the progress trend data, and obtain the mastery degree change data, which can effectively quantify the students' progress on a certain knowledge point; a mastery degree deviation data unit is used to calculate the historical average mastery change rate of different knowledge points according to the mastery degree change data, and calculate the mastery degree deviation between the real-time mastery degree change rate of students on different knowledge points and the historical average mastery change rate according to the mastery degree deviation data, to obtain the mastery degree deviation data, by comparing the current change with the historical data, it is possible to clearly identify which knowledge points are significantly behind the historical level in learning progress, which provides an important basis for timely adjusting teaching strategies and providing personalized support for students; a weak knowledge point judgment unit is used to identify the students' weak knowledge points according to the mastery degree deviation data, when the mastery degree deviation of a certain knowledge point exceeds the preset deviation threshold, the knowledge point is the student's weak knowledge point, and the weak knowledge point data set is obtained, which can effectively and automatically identify the students' weak links.

[0160] Among them, the mastery degree deviation data unit is used to calculate the historical average mastery change rate of different knowledge points according to the mastery degree change data, and calculate the mastery degree deviation between the real-time mastery degree change rate of students on different knowledge points and the historical average mastery change rate according to the mastery degree change data, and obtain the mastery degree deviation data, which specifically includes:

[0161] First, obtain the real-time changes in the students' mastery of each knowledge point in the current learning cycle. The key to this process is to compare the students' mastery of each knowledge point at the current time node with the previous time node, and calculate the mastery change rate through this comparison. In order to complete this operation, the system will extract homework grades, test scores, study time and other information from the students' learning data. This information can help the system evaluate the students' mastery of a certain knowledge point in the current cycle. By comparing it with the mastery of the previous cycle, the system calculates the changes in the mastery of each knowledge point in this cycle, forming real-time mastery change data.

[0162] Secondly, it is also necessary to obtain the historical average mastery changes of each knowledge point. This process involves extracting the student’s mastery of each knowledge point in different learning cycles from the student’s historical learning data. By analyzing these historical data, the system can calculate the average mastery changes of each knowledge point in the past multiple learning cycles. This historical data provides an important reference benchmark for evaluating the student’s current learning progress, because it can reflect the student’s long-term learning trend on each knowledge point.

[0163] Next, the system compares the real-time mastery changes with the historical average mastery changes, and calculates the mastery degree deviation of each knowledge point. By comparing the differences between the students' current mastery changes of the knowledge points and the long-term mastery changes, the system can identify abnormalities in the students' learning progress on certain knowledge points. For example, if the students' mastery changes on a certain knowledge point exceed the historical learning trajectory, the system will detect this abnormality and indicate that there may be learning difficulties or too rapid progress.

[0164] Finally, the calculated deviation data will be sorted and a data set will be generated. This data includes information such as the real-time mastery changes, historical mastery changes, and mastery degree deviations of each knowledge point. The generated data set will be stored and synchronized to the teacher and student ends for use by teachers and students. Teachers can use this data to understand students' mastery of each knowledge point, and students can use this data to clearly understand their weak links.

[0165] The preset deviation threshold is used to determine whether there is a significant mastery deviation when students learn a certain knowledge point. First, the system extracts the changes in students' mastery of different knowledge points from historical student learning data. These data include information such as student performance fluctuations, study time, and homework completion in each learning cycle. By analyzing these data, the system calculates the normal mastery fluctuation range of students on each knowledge point and sets the deviation threshold based on these fluctuation ranges.

[0166] The setting of specific values ​​is usually based on the standard deviation of changes in students' mastery of a certain knowledge point. Assuming that the rate of change in students' mastery of a certain knowledge point fluctuates within a certain range over a period of time, the system will set the deviation threshold by calculating the standard deviation of the rate of change. Generally speaking, the system will set the threshold to 1.5 times the standard deviation. That is, when the deviation in the student's mastery of a certain knowledge point exceeds this value, the knowledge point is considered to be a weak knowledge point. For example, if historical data shows that the standard deviation of the fluctuation in the rate of change in mastery of a certain knowledge point is 0.2, then the preset deviation threshold may be set to 0.3, which is 1.5 times the standard deviation. In this way, when the student's rate of change in mastery of the knowledge point exceeds 0.3, the system will judge the knowledge point to be a weak knowledge point.

[0167] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A full-cycle intelligent education evaluation system based on knowledge graph, characterized in that: The system comprises: The association module is used to perform association analysis on the student's learning behavior data set and the knowledge point data set, calculate the student's mastery of each knowledge point, and obtain the learning progress data set; The first graph module is used to identify the relationship between students and each knowledge point according to the learning progress data set, obtain knowledge relationship data, and map it into a graph structure to obtain a staged knowledge graph; The second graph module is used to obtain the students' newly added learning data set, and update the phased knowledge graph based on it, adjust the relationship between the students and each knowledge point, and obtain a dynamic knowledge graph; The cycle analysis module is used to extract the real-time mastery of each knowledge point based on the dynamic knowledge graph, obtain the real-time learning progress data set, and conduct periodic analysis on it in combination with the time dimension to obtain periodic evaluation data; The evaluation module is used to analyze students’ mastery of each knowledge point, progress trends, and weak knowledge points based on periodic evaluation data, and generate individual evaluation data sets; The feedback module is used to send individual evaluation data sets to the student and teacher terminals synchronously to adjust students' learning behaviors.

2. According to claim 1, a full-cycle intelligent education evaluation system based on knowledge graph is characterized in that: The association module includes: A learning behavior data acquisition unit is used to acquire the learning time, homework scores and test scores of students in different learning cycles to obtain a learning behavior data set; The association analysis unit is used to extract the homework scores and test scores of students related to different knowledge points in different learning cycles according to the learning behavior data set, and calculate the average homework scores and average test scores in different learning cycles to obtain the knowledge point related data set; The mastery degree calculation unit is used to integrate the learning time, average homework scores and average test scores related to different knowledge points in different learning cycles of students according to the knowledge point related data sets, calculate the mastery degree of students on different knowledge points, and obtain the learning progress data set.

3. According to claim 2, a full-cycle intelligent education evaluation system based on knowledge graph is characterized in that: The first graph module includes: The knowledge point learning data set unit is used to extract the student's access records to different knowledge points, the number of learning tasks and the accumulated learning time in the learning cycle according to the learning progress data set to obtain the knowledge point learning data set; A coverage index data set unit is used to calculate the ratio between the number of knowledge points that each student has learned in different courses and the total number of knowledge points in the course according to the knowledge point learning data set, so as to obtain a coverage index data set; The completion depth indicator data set unit is used to count the number of learning tasks and cumulative learning time of each student in different courses according to the knowledge point learning data set to obtain the completion depth indicator data set; The learning relationship analysis unit is used to determine the learning relationship between students and knowledge points in different courses based on the coverage indicator data set and the completion depth indicator data set to obtain a phased knowledge graph.

4. According to claim 3, a full-cycle intelligent education evaluation system based on knowledge graph is characterized in that: The learning relationship analysis unit comprises: The node data unit is used to define the nodes of the graph structure according to the learning relationship data, define the student identifier as a student node, define the knowledge point as a knowledge point node, and obtain the node data; The edge data unit is used to construct the edges of the graph structure according to the learning relationship data, integrate the coverage index with the completion depth index, calculate the edge weight values ​​between students and different knowledge points, and obtain edge data; The graph construction unit is used to connect student nodes with different knowledge point nodes according to node data and edge data, and combine them to form an undirected graph structure to obtain a phased knowledge graph.

5. According to claim 4, a full-cycle intelligent education evaluation system based on knowledge graph is characterized in that: The second atlas module includes: A new learning data set unit is added to obtain the learning time, homework scores and test scores in the current learning cycle, and merge them according to knowledge points to obtain a new learning data set; The standard data set updating unit is used to update the coverage index and completion depth index between the learning and different knowledge points according to the newly added learning data set, so as to obtain the updated standard data set; The learning relationship change data unit is used to calculate the difference between the updated standard data set and the staged knowledge graph to obtain the learning relationship change data; The edge increment data set unit is used to adjust the edge weight value in the staged knowledge graph according to the learning relationship change data, update the connection strength between students and different knowledge points, and obtain the edge increment data set; The graph update unit is used to construct new graph structure nodes and edges based on the edge incremental data set, and to replace the structure of the phased knowledge graph to obtain a dynamic knowledge graph.

6. According to claim 5, a full-cycle intelligent education evaluation system based on knowledge graph is characterized in that: The atlas updating unit comprises: The edge data updating unit is used to analyze the node data and edge data of the phased knowledge graph according to the edge incremental data set, identify the edge weight value that needs to be updated, and obtain the updated edge data; A new node data unit is added, which is used to identify the relationship between the newly added students and knowledge points according to the edge incremental data set, obtain the newly added relationship data, and construct new graph structure nodes and edges according to it to obtain the newly added node data; The dynamic knowledge graph unit is used to add new nodes and edges according to the newly added node data and updated edge data, and update the original edge weight values ​​to generate a new graph structure and obtain a dynamic knowledge graph.

7. A full-cycle intelligent education evaluation system based on knowledge graph according to claim 6, characterized in that: The cycle analysis module comprises: The real-time mastery degree data unit is used to extract the latest learning time, homework scores and test scores of different knowledge points according to the dynamic knowledge graph, determine the real-time mastery degree of students in different knowledge points, and obtain real-time mastery degree data; A mastery change trend data unit is used to compare the mastery degree at the real-time time point with the mastery degree at the previous time point according to the real-time mastery degree data, calculate the mastery change trend of different knowledge points, and obtain the mastery change trend data; The real-time learning progress data set unit is used to determine the learning progress of the students in the current cycle by merging the real-time mastery degree data and the mastery change trend data to obtain the real-time learning progress data set.

8. The full-cycle intelligent education evaluation system based on knowledge graph according to claim 7 is characterized in that: The cycle analysis module also includes: The periodic analysis unit is used to perform periodic analysis on the learning progress of different knowledge points in the time dimension according to the real-time learning progress data set, and to accumulate and compare the learning progress data of multiple time nodes to obtain periodic analysis results; The periodic evaluation unit is used to calculate the change rate of the mastery degree and the change rate of the learning progress of different knowledge points according to the periodic analysis results to obtain the periodic evaluation data.

9. A full-cycle intelligent education evaluation system based on knowledge graph according to claim 8, characterized in that: The evaluation module includes: The student ability unit is used to extract the students’ mastery of different knowledge points based on the periodic assessment data and obtain the students’ ability data; The progress trend unit is used to compare the students' mastery levels at different time points based on their ability data, calculate the learning progress rate of different knowledge points, and obtain the progress trend data; The weak knowledge point unit is used to calculate the deviation of students' mastery level on different knowledge points based on the progress trend data, identify students' weak knowledge points, and obtain weak knowledge point data.

10. A full-cycle intelligent education evaluation system based on knowledge graph according to claim 9, characterized in that: The weak knowledge point units include: The mastery degree change data unit is used to extract the change rate of students' mastery degree of different knowledge points at different time points according to the progress trend data, and obtain the mastery degree change data; The mastery degree deviation data unit is used to calculate the historical average mastery change rate of different knowledge points according to the mastery degree change data, and calculate the mastery degree deviation between the student's real-time mastery degree change rate on different knowledge points and the historical average mastery change rate according to the mastery degree deviation data to obtain the mastery degree deviation data; The weak knowledge point judgment unit is used to identify the students' weak knowledge points based on the mastery degree deviation data. When the mastery degree deviation of a certain knowledge point exceeds a preset deviation threshold, the knowledge point is the student's weak knowledge point, and a weak knowledge point data set is obtained.

Images