A student achievement report data management method and system

Abstract

The present application relates to the technical field of electronic data processing, and discloses a student achievement report data management method and system, which comprises the following steps: dividing students into different student groups; converting the achievements of the students into achievement grade items; calculating the item ordering indexes corresponding to the student groups; initializing an FP tree, wherein the nodes of the FP tree contain item names, group count vectors and semester sets; arranging the achievement grade items in the student achievement records in descending order and inserting them into the FP tree, updating the group count vectors and the semester sets of each node; obtaining the group count vectors and the semester sets of the end nodes corresponding to each prefix path, and judging whether the group dominance threshold is met and whether the semester distribution constraint is met; only retaining the prefix paths meeting the conditions to generate a conditional FP tree, and recursively mining to obtain frequent patterns. The present application comprehensively analyzes the student group characteristics, the item set correlation and the time sequence information in a unified framework, and improves the depth and accuracy of the mining results.

Description

Technical Field

[0001] This invention relates to the field of electronic data processing technology, specifically to a method and system for managing student grade report data. Background Technology

[0002] Student performance data mining plays a crucial role in education today, helping us to deeply analyze and understand students' learning behaviors, predict potential academic risks, and ultimately achieve personalized instruction. The core technology of performance data mining is association rule mining, which aims to discover valuable relationships between data items, such as the correlation between a specific combination of courses and students' academic performance.

[0003] Classic association rule mining algorithms, such as FP-Growth, effectively avoid generating massive candidate sets by constructing a data structure called an FP-tree, thus significantly improving mining efficiency. However, these methods typically analyze all students as a homogeneous whole, ignoring the inherent differences between student groups, such as different professional backgrounds, initial entry requirements, or learning habits; their performance patterns and behavioral patterns may be completely different.

[0004] To address this issue, some studies have attempted a "clustering + association rule" strategy. First, clustering algorithms are used to divide students into different groups, and then FP-Growth mining is performed independently on each group. However, this approach separates the grouping and mining processes, increasing overall computational complexity and making it difficult to perform horizontal comparisons and association analyses of patterns across different groups within a unified framework. Furthermore, some studies have introduced methods such as sequence pattern mining to analyze the temporal attributes of performance data. While this helps reveal temporal patterns in students' learning processes, these methods are often incompatible with FP-Growth algorithms in terms of data structure and mining objectives. This makes it difficult to comprehensively analyze student group characteristics, itemset associations, and temporal information, resulting in insufficient depth and low accuracy in the mining results. Summary of the Invention

[0005] This invention provides a student performance report data management method and system to solve the problem in the prior art that it is difficult to integrate and analyze student group characteristics, itemset correlation and time series information in an integrated manner, resulting in insufficient depth and low accuracy of the mining results.

[0006] In a first aspect, the student grade report data management method of the present invention includes the following steps:

[0007] Students are divided into different student groups based on historical academic data and course selection behavior; students' course scores are converted into discrete grade level items; for each student group, the item ranking index corresponding to the student group is calculated and obtained based on the statistical distribution characteristics of each grade level item within the student group.

[0008] Initialize the FP-tree. Each node of the FP-tree contains an item name, a group count vector for recording the count of each student group, and a semester set for recording the semester information of the path where the node is located.

[0009] Traverse each student grade record, sort the grade level items in the student grade record in descending order according to the item sorting index of the student group to which the student grade record belongs, insert the descending sorted item sequence as a path into the FP tree, and update the group count vector and semester set of each node on the path.

[0010] When constructing a conditional pattern base for items in the item header table, the population count vector and semester set of the terminal node corresponding to each prefix path are obtained. It is then determined whether the population count vector meets the preset population dominance threshold and whether the semester set meets the preset semester distribution constraint. Only the prefix paths that meet the conditions are retained to generate a conditional FP tree, and frequent patterns including population characteristics and time series information are obtained through recursive mining.

[0011] Preferably, the step of dividing students into different student groups based on historical academic data and course selection behavior includes: extracting the average grade point average of each student as a feature of historical academic data, counting the number of science, engineering, humanities, and social science courses taken as a feature of course selection behavior, vectorizing the historical academic data features and course selection behavior features, and using the K-means clustering algorithm to divide all students into K student groups, where K is a preset integer and K is greater than or equal to 2.

[0012] Preferably, the step of calculating and obtaining the item ranking index corresponding to each student group based on the statistical distribution characteristics of each grade level item within the student group includes: for each student group, counting the total number of times each grade level item appears within the student group, and using the total number of times as the item ranking index of the grade level item within the student group.

[0013] Preferably, inserting the descendingly ordered item sequence as a path into the FP-tree and updating the population count vector and semester set of each node on the path includes: when inserting the item sequence as a path into the FP-tree, for each node on the path, incrementing the component corresponding to the student's population in the node's population count vector by 1; simultaneously, adding the semester information corresponding to the student's grade record to the semester set of the node, and not adding it again if the semester information already exists.

[0014] Preferably, determining whether the population counting vector satisfies a preset population dominance threshold includes: for a target mining population, extracting the count value of the target mining population from the population counting vector of the terminal node corresponding to the prefix path. And calculate the sum of the counts for all other groups. When the inequality is satisfied When the threshold for group dominance is met, it is determined that the group dominance threshold is satisfied. The pre-set dominance coefficient, .

[0015] Preferably, whether the semester set satisfies the preset semester distribution constraint includes: counting the number of independent semesters contained in the semester set of the terminal node corresponding to the prefix path. When the inequality is satisfied When it is determined that the semester distribution constraint is satisfied, then... This is the preset minimum number of semesters required. It is an integer greater than or equal to 1.

[0016] Secondly, the student grade report data management system of the present invention includes the following modules:

[0017] The classification module is used to divide students into different student groups based on historical academic data and course selection behavior; convert students' course scores into discrete grade level items; and calculate and obtain the item ranking index corresponding to each student group based on the statistical distribution characteristics of each grade level item within the student group.

[0018] The Initialize FP Tree module is used to initialize the FP tree. The nodes of the FP tree contain an item name, a group count vector for recording the count of each student group, and a semester set for recording the semester information of the path where the node is located.

[0019] The update module is used to traverse each student grade record, sort the grade level items in the student grade record in descending order according to the item sorting index of the student group to which the student grade record belongs, insert the descending sorted item sequence as a path into the FP tree, and update the group count vector and semester set of each node on the path.

[0020] The mining module is used to obtain the population count vector and semester set of the terminal node corresponding to each prefix path when constructing the conditional pattern base for the items in the item header table. It also determines whether the population count vector meets the preset population dominance threshold and whether the semester set meets the preset semester distribution constraint. Only the prefix paths that meet the conditions are retained to generate the conditional FP tree, and frequent patterns including population characteristics and time series information are obtained through recursive mining.

[0021] Preferably, the step of dividing students into different student groups based on historical academic data and course selection behavior includes: extracting the average grade point average of each student as a feature of historical academic data, counting the number of science, engineering, humanities, and social science courses taken as a feature of course selection behavior, vectorizing the historical academic data features and course selection behavior features, and using the K-means clustering algorithm to divide all students into K student groups, where K is a preset integer and K is greater than or equal to 2.

[0022] Preferably, the step of calculating and obtaining the item ranking index corresponding to each student group based on the statistical distribution characteristics of each grade level item within the student group includes: for each student group, counting the total number of times each grade level item appears within the student group, and using the total number of times as the item ranking index of the grade level item within the student group.

[0023] Preferably, inserting the descendingly ordered item sequence as a path into the FP-tree and updating the population count vector and semester set of each node on the path includes: when inserting the item sequence as a path into the FP-tree, for each node on the path, incrementing the component corresponding to the student's population in the node's population count vector by 1; simultaneously, adding the semester information corresponding to the student's grade record to the semester set of the node, and not adding it again if the semester information already exists.

[0024] The beneficial effects of this invention are as follows: By utilizing group-specific item ranking indicators and group dominance constraints during the mining process, this invention can uncover and highlight the unique performance correlation patterns of specific student groups, avoiding information overload caused by global mining, thus better reflecting the heterogeneity of student groups. Simultaneously, by introducing semester distribution constraints during mining, the frequently discovered patterns can reflect the regularities related to specific learning stages or time evolution, organically integrating the temporal information of the data into the analysis. This invention comprehensively analyzes student group characteristics, itemset correlations, and temporal information within a unified framework, improving the depth, accuracy, and practicality of the mining results. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating the student grade report data management method provided in an embodiment of the present invention. Detailed Implementation

[0026] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0027] like Figure 1 As shown, an embodiment of the student grade report data management method provided by the present invention specifically includes the following steps:

[0028] S1. Based on historical academic data and course selection behavior, students are divided into different student groups; students' course grades are converted into discrete grade level items; for each student group, based on the statistical distribution characteristics of each grade level item within the student group, the item ranking index corresponding to the student group is calculated and obtained.

[0029] Specifically, historical academic data features for each student are extracted, such as GPA, number of failed courses, and number of scholarships received. Course selection behavior features for each student are also extracted, such as the number of courses selected per semester, the average difficulty level of selected courses, and the proportion of students selecting courses outside their major. These data features are used as feature vectors, and the K-Means clustering algorithm is used to perform unsupervised clustering of all students, grouping students with similar feature vector spatial distances into the same group. For example, students can be divided into three groups: Group 1 is academically excellent, characterized by a high GPA, no failed courses, and a large number of high-difficulty courses; Group 2 is academically stable, characterized by a moderate GPA, occasional low scores, and regular course selection behavior; Group 3 is academically risky, characterized by a low GPA, multiple failed courses, and a small number of courses selected.

[0030] For example, for students' percentage-based grades, fixed conversion rules are used for discretization. For instance, 90 and above are defined as Excellent, 80-89 as Good, 70-79 as Average, 60-69 as Pass, and below 60 as Fail. The course name is combined with this grade to form a grade level item. For example, a student's math score of 92 is converted to the grade level item: Math - Excellent. All of each student's course grades within a semester are converted into a set consisting of multiple grade level items, forming the student's grade record.

[0031] Specifically, for each student group, based on the statistical distribution characteristics of each grade level item within the student group, the item ranking index corresponding to the student group is calculated and obtained. This includes: for each student group, counting the total number of times each grade level item appears within that student group, and using this total number of times as the item ranking index for the grade level item within the student group. For example, for Group 1 students with excellent academic performance, the statistics show that "Mathematics - Excellent" appeared 200 times, "Physics - Excellent" appeared 150 times, and "Mathematics - Pass" appeared only 5 times. Therefore, for Group 1, the item ranking index for "Mathematics - Excellent" is 200, the item ranking index for "Physics - Excellent" is 150, and the item ranking index for "Mathematics - Pass" is 5. This calculation is performed for all grade level items and all student groups, thereby generating an independent list of item ranking indices containing the internal grade distribution patterns for each student group.

[0032] In an optional embodiment, the step of dividing students into different student groups based on historical academic data and course selection behavior includes: extracting the average grade point average of each student as a feature of historical academic data, counting the number of science, engineering, humanities, and social science courses taken as a feature of course selection behavior, vectorizing the historical academic data features and course selection behavior features, and then using the K-means clustering algorithm to divide all students into K student groups, where K is a preset integer and K is greater than or equal to 2.

[0033] For example, suppose we need to divide 1,000 students in a university into three groups, where K equals 3. First, we extract the historical academic data features of each student. For instance, student A has an average GPA of 3.8 and has taken 10 STEM courses, 2 humanities courses, and 1 social science course. Student A's feature vector would then be [3.8, 10, 2, 1]. We perform the same operation on all 1,000 students, obtaining 1,000 such four-dimensional feature vectors. We then input these 1,000 feature vectors into a K-means clustering algorithm. Through iterative calculation, the 1,000 students are divided into three different student groups. For example, group one might have a large number of students who have taken STEM courses and generally have high average GPAs, and could be labeled as the STEM high-achieving group; group two might have taken more humanities and social science courses and have moderate GPAs, and could be labeled as the humanities well-rounded group; group three might have a more dispersed distribution of features or a lower average GPA.

[0034] S2, initialize the FP tree. The nodes of the FP tree contain the item name, the group count vector used to record the count of each student group, and the semester set that records the semester information of the path where the node is located.

[0035] Specifically, a root node of an FP-tree is constructed. This root node does not store any item information. Other nodes in the FP-tree contain the following fields: Item Name field, used to store the name of the grade item, such as Math - Excellent; Parent Node Pointer, pointing to its parent node; Child Node Pointer List, pointing to all its child nodes; Group Count Vector, assuming there are three student groups, this group count vector is an integer array of length 3, for example (0,0,0), where the three components of the array correspond to the number of times the path passes through this node in group 1, group 2, and group 3, respectively; Semester Set, used to store a set of data, recording the semester identifiers corresponding to all paths passing through this node, such as Fall 2022 semester and Spring 2023 semester. Simultaneously, an item header table is initialized. The item header table is used to index all unique grade items in the FP-tree and their node chains in the FP-tree. The item header table contains all "grade items" that satisfy the minimum support (i.e., minimum occurrence count). The item header table contains two parts: the item name, which is the specific "grade item," and the head pointer, which points to the first node in the FP-tree where that item appears.

[0036] S3, traverse each student grade record, sort the grade level items in the student grade record in descending order according to the item sorting index of the student group to which the student grade record belongs, insert the descending sorted item sequence as a path into the FP tree, and update the group count vector and semester set of each node on the path.

[0037] Specifically, inserting the descendingly ordered item sequence as a path into the FP-tree and updating the population count vector and semester set of each node on the path includes: when inserting the item sequence as a path into the FP-tree, for each node on the path, incrementing the component corresponding to the student's population in the node's population count vector by 1; at the same time, adding the semester information corresponding to the student's grade record to the semester set of the node, if the semester information already exists, it will not be added again.

[0038] For example, consider a student's grade record belonging to Group 1. This record contains three grade levels: Math - Excellent, Physics - Good, and English - Excellent, and the record was created in the Fall 2022 semester. Query the item ranking index for Group 1. Assuming English - Excellent has the highest ranking index, followed by Math - Excellent, and then Physics - Good, rearrange the grade levels in the student's record in this order to English - Excellent, Math - Excellent, and Physics - Good, forming an item sequence. Use this item sequence as a path, inserting it starting from the root node of the FP-tree. If the root node does not yet have a child node for "English - Excellent," create a new node with the item name "English - Excellent," updating its group count vector to (1, 0, 0), and adding "Fall 2022 Semester" to the semester set. Then, using this new node as the parent node, insert "Math - Excellent," similarly creating a new node and updating its group count vector to (1, 0, 0), adding "Fall 2022 Semester" to the semester set. Continue this process to complete the insertion of the entire path. If a node already exists on the path, there is no need to create a new node. Simply add 1 to the corresponding group position in its group counting vector and add the semester identifier of the current record to its semester set.

[0039] S4. When constructing the conditional pattern base for the items in the item header table, obtain the population count vector and semester set of the terminal node corresponding to each prefix path, and determine whether the population count vector meets the preset population dominance threshold, and whether the semester set meets the preset semester distribution constraint. Only the prefix paths that meet the conditions are retained to generate the conditional FP tree, and frequent patterns including population characteristics and time series information are obtained through recursive mining.

[0040] Specifically, the conditional pattern base is a concept in the frequent itemset mining algorithm FP-Growth. The conditional pattern base consists of two parts: the first part is the prefix path, which is all paths in the FP-tree from the root node to the node preceding item X. The second part is the count value of each prefix path, where the count value of each prefix path is equal to the count value of item X. The construction process of the conditional pattern base is as follows: find item X through the item head list, and traverse all nodes of item X in the FP-tree along its node linked list. For each found item X node, backtrack from that item X node along its parent node to the root node of the FP-tree; this path, which does not include item X itself, is the prefix path. From the terminal node (i.e., the X node) corresponding to this prefix path, obtain the population count vector and the semester set. Then check whether the population count vector obtained from the X node meets the preset population dominance threshold and whether the semester set obtained from the X node meets the preset semester distribution constraint. Prefix paths that simultaneously meet the above two conditions are retained to form a conditional pattern base. Based on this, a conditional FP tree is constructed, and frequent patterns including population characteristics and time series information are obtained through recursive mining.

[0041] For example, suppose we need to construct a conditional pattern base for the item "Mathematics - Excellent". We find all "Mathematics - Excellent" nodes using the item header table. For each "Mathematics - Excellent" node, we trace its prefix path upwards to the root node. When obtaining each prefix path, we record the group count vector and semester set of the "Mathematics - Excellent" node. Assuming a preset group dominance threshold of 0.8, a condition is considered met when the group dominance of any group in the prefix path is greater than the threshold. The preset semester distribution constraint is that the path involves at least two semesters. Assuming the group count vector of the terminal node of a prefix path is (90, 5, 5), then the group dominance of group one is equal to the corresponding component of group one divided by the sum of the three components, i.e., 90 / 100 = 0.9. Since the group dominance of group one is greater than 0.8, this prefix path satisfies the preset group dominance threshold condition. Meanwhile, assuming the semester set of the terminal nodes of this prefix path includes "Fall Semester 2022" and "Spring Semester 2023", the size of the semester set of the terminal nodes of this prefix path is 2, satisfying the preset semester distribution constraint. Since the group count vector of the terminal nodes of this prefix path and the semester set satisfy two conditions, this prefix path is retained and marked as related to Group 1. All retained labeled prefix paths are collected to form a conditional pattern base, and a conditional FP-tree is constructed based on this. Then, recursively mining is performed to obtain frequent patterns. Frequent patterns include "English - Excellent", "Group 1", and "Spanning 2 semesters". Frequent patterns cover the strong correlation of a specific group and its time span information.

[0042] In an optional embodiment, determining whether the population counting vector satisfies a preset population dominance threshold includes: for a target mining population, extracting the count value of the target mining population from the population counting vector of the terminal node corresponding to the prefix path. And calculate the sum of the counts for all other groups. When the inequality is satisfied When the threshold for group dominance is met, it is determined that the group dominance threshold is satisfied. The pre-set dominance coefficient, .

[0043] For example, assume the target mining population is population 1, and the dominance coefficient α is 1.5. The population count vector of the terminal node corresponding to the prefix path is [20, 5, 5]. In this scenario, the target mining population is population 1, and its corresponding count value is... The sum is 20. The counts for the other groups are 5 and 5 respectively, therefore the sum of the counts for the other groups is 20. Since 20 is greater than 15, the prefix path is determined to meet the group dominance threshold.

[0044] In an optional embodiment, determining whether the semester set satisfies a preset semester distribution constraint includes: counting the number of independent semesters contained in the semester set of the terminal nodes corresponding to the prefix path. When the inequality is satisfied When it is determined that the semester distribution constraint is satisfied, then... This is the preset minimum number of semesters required. It is an integer greater than or equal to 1.

[0045] For example, suppose the minimum required number of semesters is... Let's set it to 3. Assume the semester set of the terminal node corresponding to the prefix path contains the following semester information: Fall 2021, Spring 2022, Fall 2021, Fall 2022, Spring 2022. After deduplication and counting the number of unique semesters included, we find three unique semesters: Fall 2021, Spring 2022, and Fall 2022. Therefore, Equals 3. (The rest of the text appears to be a series of numbers and symbols, possibly related to a specific product or service.) Compared with the preset minimum number of semesters Comparison, because Therefore, the prefix path satisfies the semester distribution constraint, indicating that this prefix path does not only appear in one semester, but is exhibited by students in at least three different semesters, and has a certain degree of universality and persistence.

[0046] The implementation principle of the student performance report data management method in this invention is as follows: This invention utilizes student-specific item ranking indicators and group dominance constraints during the data mining process to uncover and highlight unique performance correlation patterns within a student group, avoiding information overload caused by global data mining and thus better reflecting the heterogeneity of the student group. Furthermore, by introducing semester distribution constraints during mining, the frequently discovered patterns reflect regularities related to specific learning stages or time progression, organically integrating the temporal information of the data into the analysis. This invention enables comprehensive analysis of student group characteristics, itemset correlations, and temporal information within a unified framework, improving the depth, accuracy, and practicality of the mining results.

[0047] The student grade report data management system provided by this invention includes the following modules:

[0048] The classification module is used to divide students into different student groups based on historical academic data and course selection behavior; convert students' course scores into discrete grade level items; and calculate and obtain the item ranking index corresponding to each student group based on the statistical distribution characteristics of each grade level item within the student group.

[0049] The Initialize FP Tree module is used to initialize the FP tree. The nodes of the FP tree contain an item name, a group count vector for recording the count of each student group, and a semester set for recording the semester information of the path where the node is located.

[0050] The update module is used to traverse each student grade record, sort the grade level items in the student grade record in descending order according to the item sorting index of the student group to which the student grade record belongs, insert the descending sorted item sequence as a path into the FP tree, and update the group count vector and semester set of each node on the path.

[0051] The mining module is used to obtain the population count vector and semester set of the terminal node corresponding to each prefix path when constructing the conditional pattern base for the items in the item header table. It also determines whether the population count vector meets the preset population dominance threshold and whether the semester set meets the preset semester distribution constraint. Only the prefix paths that meet the conditions are retained to generate the conditional FP tree, and frequent patterns including population characteristics and time series information are obtained through recursive mining.

[0052] Specifically, the step of dividing students into different student groups based on historical academic data and course selection behavior includes: extracting the average grade point average of each student as a feature of historical academic data, counting the number of science, engineering, humanities, and social science courses taken as a feature of course selection behavior, vectorizing the historical academic data features and course selection behavior features, and then using the K-means clustering algorithm to divide all students into K student groups, where K is a preset integer and K is greater than or equal to 2.

[0053] Specifically, the step of calculating and obtaining the item ranking index corresponding to each student group based on the statistical distribution characteristics of each grade level item within the student group includes: for each student group, counting the total number of times each grade level item appears within the student group, and using the total number of times as the item ranking index of the grade level item within the student group.

[0054] Specifically, inserting the descendingly ordered item sequence as a path into the FP-tree and updating the population count vector and semester set of each node on the path includes: when inserting the item sequence as a path into the FP-tree, for each node on the path, incrementing the component corresponding to the student's population in the node's population count vector by 1; at the same time, adding the semester information corresponding to the student's grade record to the semester set of the node, if the semester information already exists, it will not be added again.

[0055] Specifically, determining whether the population counting vector satisfies a preset population dominance threshold includes: for a target mining population, extracting the count value of the target mining population from the population counting vector of the terminal node corresponding to the prefix path. And calculate the sum of the counts for all other groups. When the inequality is satisfied When the threshold for group dominance is met, it is determined that the group dominance threshold is satisfied. The pre-set dominance coefficient, .

[0056] Specifically, whether the semester set satisfies the preset semester distribution constraint includes: counting the number of independent semesters contained in the semester set of the terminal node corresponding to the prefix path. When the inequality is satisfied When it is determined that the semester distribution constraint is satisfied, then... This is the preset minimum number of semesters required. It is an integer greater than or equal to 1.

[0057] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for managing student grade report data, characterized in that, The steps include: dividing students into different student groups based on historical academic data and course selection behavior; converting students' course grades into discrete grade level items; and calculating and obtaining the item ranking index corresponding to each student group based on the statistical distribution characteristics of each grade level item within the student group. Initialize the FP-tree. Each node of the FP-tree contains an item name, a group count vector for recording the count of each student group, and a semester set for recording the semester information of the path where the node is located. Traverse each student grade record, sort the grade level items in the student grade record in descending order according to the item sorting index of the student group to which the student grade record belongs, insert the descending sorted item sequence as a path into the FP tree, and update the group count vector and semester set of each node on the path. When constructing the conditional pattern base for items in the item header table, the population count vector and semester set of the terminal nodes corresponding to each prefix path are obtained, and it is determined whether the population count vector meets the preset population dominance threshold. This includes: for a target mining population, extracting the count value of the target mining population from the population count vector of the terminal nodes corresponding to the prefix path. And calculate the sum of the counts for all other groups. When the inequality is satisfied When the threshold for group dominance is met, it is determined that the group dominance threshold is satisfied. The pre-set dominance coefficient, Determine whether the semester set satisfies the preset semester distribution constraints, including: counting the number of independent semesters contained in the semester set of the terminal node corresponding to the prefix path. When the inequality is satisfied When it is determined that the semester distribution constraint is satisfied, then... This is the preset minimum number of semesters required. For integers greater than or equal to 1, only the prefix paths that meet the conditions are retained to generate a conditional FP-tree, and frequent patterns including population characteristics and temporal information are obtained through recursive mining; By introducing semester distribution constraints during data mining, the frequently mined patterns can reflect the regularity of learning stages or time evolution, and the temporal information of the data can be organically integrated into the analysis.

2. The student grade report data management method according to claim 1, characterized in that, The method of dividing students into different student groups based on historical academic data and course selection behavior includes: The average grade point average of each student is extracted as a feature of historical academic data. The number of science, engineering, humanities, and social science courses taken is counted as a feature of course selection behavior. After vectorizing the features of historical academic data and course selection behavior, the K-means clustering algorithm is used to divide all students into K student groups, where K is a preset integer and K is greater than or equal to 2.

3. The student grade report data management method according to claim 1, characterized in that, For each student group, based on the statistical distribution characteristics of each grade level item within the student group, the item ranking index corresponding to the student group is calculated and obtained, including: For each student group, the total number of times each grade level item appears within that student group is counted, and this total number of times is used as the ranking index for the grade level item within the student group.

4. The student grade report data management method according to claim 1, characterized in that, The step of inserting the descendingly ordered sequence of items into the FP-tree as a path, and updating the population count vector and semester set of each node on the path, includes: When inserting the item sequence as a path into the FP-tree, for each node on the path, the component corresponding to the student's group in the node's group count vector is incremented by 1; at the same time, the semester information corresponding to the student's grade record is added to the semester set of the node. If the semester information already exists, it is not added again.

5. A system applied to the student grade report data management method according to claims 1-4, characterized in that, Includes the following modules: The classification module is used to divide students into different student groups based on historical academic data and course selection behavior; convert students' course scores into discrete grade level items; and calculate and obtain the item ranking index corresponding to each student group based on the statistical distribution characteristics of each grade level item within the student group. The Initialize FP Tree module is used to initialize the FP tree. The nodes of the FP tree contain an item name, a group count vector for recording the count of each student group, and a semester set for recording the semester information of the path where the node is located. The update module is used to traverse each student grade record, sort the grade level items in the student grade record in descending order according to the item sorting index of the student group to which the student grade record belongs, insert the descending sorted item sequence as a path into the FP tree, and update the group count vector and semester set of each node on the path. The mining module is used to obtain the population count vector and semester set of the terminal node corresponding to each prefix path when constructing the conditional pattern base for the items in the item header table. It also determines whether the population count vector meets the preset population dominance threshold and whether the semester set meets the preset semester distribution constraint. Only the prefix paths that meet the conditions are retained to generate the conditional FP tree, and frequent patterns including population characteristics and time series information are obtained through recursive mining.

6. The student grade report data management system according to claim 5, characterized in that, The method of dividing students into different student groups based on historical academic data and course selection behavior includes: The average grade point average of each student is extracted as a feature of historical academic data. The number of science, engineering, humanities, and social science courses taken is counted as a feature of course selection behavior. After vectorizing the features of historical academic data and course selection behavior, the K-means clustering algorithm is used to divide all students into K student groups, where K is a preset integer and K is greater than or equal to 2.

7. The student grade report data management system according to claim 5, characterized in that, For each student group, based on the statistical distribution characteristics of each grade level item within the student group, the item ranking index corresponding to the student group is calculated and obtained, including: For each student group, the total number of times each grade level item appears within that student group is counted, and this total number of times is used as the ranking index for the grade level item within the student group.

8. The student grade report data management system according to claim 5, characterized in that, The step of inserting the descendingly ordered sequence of items into the FP-tree as a path, and updating the population count vector and semester set of each node on the path, includes: When inserting the item sequence as a path into the FP-tree, for each node on the path, the component corresponding to the student's group in the node's group count vector is incremented by 1; at the same time, the semester information corresponding to the student's grade record is added to the semester set of the node. If the semester information already exists, it is not added again.

Images