Multi-objective based collaborative learning grouping and learning resource recommendation method and system

Abstract

The application discloses a multi-target-based collaborative learning grouping and learning resource recommendation method and system, which comprises four steps of obtaining a data set, performing preliminary grouping by using an improved clustering grouping algorithm, grouping the remaining learners, and assigning learning resources to the formed groups for collaborative learning. Since the application adopts a multi-target-based hierarchical clustering algorithm for grouping, the course willingness, learning behavior habit and cognitive state of the group members are considered. Similar course willingness and learning behavior habit can improve the learning efficiency of the collaborative learning group members, and the knowledge points as much as possible can help the group members to learn more knowledge and skills. The collaborative learning group formed by the grouping method is more efficient. The application has the advantages of high grouping efficiency, large grouping benefit, wide application range and the like, and can be used for learning resource recommendation, such as course recommendation, and can also be used for grouping activities in other fields.

Description

Technical Field

[0001] This invention belongs to the field of intelligent education technology, specifically relating to a method and system for collaborative learning grouping and learning resource recommendation based on multiple objectives. Background Technology

[0002] With the outbreak of the COVID-19 pandemic, online learning has become the choice for more and more people. Online learning involves not only individual learning but also group learning. Group learning helps learners consolidate and expand their skills while collaborating to discuss and solve problems and efficiently complete group tasks. This is the process of computer-supported collaborative learning (CSCL). Collaborative learning refers to learners engaging in various collaborative learning activities without the limitations of time and space. It not only allows them to fully exercise their existing knowledge but also enables them to learn unfamiliar knowledge from group members; this process is called collaborative learning.

[0003] Many factors influence the efficiency of collaborative learning, such as the composition of the group and the learning tasks. When group members share the same task motivation, they will have more consistent learning motivation and interest, significantly improving efficiency. Learners typically have different learning habits and styles; for example, some prefer watching videos, others prefer online forums, and still others prefer completing exercises to fill in knowledge gaps. When group members share similar learning habits, collaborative learning becomes more efficient. In collaborative learning, it's beneficial to ensure that group members possess diverse knowledge and skills. This allows learners to consolidate their own knowledge and skills while learning from others, broadening their knowledge base. Currently, existing collaborative learning grouping methods do not simultaneously consider the three factors of group members' course motivation, learning habits, and cognitive state, hindering the efficient completion of collaborative learning tasks.

[0004] Therefore, in the research on collaborative learning grouping, this invention proposes a multi-objective clustering grouping method. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a multi-objective collaborative learning grouping method that simultaneously considers three factors: the group members' course willingness, learning behavior habits, and cognitive state. Using an improved clustering grouping method, it ensures the similarity and differences among collaborative learning group members, thereby improving collaborative learning efficiency. This multi-objective collaborative learning grouping and learning resource recommendation method is highly efficient, yields significant benefits, and has a wide range of applications.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a multi-objective collaborative learning grouping and learning resource recommendation method, comprising the following steps:

[0007] S1 collects all learner IDs, learner course intentions, exercise IDs, knowledge point IDs contained in the exercises, and learning behavior records to obtain learner course intention vectors, learner learning behavior habit vectors, and learner cognitive state vectors; the learner set is represented as S = {s1, s2, ..., s...} n ,...,s N The set of knowledge points is represented as C = {c1, c2, ..., c}. k ,...,c K The set of all courses can be represented as R = {r1, r2, ..., r}. m ,...,r M};

[0008] S2, based on the obtained learners' course willingness vector, learners' learning behavior habit vector, and learners' cognitive state vector, the learners are initially grouped using an improved clustering grouping method;

[0009] S3, grouping the remaining learners after the initial grouping in S2.

[0010] If the number of learners in the remaining nodes meets the set group size, the node is directly added to the group set; if the number of learners in the remaining nodes is less than the set group size, the learners in the node are reassigned to the nearest group, and each group can be assigned a maximum of one new node.

[0011] S4 assigns the courses with the highest "benefits" to the groups divided by S2 and / or S3 for collaborative learning, thus realizing the recommendation of learning resources.

[0012] The vectors obtained from obtaining learners' course willingness, learning behavior habits, and cognitive state include:

[0013] 1) Obtain learners' course willingness vector

[0014] Learners have a preference ranking of all courses, represented by a vector w. i =[w i1 ,w i2 ,...,w im ,...,w iM ] indicates that each value in the vector is an integer;

[0015] 2) Obtain learners' learning behavior habit vectors

[0016] The learner's learning behavior habits vector is represented as b i =[b i1 ,b i2 ,......,b ij ], where j represents the frequency of different learning behaviors and habits, b ij For s i The probability of using the j-th learning action is:

[0017] 3) Obtain the learner's cognitive state vector

[0018] The learner's cognitive state vector is represented as a i =[a i1 ,a i2 ,...,a ik ,...,a iK ], a ik (0≤a ik ≤1) indicates s i For c k The level of proficiency.

[0019] In S2, considering the course willingness of learners in the collaborative learning group, the mean of the course willingness vector of learners in the group members is used as the group's course willingness vector; the calculation method is as shown in equation (1):

[0020]

[0021] w UM Indicates node U for course r M The sorting belongs to w U The elements in.

[0022] In S2, the learning behavior habits of learners in the collaborative learning group are considered. These learning behavior habits include the number of times videos are watched, the number of times forum interactions are conducted, the number of times exercises are completed, and the number of times interactions with the teacher are performed. The mean of the learning behavior habit vectors of the learners in the group is used as the group's learning behavior habit vector. The calculation method is as shown in Equation (2):

[0023]

[0024] b Uj Let represent the probability that node U has the j-th learning behavior habit, belonging to b. U The elements in.

[0025] In S2, the cognitive state vector of learners in the collaborative learning group is considered. The maximum proficiency of learners in each knowledge point is taken as the cognitive state vector of the group. During the collaborative learning process, learners can learn different knowledge points from other group members. After a period of time, the group members' mastery of the knowledge points will approach the maximum proficiency of the knowledge points in the group. The cognitive state is calculated as shown in equation (3):

[0026] a Uk =MAX(a ik |s i ∈U) (3)

[0027] Where a Uk This indicates that node U is at knowledge point c. k Proficiency in this area belongs to type a U The elements in.

[0028] Based on the learned learners' course willingness vector, learning behavior habit vector, and cognitive state vector, when using an improved clustering grouping algorithm to initially group learners, each learner is initially regarded as a single node U, and all learners form a node set; all nodes have the following characteristics: course willingness, learning behavior habit, and cognitive state, which are represented by vectors; when other learners join the node, the node's characteristics are recalculated.

[0029] Based on the obtained learners' course willingness vector, learners' learning behavior habit vector, and learners' cognitive state vector, an improved clustering grouping algorithm is used to initially group learners, including:

[0030] 1) Calculate the distance between every two nodes.

[0031] Let d be the distance between different nodes U and V. UV The distance between two nodes is calculated based on the distances in the course willingness vector, learning behavior habit vector, and cognitive state vector, with each of these three having different weights. The calculation method is as shown in equation (4):

[0032]

[0033] Where α + β + γ = 1, α, β, and γ represent weights. and These represent the distance between two nodes in terms of course willingness, learning behavior habits, and cognitive state, respectively.

[0034] The distance between two nodes in terms of course willingness is calculated as shown in equation (5):

[0035]

[0036] In Equation (5), the reciprocal of the course willingness represents the "benefit" of the course to the learners in the node, and the normalized Euclidean distance is used. This is used to measure the distance between the course preferences of two clusters, with the distance ranging from 0 to 1;

[0037] The distance between two nodes in terms of learning behavior habits is calculated as shown in equation (6):

[0038]

[0039] Equation (6) uses cosine similarity to measure the distance between two nodes in terms of learning behavior habits, with the distance being between 0 and 1.

[0040] The distance between two nodes in terms of learning behavior habits is calculated as shown in equation (7):

[0041]

[0042] Equation (7) is similar to Equation (5);

[0043] 2) Based on the calculated distance, find the two closest nodes.

[0044] If the total number of learners in the two most recent nodes is less than the set group size, then merge these two nodes into a new node and add it to the current node set; if the sum of the number of learners in the two most recent nodes is equal to the specified group size, then merge these two nodes into a new node and treat it as a group, add it to the group set, and delete these two nodes from the current node set.

[0045] 3) Continuously execute step 2), and eventually several groups that meet the requirements will be formed; if the node set is empty, execute S4, and if it is not empty, execute S3.

[0046] In S4, the desired "gain" of group U is represented by g. U The calculation method is shown in formulas (9) and (10):

[0047]

[0048] g U =MAX(g UM |r M ∈R) (10)

[0049] g UM Indicates group U in course r M The "benefits" of the intention on the surface belong to g U elements in

[0050] A multi-objective collaborative learning grouping and learning resource recommendation system is provided, including a dataset acquisition module, a grouping module, and a resource recommendation module;

[0051] The dataset acquisition module collects all learner IDs, learner course preferences, exercise IDs, knowledge point IDs contained in the exercises, and learning behavior records, obtaining learner course preference vectors, learner learning behavior habit vectors, and learner cognitive state vectors; the learner set is represented as S = {s1, s2, ..., s...} n ,...,s N The set of knowledge points is represented as C = {c1, c2, ..., c}. k ,...,c K The set of all courses can be represented as R = {r1, r2, ..., r}. m ,...,r M};

[0052] The grouping module is used to initially group learners based on the obtained learner course willingness vector, learner learning behavior habit vector, and learner cognitive state vector using an improved clustering grouping algorithm; at the same time, it groups the remaining learners after the initial grouping; if the number of learners in the remaining nodes meets the set group size, the node is directly added to the group set; if the number of learners in the remaining nodes is less than the specified group size, the learners in the node are reassigned to the nearest group, and each group is assigned at most one new node;

[0053] The resource recommendation module is used to assign courses that offer the greatest "benefits" to groups divided into S2 and / or S3 for collaborative learning, thus realizing the recommendation of learning resources.

[0054] In addition, the present invention also provides a computer device, including a processor and a memory, wherein the memory is used to store a computer executable program, the processor reads the computer executable program from the memory and executes it, and the processor can implement the multi-objective collaborative learning grouping and learning resource recommendation method described in the present invention when executing the computer executable program.

[0055] Simultaneously, a computer-readable storage medium can be provided, in which a computer program is stored. When the computer program is executed by a processor, it can implement the multi-objective collaborative learning grouping and learning resource recommendation method described in this invention.

[0056] Compared with the prior art, the present invention has at least the following beneficial effects:

[0057] This invention considers learners' course preferences within collaborative learning groups, ensuring a more consistent course intention among group members, guaranteeing learners' interest and motivation, and improving collaborative learning efficiency. It also considers learners' learning habits, fostering similar habits among group members, facilitating unified collaborative learning behavior and promoting efficient task completion. Furthermore, it considers learners' cognitive states—their mastery of knowledge points—and encourages them to master diverse knowledge points, allowing them to consolidate their existing skills while acquiring additional knowledge and skills. Finally, this invention combines learners' course preferences, learning habits, and cognitive states, employing a bottom-up hierarchical clustering method for collaborative learning grouping and course allocation. This ensures high similarity in course preferences and learning habits, but significant differences in knowledge and skills within each group, improving collaborative learning efficiency and promoting individual member development. Attached Figure Description

[0058] Figure 1 This is a flowchart illustrating the method described in this invention.

[0059] Figure 2 This is a schematic diagram of the steps involved in a multi-objective clustering grouping method.

[0060] Figure 3 This is a schematic diagram of the invention. Detailed Implementation

[0061] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments, but the present invention is not limited to the embodiments described below.

[0062] Example 1

[0063] refer to Figure 1 and Figure 3 The multi-objective collaborative learning grouping and learning resource recommendation method in this embodiment consists of the following steps:

[0064] A multi-objective collaborative learning grouping and learning resource recommendation method includes the following steps:

[0065] (1) Preprocessed data

[0066] Collect a suitable dataset, which should include all learner IDs, learner course preferences, exercise IDs, knowledge point IDs contained in the exercises, and learning behavior records. The learner set is represented as S = {s1, s2, ..., s}. n ,...,sN The set of knowledge points is represented as C = {c1, c2, ..., c}. k ,...,c K The set of all courses can be represented as R = {r1, r2, ..., r}. m ,...,r M}

[0067] 1) Obtain learners' course willingness vector

[0068] Learners have a preference ranking of all courses, represented by a vector w. i =[w i1 ,w i2 ,...,w im ,...,w iM The symbol ] indicates that each value in the vector is an integer. For example, w il When the value is 2, it indicates that the course r m In students i It ranked second in the course preference list.

[0069] 2) Obtain learners' learning behavior habit vectors

[0070] The learner's learning behavior habits vector is represented as b i =[b i1 ,b i2 ,......,b ij ], where j represents the frequency of different learning behaviors, such as the number of times videos are watched, the number of times forum interactions are conducted, the number of times exercises are completed, and the number of times teacher interactions are performed. b ij For s i The probability of using the j-th learning action is:

[0071] 3) Obtain the learner's cognitive state vector

[0072] The learner's cognitive state vector is represented as a i =[a i1 ,a i2 ,...,a ik ,...,a iK ], a ik (0≤a ik ≤1) indicates s i For c k The level of proficiency.

[0073] (2) Initial grouping was performed using an improved clustering grouping algorithm, referring to... Figure 2 ,

[0074] 1) Initially, each learner is considered as a single node U, then all learners form a node set. All nodes have the following characteristics: course willingness, learning behavior habits, and cognitive state, and these three characteristics are represented by vectors. The course willingness vector of a node is calculated by the average of the course willingness vectors of all learners in the node, as shown in equation (1):

[0075]

[0076] w UM Indicates node U for course r M The sorting belongs to w U The elements in.

[0077] The learning behavior habit vector of a node is calculated by averaging the learning behavior habit vectors of all learners in the node, as shown in equation (2):

[0078]

[0079] b Uj Let represent the probability that node U has the j-th learning behavior habit, belonging to b. U The elements in.

[0080] During collaborative learning, learners can learn different knowledge points from other group members. After a period of time, the group members' mastery of the knowledge points will approach the maximum proficiency of that knowledge point in the group. The calculation method is as shown in formula (3).

[0081] a Uk =MAX(a ik |s i ∈U) (3)

[0082] Where a Uk This indicates that node U is at knowledge point c. k Proficiency in this area belongs to type a U The elements in the node. In subsequent steps, other learners will join the node according to the algorithm, at which point the node's features need to be recalculated.

[0083] 2) Calculate the distance between every two nodes.

[0084] Let d be the distance between different nodes U and V. UV The distance between two nodes is calculated based on the distances in the course willingness vector, learning behavior habit vector, and cognitive state vector, with each of these three having a different weight (the weights can be specified by the user). Its calculation method is shown in equation (4):

[0085]

[0086] α+β+γ=1, α, β and γ represent weights, and These represent the distance between two nodes in terms of course willingness, learning behavior habits, and cognitive state, respectively.

[0087] The calculation method for the course willingness of the two nodes is as follows (5):

[0088]

[0089] In Equation (5), the reciprocal of the course willingness represents the "benefit" of the course to the learners in the node, and the normalized Euclidean distance is used. This is used to measure the distance between the course preferences of two clusters, with the distance ranging from 0 to 1.

[0090] The calculation method for the two nodes in terms of learning behavior habits is as shown in equation (6):

[0091]

[0092] Equation (6) uses cosine similarity to measure the distance between two nodes in terms of learning behavior habits, with the distance being between 0 and 1.

[0093] The calculation method for the two nodes in terms of learning behavior habits is as shown in equation (7):

[0094]

[0095] Equation (7) is similar to Equation (5).

[0096] 3) Find the two closest nodes

[0097] If the total number of learners in the two most recent nodes is less than the specified group size, then these two nodes are merged into a new node and added to the current node set. If the sum of the number of learners in the two most recent nodes is equal to the specified group size, then these two nodes are merged into a new node and treated as a group, added to the group set, and both nodes are removed from the current node set.

[0098] 4) Repeat step 3) over and over again until several groups that meet the requirements are formed. At this point, the node set may be empty. If it is empty, step (4) will be executed; if it is not empty, step (3) will be executed.

[0099] (3) Divide the remaining learners into groups

[0100] If the number of learners in the remaining nodes meets the prescribed group size, the node is directly added to the group set; if the number of learners in the remaining nodes is less than the prescribed group size, the learners in the node are reassigned to the nearest group. In this case, to ensure that the group sizes in the final group set are relatively even, each group will be assigned at most one new node.

[0101] (4) Assign courses that would provide the greatest "benefits" to groups for collaborative learning.

[0102] The desired "gain" of group U is represented by g. U Its calculation method is shown in formulas (9) and (10):

[0103]

[0104] g U =MAX(g UM |r M ∈R) (10)

[0105] g UM Indicates group U in course r M The "benefits" of the intention on the surface belong to g U The elements in.

[0106] Develop a multi-objective collaborative learning grouping and learning resource recommendation method.

[0107] Based on the concept of the method described in this invention, a multi-objective collaborative learning grouping and learning resource recommendation system is provided, including a dataset acquisition module, a grouping module, and a resource recommendation module;

[0108] The dataset acquisition module collects all learner IDs, learner course preferences, exercise IDs, knowledge point IDs contained in the exercises, and learning behavior records, obtaining learner course preference vectors, learner learning behavior habit vectors, and learner cognitive state vectors; the learner set is represented as S = {s1, s2, ..., s...} n ,...,s N The set of knowledge points is represented as C = {c1, c2, ..., c}. k ,...,c K The set of all courses can be represented as R = {r1, r2, ..., r}. m ,...,r M};

[0109] The grouping module is used to initially group learners based on the obtained learner course willingness vector, learner learning behavior habit vector, and learner cognitive state vector using an improved clustering grouping algorithm; at the same time, it groups the remaining learners after the initial grouping; if the number of learners in the remaining nodes meets the set group size, the node is directly added to the group set; if the number of learners in the remaining nodes is less than the specified group size, the learners in the node are reassigned to the nearest group, and each group is assigned at most one new node;

[0110] The resource recommendation module is used to assign courses that offer the greatest "benefits" to groups divided into S2 and / or S3 for collaborative learning, thus realizing the recommendation of learning resources.

[0111] In addition, the present invention can also provide a computer device, including a processor and a memory, wherein the memory is used to store a computer executable program, the processor reads part or all of the computer executable program from the memory and executes it, and the processor can implement the multi-objective collaborative learning grouping and learning resource recommendation method described in the present invention when executing part or all of the computer executable program.

[0112] On the other hand, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, can implement the multi-objective collaborative learning grouping and learning resource recommendation method described in the present invention.

[0113] The computer device may be a laptop, a desktop computer, or a workstation.

[0114] The processor can be a central processing unit (CPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), or an off-the-shelf programmable gate array (FPGA).

[0115] The memory described in this invention can be an internal storage unit of a laptop, desktop computer, or workstation, such as memory or hard disk; or it can be an external storage unit, such as a portable hard disk or flash memory card.

[0116] Computer-readable storage media can include computer storage media and communication media. Computer storage media includes volatile and non-volatile, removable and non-removable media implemented using any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer-readable storage media can include: read-only memory (ROM), random access memory (RAM), solid-state drives (SSDs), or optical discs, etc. Random access memory can include resistive random access memory (ReRAM) and dynamic random access memory (DRAM).

Claims

1. A multi-objective collaborative learning grouping and learning resource recommendation method, characterized in that, Includes the following steps: S1 collects all learner IDs, learner course intentions, exercise IDs, knowledge point IDs contained in the exercises, and learning behavior records to obtain the learner course intention vector. learner's learning behavior habit vector and the learner's cognitive state vector , Indicates the frequency of different learning behaviors and habits. express right The level of proficiency; the set of learners is represented as The set of knowledge points is represented as The set of all courses is represented as ; S2, based on the acquired learners' course willingness vector, learners' learning behavior habit vector, and learners' cognitive state vector, uses an improved clustering grouping method to initially group learners; specifically including: 1) calculating the distance between every two nodes. Different nodes and nodes The distance between them is denoted as The distance between two nodes is calculated based on the distances in the course willingness vector, learning behavior habit vector, and cognitive state vector, with each of these three having different weights. The calculation method is as shown in equation (4): (4) in, , , and Represents weight, , and These represent the distance between two nodes in terms of course willingness, learning behavior habits, and cognitive state, respectively. The distance between two nodes in terms of course willingness is calculated as shown in equation (5): (5) In Equation (5), the reciprocal of the course willingness represents the "benefit" of the course to the learners in the node, and the normalized Euclidean distance is used. This is used to measure the distance between the course preferences of two clusters, with the distance ranging from 0 to 1; The distance between two nodes in terms of learning behavior habits is calculated as shown in equation (6): (6) Equation (6) uses cosine similarity to measure the distance between two nodes in terms of learning behavior habits, and the distance is between 0 and 1; The distance between two nodes in terms of learning behavior habits is calculated as shown in equation (7): (7) Equation (7) is similar to Equation (5); 2) Based on the calculated distance, find the two closest nodes. If the total number of learners in the two most recent nodes is less than the set group size, then merge these two nodes into a new node and add it to the current node set; if the sum of the number of learners in the two most recent nodes is equal to the specified group size, then merge these two nodes into a new node and treat it as a group, add it to the group set, and delete these two nodes from the current node set. 3) Continuously execute step 2), and eventually several groups that meet the requirements will be formed; if the node set is empty, execute S4, otherwise execute S3. S3, grouping the remaining learners after the initial grouping in S2. If the number of learners in the remaining nodes meets the set group size, the node is directly added to the group set; if the number of learners in the remaining nodes is less than the set group size, the learners in the node are reassigned to the nearest group, and each group can be assigned a maximum of one new node. S4 assigns the courses with the highest desired "benefits" to the groups divided in S2 and / or S3 for collaborative learning, thus achieving learning resource recommendation; the desired "benefits" of the groups are represented as... The calculation method is shown in formulas (9) and (10): (9) (10) Representing the group In the course The "benefits" of the intention are... The elements in.

2. The method for multi-objective collaborative learning grouping and learning resource recommendation according to claim 1, characterized in that, In S2, considering the course willingness of learners in the collaborative learning group, the mean of the course willingness vector of learners in the group members is used as the group's course willingness vector; the calculation method is as shown in equation (1): (1) Represents a node For the course The sorting belongs to The elements in.

3. The method for multi-objective collaborative learning grouping and learning resource recommendation according to claim 1, characterized in that, In S2, the learning behavior habits of learners in the collaborative learning group are considered. These learning behavior habits include the number of times videos are watched, the number of times forum interactions are conducted, the number of times exercises are completed, and the number of times interactions with the teacher are performed. The mean of the learning behavior habit vectors of the learners in the group is used as the group's learning behavior habit vector. The calculation method is as shown in Equation (2): (2) Represents a node Having the first The probability of a learning behavior habit belongs to The elements in.

4. The method for multi-objective collaborative learning grouping and learning resource recommendation according to claim 1, characterized in that, In S2, the cognitive state vector of learners in the collaborative learning group is considered. The maximum proficiency of learners in each knowledge point is taken as the cognitive state vector of the group. During the collaborative learning process, learners can learn different knowledge points from other group members. After a period of time, the group members' mastery of the knowledge points will approach the maximum proficiency of the knowledge points in the group. The cognitive state is calculated as shown in equation (3): (3) in Represents a node In the knowledge points Proficiency in it belongs to The elements in.

5. The method for multi-objective collaborative learning grouping and learning resource recommendation according to claim 1, characterized in that, Based on the learned learners' course willingness vector, learning behavior habit vector, and cognitive state vector, when using the improved clustering grouping algorithm to initially group learners, each learner is initially treated as a separate node. All learners form a set of nodes; all nodes have the following characteristics: course willingness, learning behavior habits, and cognitive state, which are represented by vectors; when other learners join the node, the node's characteristics are recalculated.

6. A multi-objective collaborative learning grouping and learning resource recommendation system, characterized in that, The method for implementing the multi-objective collaborative learning grouping and learning resource recommendation method as described in any one of claims 1-5 includes a dataset acquisition module, a grouping module, and a resource recommendation module. The dataset acquisition module collects all learner IDs, learner course intentions, exercise IDs, knowledge point IDs contained in the exercises, and learning behavior records, obtaining learner course intention vectors, learner learning behavior habit vectors, and learner cognitive state vectors; the learner set is represented as... The set of knowledge points is represented as The set of all courses is represented as ; The grouping module is used to initially group learners based on the obtained learner course willingness vector, learner learning behavior habit vector, and learner cognitive state vector using an improved clustering grouping algorithm; at the same time, it groups the remaining learners after the initial grouping; if the number of learners in the remaining nodes meets the set group size, the node is directly added to the group set; if the number of learners in the remaining nodes is less than the specified group size, the learners in the node are reassigned to the nearest group, and each group is assigned at most one new node; The resource recommendation module is used to assign courses with the highest "benefits" to groups divided into S2 and / or S3 for collaborative learning, thus realizing the recommendation of learning resources.

7. A computer device, characterized in that, It includes a processor and a memory, the memory being used to store a computer-executable program, the processor reading the computer-executable program from the memory and executing it, and the processor executing the computer-executable program being able to implement the multi-objective collaborative learning grouping and learning resource recommendation method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, A computer-readable storage medium stores a computer program that, when executed by a processor, enables the implementation of the multi-objective collaborative learning grouping and learning resource recommendation method as described in any one of claims 1 to 5.

Images