Learning path planning method, system, device and storage medium

Abstract

The application provides a learning path planning method, system, device and storage medium. The method comprises: obtaining learning behavior data of a target user; determining a user characteristic portrait of the target user based on a clustering result of the learning behavior data; and performing benefit calculation on historical learning behavior data of each similar user to obtain a target learning path of the target user, wherein the similar user is selected from each user with the same user characteristic portrait. The application performs portrait modeling according to the learning behavior data of the target user, and then performs benefit calculation on the historical learning behavior data of the similar user corresponding to the user characteristic portrait to obtain the target learning path with the maximum benefit, without combining the relevance between knowledge points, thereby effectively improving the accuracy of learning path planning.

Description

Technical Field

[0001] This invention relates to the field of educational technology, and in particular to a learning path planning method, system, device, and storage medium. Background Technology

[0002] With the widespread application of the Internet, traditional education methods are gradually shifting towards online education. Due to the large number of learners in the online education model, it is particularly important to plan learning paths based on individual learning situations to improve personal learning abilities.

[0003] Current learning path planning schemes primarily use knowledge graphs as the foundation for path planning. Based on assessments and diagnostics during the course learning process, they determine the student's mastery of each knowledge point, and then combine the preorder, successor, and correlation relationships between knowledge points to derive the user's learning path. However, when faced with discrete knowledge points and when the scope of knowledge points is boundless, the structure of the knowledge graph corresponding to each knowledge point becomes quite scattered, leading to lower accuracy in the recommended learning path. Summary of the Invention

[0004] This invention provides a learning path planning method, system, device, and storage medium, aiming to improve the accuracy of learning path planning.

[0005] This invention provides a learning path planning method, comprising:

[0006] Acquire learning behavior data of target users;

[0007] Based on the clustering results of the learning behavior data, the user feature profile of the target user is determined;

[0008] The target learning path of the target user is obtained by calculating the revenue from the historical learning behavior data of each similar user, wherein the similar users are selected from users with the same user feature profile.

[0009] According to a learning path planning method provided by the present invention, determining the user feature profile of the target user based on the clustering results of the learning behavior data includes:

[0010] Multidimensional learning feature data is extracted from the learning behavior data;

[0011] Cluster analysis is performed on the multidimensional learning feature data to obtain clustering results;

[0012] Based on the clustering results, a user feature profile of the target user is generated.

[0013] According to a learning path planning method provided by the present invention, the step of calculating the benefit from the historical learning behavior data of various similar users to obtain the target learning path of the target user includes:

[0014] Based on the learning behavior data, the set of practice questions for the target user is calculated, and based on the historical learning behavior data of each of the similar users, the set of historical questions for each of the similar users is calculated.

[0015] Based on the practice question set of the target user and the historical question set of each similar user, several candidate knowledge point sets are determined;

[0016] Determine the learning benefit score corresponding to each of the candidate knowledge point sets, and determine the optimal knowledge point set based on each of the learning benefit scores;

[0017] Based on the optimal set of knowledge points, a target learning path is formed for the target user.

[0018] According to a learning path planning method provided by the present invention, the step of forming a target learning path for the target user based on the optimal set of knowledge points includes:

[0019] Based on the learning behavior data of the target user, determine the current mastery level of the target user for each knowledge point in the optimal knowledge point set;

[0020] Based on the historical learning behavior data of each similar user, the average mastery level of each knowledge point in the optimal knowledge point set is determined.

[0021] Based on the current mastery level and average mastery level of each knowledge point, the degree of dispersion of each knowledge point is determined.

[0022] Based on the degree of dispersion of each knowledge point, the target learning path is formed.

[0023] According to a learning path planning method provided by the present invention, the step of determining a plurality of candidate knowledge point sets based on the practice question set of the target user and the historical question sets of each similar user includes:

[0024] Based on the set of practice questions and the historical question sets of each of the similar users, several sets of questions to be improved are determined;

[0025] Each set of questions to be improved is processed by integrating knowledge points to obtain a set of candidate knowledge points.

[0026] According to a learning path planning method provided by the present invention, after determining the user feature profile of the target user based on the clustering results of the learning behavior data, the method further includes:

[0027] Based on the proportion of diagnostic questions corresponding to the user feature profile, a number of test questions are selected from a preset set of practice questions.

[0028] Based on the evaluation results corresponding to each of the test questions, the user profile of the target user is adjusted.

[0029] According to a learning path planning method provided by the present invention, the similar users are selected based on the following steps:

[0030] Based on the learning behavior data and the historical learning behavior data corresponding to each user, the interest similarity between each user and the target user is determined.

[0031] The interest similarity scores are sorted to select similar users based on the sorting results.

[0032] The present invention also provides a learning path planning system, comprising:

[0033] The acquisition module is used to acquire learning behavior data of the target users;

[0034] The determination module is used to determine the user feature profile of the target user based on the clustering results of the learning behavior data;

[0035] The calculation module is used to calculate the benefits from the historical learning behavior data of each similar user to obtain the target learning path of the target user, wherein the similar users are selected from users with the same user feature profile.

[0036] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the learning path planning method as described above.

[0037] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the learning path planning method as described above.

[0038] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the learning path planning method as described above.

[0039] The learning path planning method, system, device, and storage medium provided by this invention create a profile model based on the learning behavior data of the target user, and then calculate the benefit of the historical learning behavior data of similar users corresponding to the user feature profile to obtain the target learning path with the greatest benefit. This effectively improves the accuracy of learning path planning without having to consider the correlation between knowledge points. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced one by one below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating the learning path planning method provided by the present invention;

[0042] Figure 2 This is a schematic diagram of the learning behavior data provided in an embodiment of the present invention;

[0043] Figure 3 This is the second flowchart illustrating the learning path planning method provided by this invention;

[0044] Figure 4 This is the third flowchart illustrating the learning path planning method provided by this invention;

[0045] Figure 5 This is a schematic diagram of the knowledge point recommendation process provided in an embodiment of the present invention;

[0046] Figure 6 This is a schematic diagram of the learning path planning system provided by the present invention;

[0047] Figure 7 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0049] The terminology used in one or more embodiments of the present invention is for the purpose of describing particular embodiments only and is not intended to limit the scope of the invention. The singular forms “a,” “the,” and “the” used in one or more embodiments of the invention are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” used in one or more embodiments of the invention refers to and includes any or all possible combinations of one or more associated listed items.

[0050] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of the present invention, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of the present invention, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when" or "when".

[0051] Figure 1 This is one of the flowcharts illustrating the learning path planning method provided by this invention. For example... Figure 1 As shown, this learning path planning method includes:

[0052] Step 11: Obtain learning behavior data of the target users;

[0053] It should be noted that, Figure 2 This is a schematic diagram of the learning behavior data provided in the embodiments of the present invention, such as... Figure 2 As shown, the learning behavior data includes basic attribute information such as name, age, and major. Optionally, users need to fill in their basic attribute information when registering and logging into the system. In addition, the learning behavior data also includes learning behavior attribute information such as system access time, system access frequency, video viewing records, courseware viewing records, attendance, student grades, and homework submission status. Optionally, based on information such as student grades and homework submission status, a set of practice questions that the target user has already practiced can be statistically determined. The learning behavior data may also include information such as background knowledge mastery and user personality. Background knowledge mastery and user personality can be obtained by the target user completing relevant questionnaires.

[0054] Step 12: Based on the clustering results of the learning behavior data, determine the user feature profile of the target user;

[0055] It should be noted that basic attribute information such as age and major can directly yield corresponding tags. For example, age tags include those born in the 1990s, 1995s, and 2000s, while major tags include computer science, automation, mechatronics, and foreign languages. Based on the target user's age and major, various tags can be directly obtained. Therefore, there is no need to perform cluster analysis on the basic attribute information.

[0056] Specifically, in this embodiment, multi-dimensional learning feature data is extracted from the learning behavior data. This multi-dimensional learning feature data represents characteristic data that reflects the learning ability of the target user. For example, it extracts learning behavior attribute information such as system access time, system access frequency, video viewing records, courseware viewing records, attendance, student grades, and homework submission status. Further, cluster analysis is performed on the multi-dimensional learning feature data to obtain clustering results. These clustering results include multi-dimensional learning feature labels. Based on these multi-dimensional learning feature labels, a user profile of the target user is determined. For example, multi-dimensional learning feature labels include, but are not limited to, learning attitude labels and learning ability labels. When the learning attitude label is "proactive" and the learning ability label is "excellent," the user profile of the target user can be determined to be a profile of excellent learners.

[0057] Step 13: Calculate the benefit based on the historical learning behavior data of each similar user to obtain the target learning path of the target user. The similar users are selected from users with the same user feature profile.

[0058] Specifically, based on users with the same user feature profile, multiple users are randomly selected as similar users. Alternatively, the historical learning behavior data of each user and the learning behavior data of the target user can be combined to calculate the similarity between each user and the target user, thereby selecting similar users based on the similarity. In this embodiment, to improve the accuracy of learning path planning, the number of similar users can be set to multiple. Further, based on the practice question set of the target user and the historical question set of each of the similar users, several candidate knowledge point sets are calculated, and then the optimal knowledge point set is selected from each candidate knowledge point set. In one embodiment, the specific selection process is as follows: the benefit is calculated for each candidate knowledge point set, and the candidate knowledge point set with the highest learning benefit is selected as the optimal knowledge point set. In another embodiment, the scores of the questions corresponding to each knowledge point can also be determined based on the learning behavior data of the target user, thereby forming the optimal knowledge point set from several knowledge points with lower scores.

[0059] Furthermore, based on each knowledge point in the optimal knowledge point set, a target learning path is formed for the target user. To improve the accuracy of the learning path planning, more specifically, the mastery level of the target user and each similar user for each knowledge point in the optimal knowledge point set is statistically analyzed. Then, based on the mastery level of each similar user for each knowledge point in the optimal knowledge point set, the average mastery level for each knowledge point is determined. Based on the average mastery level of each knowledge point and the mastery level of the target user for each knowledge point in the optimal knowledge point set, the deviation degree of each knowledge point is calculated. Finally, based on the deviation degree of each knowledge point, a target learning path for the target user is formed.

[0060] This invention provides a model for profiling target users based on their learning behavior data. Then, it calculates the benefits of historical learning behavior data of similar users corresponding to the user's feature profile to obtain the target learning path with the greatest benefit. This approach does not require considering the correlation between knowledge points and effectively improves the accuracy of learning path planning.

[0061] Figure 3 This is the second flowchart illustrating the learning path planning method provided by this invention, as shown below. Figure 3 As shown, in one embodiment of the present invention, step 12 above: determining the user feature profile of the target user based on the clustering results of the learning behavior data, includes:

[0062] Step 121: Extract multidimensional learning feature data from the learning behavior data;

[0063] It should be noted that the multidimensional learning feature data includes system access frequency, access time, video click and viewing records, courseware resource viewing records, homework submission status, grades, knowledge point mastery level, number of posts, learning personality, and other multidimensional feature data.

[0064] Step 122: Perform cluster analysis on the multidimensional learning feature data to obtain clustering results;

[0065] It should be noted that the clustering analysis includes algorithms such as k-means clustering and Mini-Batch K-Means clustering. Furthermore, the clustering results include multi-dimensional learning feature labels. Specifically, the various feature data in the multi-dimensional learning feature data are clustered to obtain multi-dimensional user labels. Understandably, based on system access frequency and access time, learning behavior labels for the target user can be obtained. Optionally, these learning behavior labels include initiative, laziness, and mediocrity. Based on assignment submission status, grades, and knowledge point mastery level, learning attitude labels for the target user can be obtained. Optionally, these learning attitude labels include proactive labels and half-hearted abandonment labels. Based on knowledge point mastery level, learning intention labels can be obtained. In addition, based on basic attributes such as age, gender, and education level, basic image labels for the target user can be obtained.

[0066] Step 123: Based on the clustering results, generate a user feature profile of the target user.

[0067] Specifically, based on the multi-dimensional learning feature labels in the aggregation results, a user feature profile of the target user is determined. The more dimensions of the labels, the more detailed and comprehensive the learning profile of the target user is.

[0068] It should be noted that user profile characteristics may include: average learning; excellent learning; average learning now and expected to maintain average learning in the future; excellent learning now and expected to maintain excellent learning in the future; average learning now and expected to be excellent learning in the future; difficult learning now and average learning in the future; average learning now and unknown learning in the future; difficult learning now and excellent learning in the future; excellent learning now and average learning in the future; difficult learning now and unknown learning in the future.

[0069] The embodiments of the present invention achieve clustering of multi-dimensional learning feature data through the above scheme, thereby comprehensively and accurately depicting the user feature profile of the target user based on the multi-dimensional user tags after clustering.

[0070] In one embodiment of the present invention, after determining the user feature profile of the target user based on the clustering results of the learning behavior data, the method further includes:

[0071] Based on the proportion of diagnostic questions corresponding to the user feature profile, several test questions are selected from a preset set of practice questions; based on the evaluation results corresponding to each of the test questions, the user feature profile of the target user is adjusted.

[0072] It should be noted that different types of user profiles can recommend different proportions of learning assessment exercises. Specifically, determining the learning user category to which a user profile belongs can be understood as follows: user profiles corresponding to "currently average, future average," "currently average, future excellent," "currently struggling, future average," and "currently average, future unknown" are categorized as "potential learners"; user profiles corresponding to "currently excellent, future excellent," and "currently struggling, future excellent" are categorized as "high-performing learners"; and user profiles corresponding to "currently excellent, future average," "currently struggling, future unknown" are categorized as "struggling learners."

[0073] Furthermore, to improve the accuracy of user profiles depicting target users, this embodiment allows for adjustments to the user profiles. Specifically, test questions are selected from the preset practice question set based on the proportion of diagnostic questions corresponding to the learning user category. Optionally, the proportion of diagnostic questions includes the ratio of easy questions, standard questions, and extended questions, with a higher proportion of standard questions for high-potential learners, a higher proportion of extended questions for high-achieving learners, and a higher proportion of easy questions for struggling learners. It should be noted that this proportion of diagnostic questions is not fixed and is dynamically adjusted based on the target user's assessment results. For example, if the target user has a high accuracy rate on easy questions, the proportion of easy questions is reduced, while the proportion of standard and extended questions is increased. Additionally, during the selection of test questions from the preset practice question set, the benefit score for the target user for answering the test questions correctly is determined according to a benefit model, as follows:

[0074] R Y (x|θ,l)=f(x|θ,l)R Y (x|θ,l)+(1-f(x|θ,l))R Y (θ,[l,(x,0)])-R Y (θ,l)

[0075] Where Y represents the benefit score, θ represents the user profile, l represents the proportion of diagnostic questions, and x represents the test question. The next test question is selected from a preset set of practice questions to maximize the benefit score Y. Therefore, based on the evaluation results corresponding to each test question, the user profile of the target user can be adjusted. For example, if the user profile indicates excellent current learning and continued excellent future learning, and the target user has a high accuracy rate on the extended questions, then there is no need to adjust the target user's user profile. Conversely, if the user profile indicates difficulty in current learning and average future learning, but the target user has a high accuracy rate on the extended questions, then the target user's user profile will be adjusted to average current learning and excellent future learning.

[0076] Furthermore, if there is no learning behavior data for the target user, making it impossible to create a user profile, several test questions can be selected from a pre-configured set of practice questions according to a pre-set default question ratio. The process of selecting test questions is essentially the same as the process described above for selecting the next test question from the pre-configured set to maximize the benefit score Y. Based on the evaluation results corresponding to each test question, the user profile corresponding to the target user can then be determined. This eliminates the need for a large amount of answer data and a long period of learning data to create a user profile for the target user.

[0077] This invention improves the accuracy of user profiles by selecting test questions from a preset set of practice questions based on the proportion of diagnostic questions corresponding to user profiles, and then adjusting the user profiles of the target users based on the evaluation results. Furthermore, by adjusting the user profiles, adaptive planning of learning paths is achieved.

[0078] In one embodiment of the present invention, the similar users are selected based on the following steps:

[0079] Based on the learning behavior data and the historical learning behavior data corresponding to each user, the interest similarity between each user and the target user is determined; the interest similarity is sorted so as to select the similar users according to the sorting results.

[0080] Specifically, firstly, based on the user feature profile of the target user, all users identical to the user feature profile are retrieved. Further, for any user's historical learning behavior data, the following steps are performed: Based on the user's historical learning behavior data and the target user's learning behavior data, the interest similarity between the target user and the target user is calculated. Optionally, the intersection and union of the target user's practice question set and the historical question sets of similar users are calculated to obtain the interest similarity. The interest similarity calculation formula is as follows:

[0081]

[0082] Furthermore, all interest similarities are sorted to obtain a sorting result. Optionally, based on the sorting result, the user with the highest interest similarity is selected as a similar user. Preferably, in order to improve the accuracy of subsequent learning path planning, a preset number of users with high interest similarity are selected as similar users. The preset number can be set according to the actual situation and is not specifically limited here.

[0083] This invention provides an embodiment of the invention that determines the interest similarity between the target user and each user to obtain similar users. This allows for the calculation of benefits based on the historical learning behavior data of similar users, thereby obtaining the optimal target learning path.

[0084] Figure 4 This is the third flowchart of the learning path planning method provided by the present invention, as shown below. Figure 4 As shown, in one embodiment of the present invention, step 13 above: calculating the benefit from the historical learning behavior data of each similar user to obtain the target learning path of the target user, includes:

[0085] Step 131: Based on the learning behavior data, compile a set of practice questions for the target user, and based on the historical learning behavior data of each similar user, compile a set of historical questions for each similar user.

[0086] Specifically, based on information such as courseware viewing records, student grades, and homework submission status in the learning behavior data, a set of practice questions for the target user is obtained, and based on the historical learning behavior data, a set of historical questions corresponding to each of the similar users is obtained.

[0087] Step 132: Based on the practice question set of the target user and the historical question set of each similar user, determine several candidate knowledge point sets;

[0088] Specifically, for any set of historical questions from similar users, the following steps are performed:

[0089] The difference between the historical question set of the similar users and the practice question set of the target user is calculated to obtain the question set that the target user has not yet practiced. Furthermore, the knowledge points of each question in the question set that has not been practiced are integrated to obtain the candidate knowledge point set that the target user needs to improve.

[0090] Step 133: Determine the learning benefit score corresponding to each of the candidate knowledge point sets, so as to determine the optimal knowledge point set based on each of the learning benefit scores;

[0091] It should be noted that learning different knowledge points yields different benefits for students. Specifically, the following steps are performed for any set of candidate knowledge points:

[0092] Based on the target user's learning behavior data, the target user's mastery level for each knowledge point in the candidate knowledge point set is determined, as well as the average mastery level for each knowledge point in the candidate knowledge point set. The average mastery level for each knowledge point is determined based on the mastery levels of similar users for that knowledge point. In other embodiments, the average mastery level for a knowledge point can also be determined based on the mastery levels of all users in the system for that knowledge point. Understandably, the target user's mastery level for each question's knowledge point can be statistically determined based on behavioral attribute information such as grades and assignment submissions in the target user's learning behavior data.

[0093] Furthermore, based on the average mastery level of each knowledge point in the candidate knowledge point set and the target user's mastery level of each knowledge point, the learning benefit score corresponding to the candidate knowledge point set is determined. The learning benefit algorithm is as follows:

[0094]

[0095] Where KAP represents the learning achievement score, n represents the number of knowledge points, and e ki Indicates the level of mastery, e qi This represents the average level of mastery. Furthermore, based on the learning benefit scores corresponding to each candidate knowledge point set, the candidate knowledge point set with the highest learning benefit score is selected as the optimal knowledge point set.

[0096] Step 134: Based on the optimal set of knowledge points, form the target learning path for the target user.

[0097] Specifically, based on the target user's learning behavior data, the current mastery level of the target user for each knowledge point in the optimal knowledge point set is statistically analyzed; based on the historical learning behavior data of similar users, the average mastery level for each knowledge point in the optimal knowledge point set is statistically analyzed; then, based on the current mastery level and the average mastery level of any knowledge point in the optimal knowledge point set, the dispersion of any knowledge point in the optimal knowledge point set is calculated, thereby forming the target learning path based on the dispersion of each knowledge point. In other embodiments, the target learning path can also be formed based on the difference between the current mastery level and the average mastery level of any knowledge point in the optimal knowledge point set, thereby forming the difference of each knowledge point.

[0098] This invention provides an optimal set of knowledge points that maximizes learning benefits by calculating the benefits of each set of candidate knowledge points. Based on this optimal set of knowledge points, a target learning path is formed, thereby improving the accuracy of learning path planning.

[0099] In one embodiment of the present invention, forming the target learning path for the target user based on the optimal set of knowledge points includes:

[0100] Based on the learning behavior data of the target user, determine the current mastery level of the target user for each knowledge point in the optimal knowledge point set; based on the historical learning behavior data of each similar user, determine the average mastery level for each knowledge point in the optimal knowledge point set; based on the current mastery level and average mastery level of each knowledge point, determine the dispersion of each knowledge point; based on the dispersion of each knowledge point, form the target learning path.

[0101] Specifically, Figure 5 This is a schematic diagram of the knowledge point recommendation process provided in the embodiments of the present invention, such as... Figure 5 As shown, based on behavioral attribute information such as scores and homework submissions in the learning behavior data of the target user, the current mastery level of the target user for each knowledge point in the optimal knowledge point set can be statistically calculated. For example, if the target user has a high practice score for all questions corresponding to a certain knowledge point, the average of the practice scores for all questions corresponding to that knowledge point can be used as the current mastery level of that knowledge point. Multiple score thresholds can also be preset, where each score threshold corresponds to a mastery level. The average of the practice scores for all questions corresponding to that knowledge point is compared with multiple score thresholds, so that the mastery level that meets the score threshold is used as the current mastery level of the knowledge point.

[0102] Furthermore, for any similar user, the following steps are performed: Based on behavioral attribute information such as grades and assignment submissions in the similar user's historical learning behavior data, the similar user's mastery level for each knowledge point in the optimal knowledge point set can be statistically analyzed. Further, for any knowledge point in the optimal knowledge point set: Based on the mastery level of each similar user for that knowledge point, the average mastery level of that knowledge point is calculated; then, based on the current mastery level and the average mastery level corresponding to that knowledge point, the dispersion of that knowledge point is calculated.

[0103] Furthermore, the discreteness of each knowledge point is sorted to obtain a sorting result. Based on the sorting result, the knowledge points are divided into "priority learning", "delayed learning" and "advanced learning", thereby forming recommended knowledge points and obtaining the target learning path.

[0104] This invention improves the accuracy of learning path planning by calculating the degree of dispersion based on the target user's mastery of each knowledge point and the average mastery of each knowledge point.

[0105] In one embodiment of the present invention, determining a plurality of candidate knowledge point sets based on the practice question set of the target user and the historical question sets of each of the similar users includes:

[0106] Based on the set of practice questions and the historical question sets of each of the similar users, several sets of questions to be improved are determined; knowledge point integration processing is performed on each set of questions to be improved to obtain each set of candidate knowledge points.

[0107] Specifically, for any set of historical questions from similar users, the following steps are performed:

[0108] Based on the historical question set of similar users and the practice question set of the target user, questions that the target user has not yet practiced relative to the historical question set of similar users can be identified. That is, questions that do not overlap with the practice question set of the target user are selected from the historical question set. Then, it is determined which questions the similar users practiced and which improved their abilities. This yields the target user's set of questions to be improved. The formula for calculating the set of questions to be improved is as follows: O = VU, where O represents the set of questions to be improved, V represents the historical question set, and U represents the practice question set. It should be noted that each practice question has its own knowledge point. By integrating the knowledge points of each question in the set of questions to be improved, a set of candidate knowledge points that the target user needs to improve is obtained.

[0109] This invention determines the set of questions that the target user needs to improve by comparing the set of historical questions from similar users with the set of practice questions from the target user. Then, based on the set of candidate knowledge points, it determines the set of candidate knowledge points and accurately calculates the learning path corresponding to the optimal knowledge point based on each knowledge point in the set of candidate knowledge points.

[0110] The learning path planning system provided by this invention is described below. The learning path planning system described below can be referred to in correspondence with the learning path planning method described above.

[0111] Figure 6 This is a schematic diagram of the learning path planning system provided by the present invention, as shown below. Figure 6 As shown in the figure, an embodiment of the present invention provides a learning path planning system, which includes:

[0112] Module 61 is used to acquire learning behavior data of the target user;

[0113] The determination module 62 is used to determine the user feature profile of the target user based on the clustering results of the learning behavior data;

[0114] The calculation module 63 is used to calculate the benefit based on the historical learning behavior data of each similar user to obtain the target learning path of the target user, wherein the similar users are selected from users with the same user feature profile.

[0115] The determining module 62 is further configured to:

[0116] Multidimensional learning feature data is extracted from the learning behavior data;

[0117] Cluster analysis is performed on the multidimensional learning feature data to obtain clustering results;

[0118] Based on the clustering results, a user feature profile of the target user is generated.

[0119] The computing module 63 is also used for:

[0120] Based on the learning behavior data, the set of practice questions for the target user is calculated, and based on the historical learning behavior data of each of the similar users, the set of historical questions for each of the similar users is calculated.

[0121] Based on the practice question set of the target user and the historical question set of each similar user, several candidate knowledge point sets are determined;

[0122] Determine the learning benefit score corresponding to each of the candidate knowledge point sets, and determine the optimal knowledge point set based on each of the learning benefit scores;

[0123] Based on the optimal set of knowledge points, a target learning path is formed for the target user.

[0124] The computing module 63 is also used for:

[0125] Based on the learning behavior data of the target user, determine the current mastery level of the target user for each knowledge point in the optimal knowledge point set;

[0126] Based on the historical learning behavior data of each similar user, the average mastery level of each knowledge point in the optimal knowledge point set is determined.

[0127] Based on the current mastery level and average mastery level of each knowledge point, the degree of dispersion of each knowledge point is determined.

[0128] Based on the degree of dispersion of each knowledge point, the target learning path is formed.

[0129] The computing module 63 is also used for:

[0130] Based on the set of practice questions and the historical question sets of each of the similar users, several sets of questions to be improved are determined;

[0131] Each set of questions to be improved is processed by integrating knowledge points to obtain a set of candidate knowledge points.

[0132] The learning path planning system also includes:

[0133] Based on the proportion of diagnostic questions corresponding to the user feature profile, a number of test questions are selected from a preset set of practice questions.

[0134] Based on the evaluation results corresponding to each of the test questions, the user profile of the target user is adjusted.

[0135] The learning path planning system also includes:

[0136] Based on the learning behavior data and the historical learning behavior data corresponding to each user, the interest similarity between each user and the target user is determined.

[0137] The interest similarity scores are sorted to select similar users based on the sorting results.

[0138] It should be noted that the system provided in this embodiment of the invention can implement all the method steps implemented in the above method embodiment and can achieve the same technical effect. Therefore, the parts that are the same as those in the method embodiment and the beneficial effects will not be described in detail here.

[0139] Figure 7 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 7 As shown, the electronic device may include a processor 710, a memory 720, a communications interface 730, and a communications bus 740, wherein the processor 710, memory 720, and communications interface 730 communicate with each other via the communications bus 740. The processor 710 can call logical instructions in the memory 720 to execute a learning path planning method, which includes: acquiring learning behavior data of a target user; determining a user feature profile of the target user based on the clustering results of the learning behavior data; and calculating the benefit of historical learning behavior data of similar users to obtain the target learning path of the target user, wherein the similar users are selected from users with the same user feature profile.

[0140] Furthermore, the logical instructions in the aforementioned memory 720 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0141] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program implements the learning path planning method provided by the above methods. The method includes: acquiring learning behavior data of a target user; determining a user feature profile of the target user based on the clustering results of the learning behavior data; and calculating the benefits of historical learning behavior data of each similar user to obtain the target learning path of the target user, wherein the similar users are selected from users with the same user feature profile.

[0142] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the learning path planning method provided by the above methods. The method includes: acquiring learning behavior data of a target user; determining a user feature profile of the target user based on the clustering results of the learning behavior data; and calculating the benefits of historical learning behavior data of each similar user to obtain the target learning path of the target user, wherein the similar users are selected from users with the same user feature profile.

[0143] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0144] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0145] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A learning path planning method characterized by, The method comprises the following steps: obtaining learning behavior data of a target user; determining a user feature portrait of the target user based on a clustering result of the learning behavior data; calculating the learning behavior data of each similar user to obtain a target learning path of the target user, wherein the similar user is selected from each user with the same user feature portrait as the target user; wherein the calculation of the learning behavior data of each similar user to obtain the target learning path of the target user comprises: based on the learning behavior data, statistics of the target user's exercise question set, and based on the historical learning behavior data of each similar user, statistics of the historical question set of each similar user; based on the target user's exercise question set and each similar user's historical question set, determine a plurality of candidate knowledge point sets; based on the current mastery level and the average mastery level of each knowledge point in each candidate knowledge point set, determine the learning benefit score corresponding to each candidate knowledge point set, and determine the optimal knowledge point set according to each learning benefit score, wherein the current mastery level of any knowledge point is determined based on the learning behavior data of the target user, and the average mastery level of any knowledge point is determined based on the mastery level of each similar user or all users in the system for the knowledge point; based on the current mastery level and the average mastery level of each knowledge point in the optimal knowledge point set, determine the dispersion degree of each knowledge point in the optimal knowledge point set; based on the dispersion degree of each knowledge point in the optimal knowledge point set, form the target learning path.

2. The learning path planning method of claim 1, wherein, The determination of the user feature portrait of the target user based on the clustering result of the learning behavior data comprises: extracting multi-dimensional learning feature data from the learning behavior data; performing clustering analysis on the multi-dimensional learning feature data to obtain a clustering result; generating the user feature portrait of the target user based on the clustering result.

3. The learning path planning method of claim 1, wherein, The determination of the dispersion degree of each knowledge point in the optimal knowledge point set based on the current mastery level and the average mastery level of each knowledge point in the optimal knowledge point set comprises: determining the current mastery level of each knowledge point in the optimal knowledge point set for the target user based on the learning behavior data of the target user; determining the average mastery level of each knowledge point in the optimal knowledge point set based on the historical learning behavior data of each similar user; determining the dispersion degree of each knowledge point based on the current mastery level and the average mastery level of each knowledge point.

4. The learning path planning method of claim 1, wherein, The determination of a plurality of candidate knowledge point sets based on the target user's exercise question set and each similar user's historical question set comprises: determining a plurality of to-be-improved question sets based on the exercise question set and each similar user's historical question set; performing knowledge point integration processing on each to-be-improved question set to obtain each candidate knowledge point set.

5. The learning path planning method of claim 1, wherein, After determining the user feature portrait of the target user based on the clustering result of the learning behavior data, the method further includes: selecting a plurality of test questions from a pre-set exercise question set based on the proportion of diagnostic questions corresponding to the user feature portrait; adjusting the user feature portrait of the target user based on the evaluation results respectively corresponding to each of the test questions.

6. The learning path planning method of claim 1, wherein, The similar users are selected based on the following steps: determining the interest similarity between each of the users and the target user based on the learning behavior data and the historical learning behavior data respectively corresponding to each of the users; sorting the interest similarities to select each of the similar users according to the sorting result.

7. A learning path planning system characterized by comprising: The method includes: an acquisition module configured to acquire learning behavior data of a target user; a determination module configured to determine a user feature portrait of the target user based on a clustering result of the learning behavior data; a calculation module configured to perform benefit calculation on historical learning behavior data of each of the similar users to obtain a target learning path of the target user, wherein the similar users are selected from each of the users having the same user feature portrait as the target user; The calculation module is specifically configured to: statistically determine an exercise question set of the target user based on the learning behavior data, and statistically determine a historical question set of each of the similar users based on the historical learning behavior data of each of the similar users; determine a plurality of candidate knowledge point sets based on the exercise question set of the target user and the historical question set of each of the similar users; determine a learning benefit score corresponding to each of the candidate knowledge point sets based on a current mastery level and an average mastery level respectively corresponding to each of the knowledge points in each of the candidate knowledge point sets, and determine an optimal knowledge point set based on the learning benefit scores, wherein the current mastery level of any knowledge point is determined based on the learning behavior data of the target user, and the average mastery level of the any knowledge point is determined based on the mastery levels of the any knowledge point of each of the similar users or all users in the system; determine the dispersion degree of each of the knowledge points in the optimal knowledge point set based on the current mastery level and the average mastery level respectively corresponding to each of the knowledge points in the optimal knowledge point set; form the target learning path based on the dispersion degree of each of the knowledge points in the optimal knowledge point set.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor implements the learning path planning method according to any one of claims 1 to 6 when executing the program. 9.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program implements the learning path planning method according to any one of claims 1 to 6 when executed by the processor.

Images