Online interaction method and system

Abstract

The invention belongs to the technical field of data processing, and particularly relates to an online interaction method and system, and the method comprises the steps: constructing a recommended learning path through employing an ant colony algorithm according to a course knowledge graph, obtaining the learning target heuristic, value heuristic, difficulty heuristic and interest heuristic of a course in each iteration process of the ant colony algorithm, and obtaining a recommended learning path; constructing a heuristic function of the course; according to the number of courses contained in each path in the last round of iteration result of the ant colony algorithm and the standard learning duration of the courses, pheromones of edges between every two courses are obtained; and determining the selection probability of each selectable course according to the pheromone of the edge between the course corresponding to the current node and each selectable course and the heuristic function of each selectable course, and selecting the selectable course as the next node of the path according to the selection probability. According to the method, personalized learning path recommendation is realized, and the learning efficiency and knowledge mastering effect of students can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing. More specifically, the present invention relates to an online interaction method and system. Background Art

[0002] With the continuous advancement of educational informatization, personalized learning path recommendation technology has become an important means to improve teaching efficiency and meet the diverse needs of students. At present, existing research has attempted to combine knowledge modeling and recommendation algorithms to provide customized learning paths for students. However, most methods still have obvious limitations.

[0003] Existing methods mostly rely on simple rule matching or linear sorting strategies to generate learning paths, lacking in-depth modeling of the complex semantic relationships between courses and knowledge points, and it is difficult to accurately reflect the overall structure of the knowledge system. This results in the recommended learning paths often ignoring the prerequisite knowledge requirements, causing cognitive breaks in the learning process and affecting the students' understanding depth and learning coherence.

[0004] In addition, existing methods often have a weak description of individual student differences. Usually, they only make rough recommendations based on interest tags, lacking comprehensive analysis of multi-dimensional information such as students' learning behavior characteristics and mastery of knowledge points, and it is difficult to achieve true "teaching students in accordance with their aptitude". At the same time, some recommendation mechanisms do not have the ability to dynamically adjust, and cannot optimize the path in real time according to the students' learning progress, reducing the adaptability and practicality of the recommendation results.

[0005] Therefore, there is an urgent need to introduce a more intelligent and efficient path generation method that can deeply integrate the knowledge structure and student characteristics and has the ability to dynamically respond to significantly improve the accuracy and practicality of personalized learning path recommendations. Summary of the Invention

[0006] To solve the technical problems that the existing methods are difficult to accurately reflect the overall structure of the knowledge system, difficult to achieve teaching students in accordance with their aptitude, and do not have the ability to dynamically adjust, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides an online interaction method, including: Construct a curriculum knowledge graph; according to the curriculum knowledge graph, use the ant colony algorithm to construct a recommended learning path; visualize the recommended learning path in the curriculum knowledge graph to guide students to learn in sequence; wherein, in each round of iteration of the ant colony algorithm: obtain learning objective inspiration according to the importance of the knowledge points included in the curriculum and the relevance between the students' learning objectives and the fields to which the knowledge points belong; obtain value inspiration according to the students' mastery of the learned knowledge points and the relevance between the learned knowledge points and the knowledge points included in the curriculum; obtain difficulty inspiration according to the difficulty of each knowledge point included in the curriculum; obtain interest inspiration according to the students' preference indexes for each resource type; construct an inspiration function for the curriculum according to the learning objective inspiration, value inspiration, difficulty inspiration, and interest inspiration; obtain the pheromone of the edges between all courses pairwise according to the number of courses included in each path and the standard learning duration of the courses in the result of the previous round of iteration of the ant colony algorithm; use the courses that meet the preconditions, have not been learned by the students and have not been added to the path yet as optional courses, and determine the selection probabilities of the optional courses according to the pheromone of the edges between the course corresponding to the current node and the optional courses and the inspiration functions of the optional courses, and select an optional course as the next node of the path according to the selection probabilities.

[0008] The present invention realizes the semantic modeling of courses, knowledge points and their mutual relationships by constructing a structured curriculum knowledge graph, making the knowledge system clearer and computable. On this basis, combined with multi-dimensional information such as students' learning behaviors, learning achievements, mastery levels and interest preferences, it can accurately identify their learning objectives and weak links, providing comprehensive data support for personalized path recommendation. The present invention uses the ant colony algorithm for learning path planning, fully considering factors such as the logical coherence of the path, learning efficiency, difficulty adaptability and students' interests. It can dynamically generate high-quality and personalized learning paths on the premise of meeting the pre-dependency relationships, with good adaptability and search capabilities, improving the accuracy and practicality of the recommendation results, and solving problems such as path breaks and cognitive breaks in traditional methods. The present invention dynamically adjusts the recommended path according to the students' learning progress, enhancing the flexibility and interactivity of the recommendation mechanism and improving the intelligent level of learning path recommendation.

[0009] Preferably, the obtaining of the learning objective inspiration includes: performing text preprocessing on the students' learning objectives, and converting the preprocessed text into a number of word vectors; converting the fields to which each knowledge point included in the curriculum belongs into word vectors; calculating the average value of the cosine similarities between the word vectors corresponding to the learning objectives and the word vectors corresponding to the fields to which the knowledge points belong as the relevance between the learning objectives and the fields to which the knowledge points belong; and performing weighted summation on the relevance between the learning objectives and the fields to which the knowledge points belong according to the importance of the knowledge points included in the curriculum to obtain the learning objective inspiration of the curriculum.

[0010] By vectorizing the semantic information of learning objectives and course knowledge points and calculating the cosine similarity between the two, the present invention can effectively measure the matching degree between learning objectives and course content in the semantic space. On this basis, the relevance results are weighted and summed in combination with the importance of knowledge points, making the learning objective inspiration of the course more in line with the actual knowledge structure and students' needs, thereby improving the accuracy and personalization of the recommended path and helping to guide students to preferentially select courses that highly match their own learning objectives.

[0011] Preferably, the value inspiration satisfies the expression: ; where represents the value inspiration of course K; represents the th knowledge point in course K; represents the th learned knowledge point; represents the relevance between the th learned knowledge point and the th knowledge point in course K; represents the number of knowledge points included in course K; represents the number of learned knowledge points; represents the mastery level of the student for the th learned knowledge point.

[0012] By comprehensively considering the relevance between the learned knowledge points and each knowledge point in the candidate courses and the student's mastery level of the learned knowledge points, the present invention can effectively measure the value of the course for the improvement of the student's knowledge system. Among them, the knowledge points with a lower mastery level correspond to a higher learning demand weight, and the courses to which the knowledge points with high relevance belong will obtain a higher value inspiration, thus encouraging students to preferentially learn the content closely related to the weak knowledge points, achieving the purpose of strengthening understanding through repeated learning and gradually making up for the knowledge short board, and improving the pertinence and personalization level of path recommendation.

[0013] Preferably, the difficulty inspiration satisfies the expression: ; where represents the difficulty inspiration of course K; represents the relevance between the th learned knowledge point and the th knowledge point in course K; represents the number of knowledge points included in course K; represents the number of learned knowledge points; represents the mastery level of the student for the th learned knowledge point; represents the The difficulty of each knowledge point; is the maximum value function.

[0014] The present invention dynamically evaluates the actual learning difficulty of a course for a student by comprehensively considering the inherent difficulty of each knowledge point in the course and the student's mastery of relevant previously learned knowledge points. Among them, for each knowledge point, the most relevant mastered knowledge point is selected and weighted adjustment is performed in combination with its mastery level, so that the difficulty evaluation is more in line with the student's actual cognitive level, and courses that are too difficult for the student and lack sufficient prerequisite support can be reasonably avoided in the recommended path, thereby improving the adaptability and acceptability of the learning path and helping the student to gradually complete knowledge construction.

[0015] Preferably, the obtaining of interest inspiration includes: taking the resource type of the course as the target resource type and taking the preference index of the student for the target resource type as the interest inspiration of the course.

[0016] Preferably, the construction of the inspiration function of the course includes: performing weighted summation on the normalized result of the learning target inspiration, the normalized result of the value inspiration, the normalized result of the interest inspiration, and the negatively correlated normalized result of the difficulty inspiration to obtain the inspiration function of the course.

[0017] By normalizing the learning target inspiration, value inspiration, interest inspiration, and difficulty inspiration and performing weighted fusion according to the set weights, the present invention constructs a comprehensive inspiration function, which can realize the organic integration of multi-dimensional information, taking into account the matching degree between the course and the learning target, the supplementary value for knowledge weak points, the student's interest preference, and the learning difficulty of the course. The finally formed inspiration function more comprehensively and accurately reflects the adaptability degree of the course to the student's learning path, providing a scientific and personalized decision-making basis for the recommendation system.

[0018] Preferably, the method for obtaining the pheromone of the edge between two courses includes: obtaining the quality score of the path according to the number of courses included in the path and the standard learning duration of the course; for any two courses, if the two courses are adjacent nodes in a certain path of the previous iteration result, then taking the corresponding path as the reference path for the two courses; updating the pheromone of the edge between the two courses according to the quality scores of all the reference paths of the two courses: , represents the course and the course the pheromone of the edge between; represents the pheromone of the edge between the course and the course in the previous iteration; represents the course and the course The average quality score of all reference paths therebetween; Denote the pheromone evaporation factor.

[0019] By comprehensively considering the number of courses and the standard learning duration in a path, calculating the quality score of the path, and dynamically adjusting the pheromone intensity of the edge based on the co-occurrence of adjacent courses in the historical path, the present invention can effectively reflect the value of the transfer between different courses, improve the recognition ability of efficient learning paths, enhance the stability and adaptability of path recommendation, and help guide students to construct a more reasonable and coherent learning sequence.

[0020] Preferably, the method for obtaining the mastery level of the knowledge points learned by the student includes: for any learned knowledge point, obtaining the average value of the simulation scores corresponding to all simulation operation tasks related to the knowledge point by the student, normalizing the average value of the simulation scores to obtain the operation proficiency of the student for the knowledge point; using the correct rate of the knowledge point by the student in the periodic tests and examinations as the theoretical proficiency of the student for the knowledge point, and performing weighted average on the operation proficiency and the theoretical proficiency to obtain the mastery level of the student for the knowledge point.

[0021] Preferably, the method for obtaining the preference index of the student for each resource type includes: for any resource type, taking the learning duration, access times, task completion rate, collection times, viewing times, and simulation operation times of the student in all courses of the resource type as a judgment dimension for the resource type respectively, normalizing each judgment dimension respectively, and performing weighted summation on the normalized results of each judgment dimension to obtain the preference index of the student for the resource type.

[0022] In a second aspect, the present invention provides an online interaction system, including a processor and a memory, where the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned online interaction method is implemented.

[0023] By adopting the above technical solution, generating a computer program for the above-mentioned online interaction method and storing it in the memory to be loaded and executed by the processor, so as to manufacture a terminal device according to the memory and the processor, which is convenient to use.

[0024] The beneficial effects of the present invention are as follows: The present invention realizes the semantic modeling of courses, knowledge points and their interrelationships by constructing a structured curriculum knowledge graph, making the knowledge system clearer and computable. On this basis, by combining multi-dimensional information such as students' learning behaviors, academic achievements, mastery levels and interest preferences, it can accurately identify their learning goals and weak links, providing comprehensive data support for personalized path recommendation. The present invention uses the ant colony algorithm for learning path planning, fully considering factors such as the logical coherence of the path, learning efficiency, difficulty adaptability and students' interests. It can dynamically generate high-quality and personalized learning paths while satisfying the prerequisite dependency relationships, with good self-adaptability and search capabilities, improving the accuracy and practicality of the recommendation results, and solving problems such as path breaks and cognitive breaks in traditional methods. The present invention dynamically adjusts the recommended path according to the students' learning progress, enhancing the flexibility and interactivity of the recommendation mechanism and improving the intelligent level of learning path recommendation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a flowchart schematically showing an online interaction method in the present invention; Figure 2 is a flowchart schematically showing the construction of the path in each round of the ant colony algorithm. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0026] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0027] The following will describe the specific embodiments of the present invention in detail in conjunction with the accompanying drawings.

[0028] An embodiment of the present invention discloses an online interaction method, referring to Figure 1 , including steps S1 - S4: S1. Construct a curriculum knowledge graph.

[0029] Specifically, collect data from multiple sources such as teaching syllabuses, textbooks, teaching objectives, teaching contents, and past examination papers, and perform text preprocessing on the original data, including operations such as word segmentation, part-of-speech tagging, text cleaning, and entity standardization. Based on natural language processing technology, identify core entities such as "courses" and "knowledge points" from them, and further extract the semantic relationships (such as prerequisite relationships, inclusion relationships, correlation relationships, etc.) between the entities and the attribute information of each entity.

[0030] The attributes of a course include: the field it belongs to, the resource type (such as text, video, audio, simulation), the standard learning duration, prerequisite courses, and the knowledge points it contains, etc.; the attributes of knowledge points include: difficulty, the field it belongs to, the course it belongs to, importance, and the relevance to other knowledge points, etc.

[0031] Among them, the difficulty of a knowledge point is quantitatively obtained based on the average correct rate of the knowledge point in previous years' test papers. The lower the correct rate, the higher the difficulty, and the value range is [0, 1]; the importance of a knowledge point is evaluated by the frequency of its appearance in the teaching syllabus and the examination syllabus. The higher the appearance frequency, the higher the importance, and the value range is [0, 1]; the relevance between knowledge points is automatically extracted through co-occurrence analysis and semantic similarity models, and then manually corrected and supplemented by educational experts to ensure the accuracy of the logical relationship between knowledge, and the value range is [0, 1].

[0032] Finally, the extracted entities, relationships, and their attributes are organized in the form of a graph structure to construct a course knowledge graph.

[0033] S2. Collect the learning behaviors, learning achievements, and learning goals of students. Obtain the mastery level of students for the learned knowledge points based on their learning achievements, and construct a preference index for students for each resource type according to their learning behaviors.

[0034] Specifically, for any student, collect learning information in multiple dimensions such as the student's learning behaviors, learning achievements, and learning goals.

[0035] Learning behaviors mainly include the learning duration, access frequency, task completion rate, and interaction behaviors (such as collection, repeated viewing, number of simulation operations) of students in each course, etc.

[0036] Learning achievements include two categories: theoretical achievements and simulation achievements. Theoretical achievements come from the periodic tests and examinations corresponding to courses with resource types of text, video, or audio, and include the correct rates of students for each knowledge point in the periodic tests and examinations; simulation achievements come from the simulation operation tasks set in courses with resource types of simulation.

[0037] Learning goals refer to the degree of knowledge mastery or the direction of skill improvement that students expect to achieve through learning, which are actively set by students. For example, "be familiar with the functions and usage methods of control cores such as PLC and single-chip microcomputers", "master the installation, debugging, and data acquisition methods of various sensors", "understand the basic structure and working principle of control systems", etc.

[0038] Furthermore, by analyzing the students' theoretical scores and simulation scores, the mastery level of each learned knowledge point by the students is comprehensively calculated to measure the students' understanding and application level of the knowledge point. By statistically analyzing the learning duration distribution, access frequency, task completion rate, and interaction behaviors (such as favorites, repeated viewings, and the number of simulation operations) of students in courses of different resource types (graphic, video, audio, simulation), a preference index of students for each resource type is constructed to reflect the students' preference for courses of different resource types. For example, if a student actively completes simulation tasks multiple times and stays for a long time, it can be judged that the student is more inclined to practical learning; if the student frequently watches video courses, it indicates that the student prefers audio-visual resources.

[0039] In one embodiment, the method for obtaining the mastery level of each learned knowledge point by the student includes: for any learned knowledge point, obtaining the mean value of the simulation scores corresponding to all simulation operation tasks related to the knowledge point by the student, normalizing the mean value to obtain the operation proficiency of the student for the knowledge point, using the correct rate of the knowledge point by the student in the periodic tests and exams as the theoretical proficiency of the student for the knowledge point, and performing a weighted average of the operation proficiency and the theoretical proficiency to obtain the mastery level of the student for the knowledge point. When performing the weighted average of the operation proficiency and the theoretical proficiency, the weights of the operation proficiency and the theoretical proficiency are set by the implementer according to the nature of the course. For example, if the course emphasizes theoretical understanding, the weight of the operation proficiency can be set to 0.4 and the weight of the theoretical proficiency can be set to 0.6; if the course emphasizes practical ability, the weight of the operation proficiency can be set to 0.6 and the weight of the theoretical proficiency can be set to 0.4. It should be noted that to ensure the rationality of the weighted average, the sum of the weights of the operation proficiency and the theoretical proficiency needs to be 1.

[0040] It should be noted that when the student does not perform the simulation operation tasks related to the knowledge point, the simulation score of the student under the corresponding simulation operation task is set to 0; when there are no simulation operation tasks related to the knowledge point, the theoretical proficiency of the student for the knowledge point is used as the mastery level of the student for the knowledge point; when the student does not participate in the periodic tests and exams, or the content of the periodic tests and exams participated in does not include the knowledge point, the theoretical proficiency of the student for the knowledge point is set to 0; when the content of all periodic tests and exams of the course does not include the knowledge point, the operation proficiency of the student for the knowledge point is used as the mastery level of the student for the knowledge point. When there are no simulation operation tasks related to the knowledge point and the content of all periodic tests and exams of the course does not include the knowledge point, the mastery level of the student for the knowledge point is set to 0.

[0041] In one embodiment, the construction of the preference index of students for each resource type includes: For any resource type, the learning duration, access times, task completion rate, collection times, viewing times, and simulation operation times of the student in all courses of this resource type are respectively used as a judgment dimension of this resource type. Each judgment dimension is normalized separately, and the normalized results of each judgment dimension are weighted and summed to obtain the preference index of the student for this resource type. Among them, the separate normalization of each judgment dimension includes: for any judgment dimension, dividing the data of this judgment dimension of this resource type by the sum of the data of this judgment dimension of all resource types to realize the normalization of the data of this judgment dimension of this resource type.

[0042] It should be noted that when the normalized results of each judgment dimension are weighted and summed in this embodiment, the weights of each judgment dimension are the same, all being , represents the number of judgment dimensions. In other embodiments, the implementer can also set the weights of each judgment dimension according to the actual implementation situation.

[0043] S3. According to the learning goals of the student, the student's mastery of the learned knowledge points, the student's preference index for each resource type, and the course knowledge graph, use the ant colony algorithm to construct a recommended learning path.

[0044] Specifically, in each round of iteration of the ant colony algorithm, the last course learned by the student is used as the starting node, and the construction of each round of path starts from the starting node. In response to the iteration round of the ant colony algorithm reaching the maximum number of iterations or the optimal paths obtained in consecutive H rounds of iteration being the same, stop the ant colony algorithm, remove the starting node from the finally obtained optimal path, and the remaining path is used as the recommended learning path. Among them is a preset quantity. In one embodiment, is set to 5, the maximum number of iterations in the ant colony algorithm is 100, the number of ants is 50, the pheromone factor is 1, the heuristic function factor is 1, the pheromone evaporation factor is 1, and the initial pheromone of the edge between any two courses is 1. In other embodiments, the implementer can set according to the actual implementation situation.

[0045] It should be noted that in the present invention, the last course learned by the student is used as the starting node, and the construction of each round of path starts from the starting node, which can ensure the continuity and logic of the learning path, ensure that the newly recommended courses have a good connection with the student's existing learning foundation in terms of knowledge structure, thereby improving the rationality and effectiveness of personalized recommendation. Removing the starting node from the finally obtained optimal path can avoid repeated recommendation of the learned content and ensure that the output learning path only contains course recommendation results that have not been learned but are highly matched with the learning goals.

[0046] Further, the flowchart of each round of path construction refers to Figure 2 , including steps S301 to S308, specifically: S301. Obtain the learning objective inspiration of the course according to the importance of the knowledge points included in the course and the relevance between the learning objective and the field to which the knowledge points included in the course belong.

[0047] S302. Obtain the value inspiration of the course according to the student's mastery of the learned knowledge points and the relevance between the learned knowledge points and the knowledge points included in the course.

[0048] S303. Obtain the difficulty inspiration of the course according to the student's mastery of the learned knowledge points, the relevance between the learned knowledge points and the knowledge points included in the course, and the difficulty of each knowledge point in the course.

[0049] S304. Obtain the interest inspiration of the course according to the resource type of the course and the student's preference index for each resource type.

[0050] S305. Construct the inspiration function of the course according to the learning objective inspiration, value inspiration, difficulty inspiration, and interest inspiration of the course.

[0051] S306. Obtain the pheromone of the edges between all courses pairwise according to the number of courses included in each path and the standard learning duration of the course in the previous iteration result of the ant colony algorithm.

[0052] S307. Take the courses that meet the preconditions, have not been learned by the student, and have not been added to the path yet as optional courses. Determine the selection probability of each optional course according to the pheromone of the edges between the course corresponding to the current node and each optional course and the inspiration function of each optional course. Select an optional course as the next node of the path according to the selection probability.

[0053] S308. Repeat step S307 until the current iteration of the ant colony algorithm stops when the stop condition is met, and the construction of the current round of path is realized.

[0054] In one embodiment, obtaining the learning objective inspiration of the course according to the importance of the knowledge points included in the course and the relevance between the learning objective and the field to which the knowledge points included in the course belong in step S301 includes: Preprocess the learning objectives of students, including word segmentation and stop word removal. Convert the preprocessed text into a number of word vectors using a natural language processing model (Natural Language Processing, NLP). Convert the fields to which each knowledge point included in the course belongs into word vectors using the NLP model. Calculate the mean of the cosine similarities between the word vectors corresponding to the learning objectives and the word vectors corresponding to the fields to which the knowledge points belong as the relevance between the learning objectives and the fields to which the knowledge points belong. Determine the learning objective inspiration of the course based on the importance of the knowledge points included in the course and the relevance between the learning objectives and the fields to which the knowledge points belong: ; Among them, represents the learning objective inspiration of course K; represents the importance of the th knowledge point in course K; represents the number of knowledge points included in course K; represents the learning objective of the student, represents the th knowledge point in course K; represents the relevance between the learning objective of the student and the field to which the th knowledge point in course K belongs. Since the relevance between the learning objective and the field to which the knowledge point belongs is obtained based on the mean of the cosine similarities, and the value range of the cosine similarity is [-1, 1], the relevance also has a value range of [-1, 1]. Therefore, in this embodiment, is used to normalize the relevance . represents the importance weight of the th knowledge point in course K. When the relevance between the learning objective of the student and the field to which the more important knowledge point in course K belongs is greater, the learning objective inspiration of course K is greater, and course K should be given higher priority for the student to learn.

[0055] It should be noted that the NLP model used to obtain word vectors in this embodiment is the Word2Vec model. In other embodiments, the implementer can select the NLP model according to the actual implementation situation, such as BERT.

[0056] In one embodiment, in step S302, according to the student's mastery of the learned knowledge points and the relevance between the learned knowledge points and the knowledge points included in the course, obtain the value inspiration of the course, including: The value inspiration satisfies the expression: ; Among them, Represents the value inspiration of course K; Represents the th knowledge point in course K; Represents the th learned knowledge point; Represents the th learned knowledge point and the th knowledge point in course K; Represents the number of knowledge points included in course K; Represents the number of learned knowledge points; Represents the mastery level of the th learned knowledge point by the student; when the mastery level of the th learned knowledge point is smaller, the student pays more attention to the course to which the related knowledge point belongs, so as to encourage the student to learn the course related to the weak knowledge point and improve the mastery level of the knowledge point through repeated learning of the related knowledge point.

[0057] In one embodiment, in step S303, according to the mastery level of the learned knowledge points by the student, the relevance between the learned knowledge points and the knowledge points included in the course, and the difficulty of each knowledge point in the course, the difficulty inspiration of the course is obtained, including: The difficulty inspiration satisfies the expression: ; Where Represents the difficulty inspiration of course K; Represents the th learned knowledge point and the th knowledge point in course K; Represents the number of knowledge points included in course K; Represents the number of learned knowledge points; Represents the mastery level of the th learned knowledge point by the student; Represents the th knowledge point in course K; Is the maximum value function; when the relevance between a learned knowledge point and the th knowledge point in course K is greater, and the mastery level of the learned knowledge point by the student is greater, for the student, the difficulty of learning the th knowledge point in course K is relatively reduced. Therefore, the present invention takes as the difficulty reduction coefficient of the th knowledge point in course K to obtain the actual learning difficulty of the th knowledge point in course K, and combines the actual learning difficulties of all knowledge points in course K to obtain the difficulty inspiration of course K.

[0058] In one embodiment, in step S304, obtaining the interest inspiration of the course according to the resource type of the course and the preference index of the student for each resource type includes: Taking the resource type of the course as the target resource type, and taking the preference index of the student for the target resource type as the interest inspiration of the course.

[0059] In one embodiment, in step S305, constructing the inspiration function of the course according to the learning target inspiration, value inspiration, difficulty inspiration, and interest inspiration of the course includes: The inspiration function of the course satisfies the expression: ; where, represents the inspiration function of course K; , , , respectively represent the weights of the learning target inspiration, value inspiration, difficulty inspiration, and interest inspiration; represents the learning target inspiration of course K of the normalized result; represents the value inspiration of course K of the normalized result; represents the difficulty inspiration of course K reciprocal of of the normalized result; represents the interest inspiration of course K of the normalized result. In this embodiment, the learning target inspiration of all courses that the student has not yet learned is used to perform maximum-minimum normalization on the learning target inspiration of course K, the value inspiration of all courses that the student has not yet learned is used to perform maximum-minimum normalization on the value inspiration of course K, the reciprocal of the difficulty inspiration of all courses that the student has not yet learned is used to perform maximum-minimum normalization on the reciprocal of the difficulty inspiration of course K, the interest inspiration of all courses that the student has not yet learned is used to perform maximum-minimum normalization on the interest inspiration of course K. In other embodiments, the implementer can select other normalization methods according to the actual implementation situation, such as maximum value normalization; when the learning target inspiration of course K is larger, the value inspiration is larger, the difficulty inspiration is smaller, and the interest inspiration is larger, the inspiration function of course K is larger, encouraging students to give priority to learning courses that highly match their own learning goals, help make up for weak knowledge points, have a smaller difficulty, and conform to personal interest preferences.

[0060] It should be noted that since the learning objective inspiration plays a core guiding role in path recommendation and directly affects the matching degree between the courses and the student's target knowledge system, the weight of the learning objective inspiration in this embodiment is set to 0.4, and the weights of the value inspiration , the difficulty inspiration , and the interest inspiration are respectively set to 0.2. In other embodiments, the implementer can set the weights of the learning objective inspiration, the value inspiration, the difficulty inspiration, and the interest inspiration according to the actual implementation situation.

[0061] In one embodiment, in step S306, the pheromone of the edges between all courses is obtained according to the number of courses included in each path and the standard learning duration of the courses in the previous iteration result of the ant colony algorithm, including: For any path in the previous iteration result, the quality score of the path is obtained according to the number of courses included in the path and the standard learning duration of the courses: ; wherein, represents the quality score of path ; represents the number of courses included in path ; represents the sum of the standard learning durations of all courses included in path , that is, the total standard learning duration of path ; represents the exponential function with the natural constant as the base, which is used to perform a negative correlation mapping on ; is a normalization function. In this embodiment, the number of courses included in path is normalized by the maximum and minimum values using the number of courses included in all paths in the previous iteration result, and the total standard learning duration of path is normalized by the maximum and minimum values using the total standard learning durations of each path in the previous iteration result. In other embodiments, the implementer can select other normalization methods according to the actual implementation situation, such as maximum value normalization; when the number of courses included in path is smaller, the total standard learning duration of path is smaller, the learning efficiency of path is higher, and the quality score of path is larger. is higher, and the quality score of path is larger.

[0062] For any two courses, if the two courses are adjacent nodes in a certain path of the previous iteration result, the corresponding path is used as the reference path for the two courses. Update the pheromone of the edge between the two courses according to the quality scores of all the reference paths of the two courses: ; Among them, represents the pheromone of the edge between course and course ; represents the pheromone of the edge between course and course in the previous iteration; represents the average value of the quality scores of all the reference paths between course and course . When there is no reference path between course and course , it is stipulated that is 0; represents the pheromone evaporation factor.

[0063] In one embodiment, the satisfaction of the precondition in step S307 means that all the prerequisite courses of the course have been learned or added to the path.

[0064] In one embodiment, determining the selection probabilities of the optional courses according to the pheromones of the edges between the course corresponding to the current node and the optional courses and the heuristic functions of the optional courses in step S307 includes: ; Among them, the course corresponding to the current node is represented by , represents the selection probability of taking the optional course as the next node of the current node; represents the pheromone of the edge between course and the optional course ; represents the heuristic function of the optional course ; represents the set composed of the optional courses; represents the pheromone factor; represents the heuristic function factor. When the pheromone of the edge between course and the optional course is larger, and the heuristic function of the optional course is larger, the optional course The selection probability is used to encourage students to preferentially study courses that highly match their own learning goals, help make up for weak knowledge points, are of lower difficulty, conform to personal interest preferences, and have a relatively high overall learning efficiency.

[0065] In one embodiment, in step S307, selecting an optional course as the next node of the path according to the selection probability includes: Selecting one course from all optional courses as the next node of the path according to the selection probabilities of the optional courses by means of sampling with unequal probabilities.

[0066] In one embodiment, the setting method of the stop condition in step S308 is as follows: Using the method in step S301 to obtain the relevance between the learning goals of the student and the fields to which all knowledge points belong, and regarding all knowledge points with a relevance greater than a preset relevance threshold as the knowledge points covered by the learning goals. In response to the knowledge points included in the courses in the path already covering all the knowledge points covered by the learning goals, stop constructing this path. In response to all the paths corresponding to each ant having stopped being constructed, this round of iteration ends.

[0067] Among them, the relevance threshold is set by the implementer according to the actual implementation situation. In this embodiment, the relevance threshold is 0.3.

[0068] S4. Visualize the recommended learning path in the knowledge graph to guide students to study in sequence.

[0069] Specifically, clearly identify the recommended learning path in the knowledge graph and display it to students in a visual form to help them intuitively understand the logical relationship between each course and knowledge points and the learning order between courses, and recommend that students study courses according to the recommended learning path to ensure the gradual and systematic mastery of knowledge, thereby improving learning efficiency and effect.

[0070] An embodiment of the present invention also discloses an online interaction system, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, an online interaction method according to the present invention is implemented.

[0071] The above system also includes other components well known to those skilled in the art such as a communication bus and a communication interface, and their settings and functions are known in the art, so they will not be described in detail here.

Claims

1. An online interaction method, characterized in that, Including: Constructing a curriculum knowledge graph; Based on the curriculum knowledge graph, using the ant colony algorithm to construct a recommended learning path; Visualizing the recommended learning path in the curriculum knowledge graph to guide students to learn in sequence; Among them, in each round of iteration of the ant colony algorithm: Obtain learning objective inspiration according to the importance of the knowledge points included in the curriculum and the relevance between the student's learning objectives and the fields to which the knowledge points belong; obtain value inspiration according to the student's mastery of the learned knowledge points and the relevance between the learned knowledge points and the knowledge points included in the curriculum; obtain difficulty inspiration according to the difficulty of each knowledge point included in the curriculum; obtain interest inspiration according to the student's preference index for each resource type; construct an inspiration function for the curriculum according to the learning objective inspiration, value inspiration, difficulty inspiration, and interest inspiration; According to the number of courses included in each path and the standard learning duration of the courses in the previous iteration result of the ant colony algorithm, obtain the pheromone of the edges between all pairs of courses; regard the courses that meet the preconditions, have not been learned by the student, and have not been added to the path currently as optional courses, and determine the selection probability of each optional course according to the pheromone of the edges between the course corresponding to the current node and each optional course and the inspiration function of each optional course, and select an optional course as the next node of the path according to the selection probability.

2. The online interaction method according to claim 1, characterized in that The obtaining of the learning objective inspiration includes: Performing text preprocessing on the student's learning objectives, and converting the preprocessed text into several word vectors; converting the fields to which each knowledge point included in the curriculum belongs into word vectors; calculating the mean value of the cosine similarities between the word vectors corresponding to the learning objectives and the word vectors corresponding to the fields to which the knowledge points belong, as the relevance between the learning objectives and the fields to which the knowledge points belong; performing weighted summation on the relevance between the learning objectives and the fields to which the knowledge points belong according to the importance of the knowledge points included in the curriculum, to obtain the learning objective inspiration of the curriculum.

3. An online interaction method according to claim 1, characterized in that, The value inspiration satisfies the expression: ； Among them, represents the value inspiration of course K; represents the th knowledge point in course K; represents the th knowledge point that has been learned; represents the th knowledge point that has been learned and the th knowledge point in course K; represents the number of knowledge points included in course K; represents the number of knowledge points that have been learned; represents the student's mastery level of the th knowledge point that has been learned.

4. An online interaction method according to claim 1, characterized in that The difficulty inspiration satisfies the expression: ； Among them, represents the difficulty inspiration of course K; represents the th learned knowledge point and the th knowledge point in course K; represents the number of knowledge points included in course K; represents the number of learned knowledge points; represents the student's mastery level of the th learned knowledge point; represents the th knowledge point's difficulty in course K; is the maximum value function.

5. The online interaction method according to claim 1, characterized in that The obtaining of the interest inspiration includes: Regarding the resource type of the curriculum as the target resource type, and taking the student's preference index for the target resource type as the interest inspiration of the curriculum.

6. The online interaction method according to claim 1, characterized in that, The constructing of the inspiration function for the curriculum includes: Performing weighted summation on the normalized result of the learning objective inspiration, the normalized result of the value inspiration, the normalized result of the interest inspiration, and the negatively correlated normalized result of the difficulty inspiration, to obtain the inspiration function of the curriculum.

7. An online interaction method according to claim 1, wherein The method for obtaining the pheromone of the edges between pairs of courses includes: Obtain the quality score of a path based on the number of courses included in the path and the standard learning duration of the courses; for any two courses, if the two courses are adjacent nodes in a certain path of the previous iteration result, then use the corresponding path as the reference path for the two courses; update the pheromone of the edge between the two courses according to the quality scores of all the reference paths of the two courses: , represents the pheromone of the edge between course and course represents the pheromone of the edge between course and course in the previous iteration; represents the average value of the quality scores of all the reference paths between course and course represents the pheromone evaporation factor.

8. An online interaction method according to claim 1 or 3 or 4, characterized in that The method for obtaining the student's mastery of the learned knowledge points includes: For any learned knowledge point, obtain the mean value of the simulation scores corresponding to all simulation operation tasks related to the knowledge point by the student, normalize the mean value of the simulation scores, to obtain the operation proficiency of the student for the knowledge point; take the correct rate of the knowledge point by the student in the stage tests and exams as the theoretical proficiency of the student for the knowledge point, and perform weighted averaging on the operation proficiency and the theoretical proficiency, to obtain the student's mastery of the knowledge point.

9. An online interaction method according to claim 1 or 5, characterized in that, The method for obtaining the student's preference index for each resource type includes: For any type of resource, the learning duration, access times, task completion rate, collection times, viewing times, and simulation operation times of the student in all courses of this resource type are respectively used as a judgment dimension for this resource type. Each judgment dimension is normalized respectively, and the normalized results of each judgment dimension are weighted and summed to obtain the preference index of the student for this resource type.

10. An online interaction system, characterized in that, Including: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an online interaction method according to any one of claims 1-9 is implemented.

Images