A knowledge graph-based adaptive learning content generation method and system

Abstract

The application provides a knowledge graph-based adaptive learning content generation method and system, which comprises the following steps: constructing a course knowledge graph; obtaining the cognitive mastery degree of a learner on a knowledge point in a current course; obtaining a cognitive development knowledge cluster of the learner at a current stage based on the cognitive mastery degree, a teaching organization layer and a knowledge representation layer; generating a cognitive difference subgraph based on the cognitive development knowledge cluster at the current stage and the dependency relationship between each entity in the course knowledge graph; retrieving associated resources in a resource aggregation layer in the course knowledge graph based on the knowledge points included in the cognitive difference subgraph, and mapping the associated resources and the aggregation relationship between the resources to the cognitive difference subgraph to generate adaptive learning content for the learner at the current stage. The application generates dynamic adaptive learning content containing logical relationship dependencies, solves the problems of adaptive learning content reasoning and dynamic generation in online learning, and provides a basis for the implementation of an adaptive learning system.

Description

Technical Field

[0001] This invention relates to the field of computer application technology, and in particular to an adaptive learning content generation method and system based on knowledge graphs. Background Technology

[0002] With the rapid development of internet technology, online educational resources have exploded, profoundly impacting the education field. However, faced with the abundance and disorganization of online resources, learners often encounter problems such as "knowledge disorientation" and "information overload," leading to poor learning outcomes and low efficiency. Therefore, personalized adaptive learning, acquired in a non-linear manner, has received widespread attention both domestically and internationally, becoming a new paradigm and important proposition for future educational development.

[0003] Most existing adaptive learning methods focus on adaptive learning path planning and adaptive sorting of learning objects. They are mostly based on fixed learning objects such as resources and knowledge points, and lack methods for generating adaptive learning content.

[0004] However, learning is a dynamic and evolving process. If learning content is generated based solely on a fixed learning object, it will be impossible to dynamically generate learning content that conforms to teaching and cognitive principles based on changes in the learner's cognitive state. This not only results in the generated learning content not conforming to the learner's current cognition, but also makes it impossible to fill in the gaps in knowledge that have not been mastered. Summary of the Invention

[0005] This invention provides a knowledge graph-based adaptive learning content generation method and system to address the problem that existing methods lack adaptive learning content generation methods.

[0006] On the one hand, this invention provides an adaptive learning content generation method based on knowledge graphs, the method comprising: The knowledge cluster entities, knowledge point entities, and resource entities of the course are obtained, and a course knowledge graph is constructed. The course knowledge graph includes a teaching organization layer based on knowledge cluster entities, a knowledge representation layer based on knowledge point entities, and a resource aggregation layer based on resource entities. Obtain the learner's cognitive mastery of the knowledge points in the current course; based on the cognitive mastery, and according to the teaching organization layer and the knowledge representation layer, obtain the learner's cognitive development knowledge cluster at the current stage; Based on the current cognitive development knowledge cluster and the dependencies between various entities in the course knowledge graph, a cognitive difference subgraph is generated according to the full dependency reasoning rule and the pruning processing rule. Based on the knowledge points included in the cognitive difference subgraph, related resources in the resource aggregation layer of the course knowledge graph are retrieved, and the related resources and the aggregation relationships between resources are mapped to the cognitive difference subgraph to generate adaptive learning content for the learner's current stage.

[0007] Furthermore, the method for constructing the course knowledge graph includes: Based on the IKR three-layer course knowledge graph model, the three-layer entities of knowledge cluster entities, knowledge point entities, and resource entities, as well as the inter-layer and intra-layer relationships, are extracted. Then, according to the formal representation model of the three-layer entities and inter-layer and intra-layer relationships in the IKR three-layer course knowledge graph model, they are represented in the form of tuples to construct the course knowledge graph.

[0008] Furthermore, the aforementioned cognitive development knowledge cluster The process of obtaining it includes: S201: Determine whether the learner's cognitive mastery of the knowledge points has been updated; if not, use the previously obtained cognitive development knowledge cluster as the current cognitive development knowledge cluster. Conversely, if the condition is not met, then execute S202. S202: Determine whether the current learner has a stored record of a cognitive development knowledge cluster; if so, use the stored cognitive development knowledge cluster as the cognitive development knowledge cluster for the current stage. Conversely, the initial knowledge cluster of the current course is retrieved from the course knowledge graph. and the initial knowledge cluster As a cluster of knowledge for cognitive development at the current stage ; S203: Retrieve the course knowledge graph to obtain the cognitive development knowledge clusters for the current stage. ; S204: Acquiring Knowledge Clusters for Cognitive Development The set of knowledge points contained in the set and the degree of cognitive mastery of the knowledge points contained in the set of knowledge points are used to obtain the corresponding cognitive development knowledge cluster. Cognitive mastery level; S205: Assessing knowledge clusters in cognitive development If the level of cognitive mastery exceeds a set threshold, then the cognitive development knowledge cluster for the next stage of learning is inferred from the course knowledge graph and used as the cognitive development knowledge cluster for the current stage. Conversely, directly output the knowledge cluster of cognitive development at the current stage. .

[0009] Furthermore, in S203, the course knowledge graph is retrieved to obtain the cognitive development knowledge clusters for the current stage. ,include: If the conditions are met: If so, then execute S204; Conversely, through Acquire cognitive development knowledge cluster Non-dependent knowledge clusters in subclass hierarchical knowledge clusters and will not depend on knowledge clusters As a cluster of knowledge for cognitive development at the current stage ; in, This represents the knowledge clusters of the teaching organization layer; Representing the order relation triple ; Represents a hierarchical triplet .

[0010] Furthermore, the corresponding knowledge clusters for cognitive development Cognitive mastery The calculation methods include: ; in, Represents a set of knowledge points. ; Indicates the learner's understanding of knowledge points The level of cognitive mastery; Set a threshold to indicate that the learner has mastered the knowledge point. ; This indicates that learners are in the knowledge cluster of cognitive development. Of the knowledge points included, those that have already been mastered The number of; This indicates that learners are in the knowledge cluster of cognitive development. The number of all knowledge points contained therein.

[0011] Furthermore, the cognitive development knowledge clusters for the next stage of learning are inferred from the aforementioned course knowledge graph. And as a knowledge cluster for cognitive development at the current stage. ,include: S301: If ,but The next stage of learning will focus on the cognitive development knowledge clusters. As a cluster of knowledge for cognitive development at the current stage If the condition is not met, then execute S203; otherwise, execute S302. S302: If Then the knowledge clusters in the knowledge graph As a cluster of knowledge for cognitive development at the current stage If the program returns to its previous state, it will execute S301; otherwise, it indicates that the course has been completed and the course learning has ended.

[0012] Furthermore, the generation step of the cognitive difference subgraph includes: Create an empty subgraph as the initial cognitive difference subgraph; Retrieve the course knowledge graph, obtain the fully dependent entity set based on the fully dependent reasoning rule, and add the entity set to the initial cognitive difference subgraph to obtain the first cognitive difference subgraph; Based on the entity set, the dependencies between all entities in the entity set in the course knowledge graph are obtained, and the dependencies are mapped to the first cognitive difference subgraph to obtain the second cognitive difference subgraph. The learner's cognitive mastery of the knowledge points in the current course is obtained, and the second cognitive difference subgraph is pruned to generate the cognitive difference subgraph.

[0013] Furthermore, the fully dependent reasoning rule includes: First rule: Retrieve cognitive development knowledge clusters from the aforementioned course knowledge graph. The knowledge clusters of cognitive development are then added to the fully dependent initial cognitive difference subgraph. Second rule: According to ,get The collection of knowledge points included And add the set of knowledge points to the initial cognitive difference subgraph of full dependence; Third rule: According to To obtain the knowledge cluster related to cognitive development A set of higher-level knowledge clusters with sequential dependencies And add it to the initial cognitive difference subgraph of full dependence; Fourth rule: According to To obtain the set of knowledge points contained in the higher-level knowledge cluster. And add it to the initial cognitive difference subgraph of full dependence; Fifth rule: According to ,get The set of prior knowledge points it relies on And add it to the initial cognitive difference subgraph of full dependence.

[0014] Furthermore, the pruning rules include: pruning of already mastered knowledge points, pruning of free-state knowledge points, and pruning of knowledge clusters that do not exist in the cognitive development knowledge cluster; The knowledge point pruning process I have mastered includes: knowledge point entities in the cognitive difference subgraph of full dependency. If there are knowledge points Based on learners' understanding of knowledge points The level of cognitive mastery is considered when it is believed that the learner has mastered the knowledge point. Then, the entity of the knowledge point and its related relationships are pruned from the second cognitive difference subgraph; The pruning process for detached knowledge points includes: after completing the resource entity, based on... This yields detached knowledge point entities, which are then pruned along with their associated relationships. The pruning of knowledge clusters in the absence of cognitive development knowledge clusters includes: after completing the resource entities, for leaf knowledge clusters, if the conditions are not met... If a leaf knowledge cluster is found to contain no knowledge point entities, then that entity and its relationships are pruned.

[0015] On the other hand, the present invention also provides an adaptive learning content generation system based on knowledge graphs, which includes at least a course knowledge graph construction module, a cognitive development knowledge cluster generation module, a cognitive difference subgraph generation module, and a learning content output module, for performing the steps of any of the above methods.

[0016] In general, the technical solution conceived in this invention can achieve the following beneficial effects compared with the prior art: This invention provides a method and system for generating adaptive learning content based on knowledge graphs. On one hand, it constructs a course knowledge graph including a teaching organization layer, a knowledge representation layer, and a resource aggregation layer. This not only dynamically generates learning content that better matches the teacher's teaching and the learner's cognitive level based on changes in the learner's cognitive state, but also fills in knowledge gaps, making the dynamically generated adaptive learning content more reasonable. On the other hand, it uses knowledge graphs to achieve abstract modeling of the course and aggregation of online resources, incorporating the learner's cognitive mastery of the knowledge points in the current course, dynamically updating the learner's cognitive state data on the knowledge points, and then generating dynamic adaptive learning content containing logical dependencies through steps such as reasoning of cognitive development knowledge clusters, generation of cognitive difference subgraphs, and supplementation of related resources. This provides a foundation for the implementation of an adaptive learning system. Attached Figure Description

[0017] 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 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.

[0018] Figure 1 This is a schematic diagram of the method flow of an adaptive learning content generation method and system based on knowledge graphs provided by the present invention; Figure 2This is an abstract schematic diagram of entity relationships in an adaptive learning content generation method and system based on knowledge graphs provided by the present invention. Figure 3 This is a partial hierarchical relationship structure diagram of the course knowledge graph of an adaptive learning content generation method and system based on knowledge graph provided by the present invention; Figure 4 This is a schematic diagram of the second cognitive difference subgraph of an adaptive learning content generation method and system based on knowledge graphs provided by the present invention; Figure 5 This is a schematic diagram of the pruned cognitive difference subgraph of an adaptive learning content generation method and system based on knowledge graphs provided by this invention. Detailed Implementation

[0019] 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 and embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0020] It should be noted that, in the description of the embodiments of the present invention, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, or system. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, or system that includes said element.

[0021] This application addresses the content reasoning and dynamic generation problems in existing adaptive learning methods. With the support of artificial intelligence technology, it proposes a knowledge graph-based adaptive learning content generation method that adaptively generates learning content using knowledge graphs, relational reasoning, and other technologies. Figure 1 As shown, the specific methods include: Step S101: Obtain the knowledge cluster entities, knowledge point entities, and resource entities of the course, and construct the course knowledge graph.

[0022] In other words, based on the characteristics of the course, a course model is abstracted, and the three-layer entities of knowledge cluster entities, knowledge point entities, and resource entities, as well as the inter-layer and intra-layer relationships of the entities, are extracted. The course knowledge graph is then constructed using the knowledge cluster entities, knowledge point entities, and resource entities.

[0023] Among them, such as Figure 2As shown, the course knowledge graph includes a teaching organization layer based on knowledge cluster entities, a knowledge representation layer based on knowledge point entities, and a resource aggregation layer based on resource entities.

[0024] As a preferred embodiment of the present invention, the method for constructing a course knowledge graph includes: based on the IKR three-layer course knowledge graph model, extracting three layers of entities—knowledge cluster entities, knowledge point entities, and resource entities—as well as inter-layer and intra-layer relationships, and representing them in a tuple form according to the formal representation model of the three layers of entities and inter-layer and intra-layer relationships in the IKR three-layer course knowledge graph model, thereby constructing the course knowledge graph.

[0025] Taking the "Database" course as an example, the IKR three-layer course knowledge graph model was used to model the teaching organization, knowledge structure, and unordered online support resource aggregation of the current course. The IKR three-layer course knowledge graph model includes a teaching organization layer, a knowledge representation layer, and a resource aggregation layer, and its entity relationships are abstractly represented as follows: Figure 2 As shown.

[0026] The teaching organization layer models the teaching knowledge clusters and teaching sequence. The modeling objects of this layer are abstract knowledge cluster entities and their inter-layer and intra-layer relationships.

[0027] A knowledge cluster entity refers to a meaningful knowledge module composed of n knowledge points combined according to a certain pedagogical logical relationship. The formal representation model of a knowledge cluster entity is represented by a tuple.

[0028] As an embodiment of the present invention, the formal representation model of the knowledge cluster entity G is expressed as: G = {t*, c*, p*, d, n, if_top*, date*}; where * indicates a non-empty attribute; t indicates the title of the knowledge cluster entity; c represents the category of the knowledge cluster entity, which is an important representation attribute for classifying knowledge clusters. For example, 1 is used to mark a compulsory knowledge cluster, 2 to mark an elective knowledge cluster, and 3 to mark a review knowledge cluster; p indicates the proposer of the knowledge cluster entity, for example, the proposer of the current knowledge cluster is a teacher, an expert, or a teaching and research group, which are marked as A, B, and C respectively; d indicates the description of the knowledge cluster entity, that is, the explanation of the knowledge content explained by the knowledge cluster; n indicates the knowledge cluster sequence number of the knowledge cluster entity, that is, the number or label of a series of logically ordered knowledge clusters designed by the teacher or teaching and research group; if_top describes whether the knowledge cluster entity is the starting knowledge cluster, which is used to mark whether a knowledge cluster is the starting knowledge cluster, that is, whether it can be used as the first lesson in the course teaching level; date indicates the addition time of the knowledge cluster entity.

[0029] As a specific embodiment of the present invention, taking the knowledge cluster "Basic Knowledge of Database Management Systems" in the course "Database Principles" as an example, its entity formal representation is as follows:

[0030] It should be noted that the inter-layer and intra-layer relationships of knowledge cluster entities include: the order relationship between knowledge cluster entities (NextGsOf); the hierarchical relationship between knowledge cluster entities (PartGsOf); and the inclusion relationship between knowledge cluster entities and knowledge point entities in the knowledge representation layer (BelongTo).

[0031] Among these, the sequential relationship connects knowledge clusters as follows: Sequential relationship (knowledge cluster-knowledge cluster); in teaching or curriculum design, the order in which knowledge clusters are learned constitutes the sequential relationship between them, such as database system schema-three-level schema structure. Hierarchical relationship (knowledge cluster-knowledge cluster); in teaching content, a large knowledge cluster A can be divided into several smaller knowledge clusters A1, A2, etc., for learning. That is, A and A1, A2, etc., have a hierarchical relationship, such as the chapter and section definitions in a textbook, for example, basic data knowledge-common data types. Containment relationship (knowledge cluster-knowledge point); a knowledge cluster is a reorganization of knowledge points; therefore, the connection between a knowledge cluster and a knowledge point is a containment relationship, for example, common data types-integer.

[0032] The teaching organization layer extracts entities and relationships. The teaching organization layer is an abstract model of the course teaching organization structure, which is extracted from the existing course structure or teaching syllabus.

[0033] It should be noted that the extraction of intra-level relationships between knowledge cluster entities involves obtaining the course hierarchy directory as candidate knowledge cluster entities, extracting the hierarchical relationships between knowledge cluster entities based on the hierarchical relationship of course chapters, and determining the sequential relationship between knowledge cluster entities at the same level based on the sequential learning relationship of chapters within the same level of the course. For the extraction of inter-level relationships between knowledge cluster entities, leaf knowledge cluster entities (i.e., those without lower-level knowledge clusters) are converted into text using speech-to-text and image recognition technologies. A CRF model is then used for entity recognition, selecting the top ten ranked knowledge point entities as the candidate set of knowledge point entities with inclusion relationships with the knowledge cluster. The rationality of the relationships is then determined, completing the extraction of inclusion relationships between knowledge cluster entities and knowledge point entities in the knowledge representation layer.

[0034] The knowledge representation layer models the course knowledge and its inherent logical dependencies. The entity objects modeled in this layer are specific knowledge points and their inter-layer and intra-layer relationships.

[0035] Specifically, the model construction of the knowledge representation layer includes: knowledge classification modeling, knowledge point entity representation modeling, and inter-layer and intra-layer relationship modeling of knowledge point entities.

[0036] Knowledge classification modeling, based on Bloom's Taxonomy of Learning Objectives, can be divided into factual knowledge, procedural knowledge, and conceptual knowledge. As an embodiment of this invention, knowledge classification modeling further refines knowledge into: symbols, events, people, software, concepts, theories, algorithms, operations, rules, and methods.

[0037] Knowledge point entity representation is based on classification. First, it abstracts the common attribute descriptions shared by all knowledge points. Then, based on the characteristics of each type of knowledge, it abstracts targeted feature attribute descriptions.

[0038] In the knowledge entity representation modeling, the formal representation model of each knowledge entity K is represented by a tuple.

[0039] As an embodiment of the present invention, each knowledge point entity K can be represented as: K = {t*, c*, o, e,es, d, cog} i , a,p i}; where * indicates a non-empty attribute; t indicates the name of the knowledge point entity; c indicates the category to which the knowledge point entity belongs; o indicates synonyms, near-synonyms, or other aliases of the knowledge point entity; e indicates the English name of the knowledge point entity; es indicates the English abbreviation of the knowledge point entity; d indicates the explanation and description of the knowledge point entity; cog indicates the cognitive level to be achieved for the knowledge point entity; cogi indicates different levels of cognitive understanding to be achieved for the knowledge point; a indicates supplementary content of the knowledge point entity; pi indicates a certain characteristic attribute of the knowledge point entity, and the number and types of pi vary for different knowledge points.

[0040] As a specific embodiment of the present invention, taking the knowledge point "relation" in the "Database Principles" course as an example, its entity formal representation is as follows:

[0041] It should be noted that the relationships between and within knowledge point entities include: parent-child relationships (SubClassof / SuperClassof) between knowledge point entities, with the connecting entity being knowledge point-knowledge point, e.g., key-primary key; whole-part relationships (PartOf / IsPartOf) between knowledge point entities, with the connecting entity being knowledge point-knowledge point, e.g., relation-tuple; prior-successor relationships (Requires / BeRequiredTo) between knowledge point entities, with the connecting entity being knowledge point-knowledge point, e.g., relation-relational algebra; sibling relationships (BrotherClass) between knowledge point entities, with the connecting entity being knowledge point-knowledge point, e.g., primary key-foreign key; correlation relationships (HasRelatedOf) between knowledge point entities, with the connecting entity being knowledge point-knowledge point, e.g., frequent pattern-association rule; and resource relationships (RelResource) between knowledge point entities and resource entities in the resource aggregation layer, with the connecting entity being knowledge point-resource.

[0042] Entities and relations in the knowledge representation layer can be extracted using existing automatic extraction techniques.

[0043] The resource aggregation layer models the online resources related to the course. The entities modeled in this layer are multimodal resource entities and their inter-layer and intra-layer relationships. The formal representation model of resource entity S is represented by tuples.

[0044] As an embodiment of the present invention, each resource entity S can be represented as: S = {t*, c*, p, u*, step*, d, date}; where * indicates a non-empty attribute; t indicates the title of the resource entity; c indicates the category of the resource entity, such as text, video, animation, etc.; p indicates the instructor of the resource entity, such as the instructor, expert, or teaching and research group; u indicates the URL address of the resource entity, that is, the online access address of the resource; step indicates the teaching stage of the resource entity, such as including the stages of introduction, overview, review, lecture, integration, and application; d indicates the explanation and description of the resource entity; and date indicates the publication time of the resource entity.

[0045] As a specific embodiment of the present invention, taking a course resource in China University MOOC as an example, its entity formal representation is as follows:

[0046] It should be noted that the inter-layer and intra-layer relationships of resource entities include: the related knowledge point relationship (Has_relateKG_of) between the resource entity and the knowledge point entity of the knowledge representation layer, that is, the knowledge points included in the resource, and the connecting entity is resource-knowledge point; the main knowledge point relationship (Has_mainKG_of) between the resource entity and the knowledge point entity of the knowledge representation layer, that is, the main knowledge points described by the resource, and the connecting entity is resource-knowledge point; and the sequential relationship (NextS / LastS) between resource entities, that is, the original order of the resources, and the connecting entity is resource-resource.

[0047] In one embodiment of this invention, the entity and relation extraction of the resource aggregation layer involves selecting seven database-related courses from the China University MOOC network as representatives of online resources to obtain resource entities. These entities are then converted into text using speech-to-text and image recognition technologies. A CRF model is then used for entity recognition, selecting the entity with the highest frequency as the candidate main knowledge point entity for the resource, and selecting the second to fifth ranked knowledge point entities as the candidate related knowledge entity set for the resource. The rationality of the relations is then determined, thus completing the entity and relation extraction of the resource aggregation layer.

[0048] Thus, the extraction of three layers of entities—knowledge cluster entities, knowledge point entities, and resource entities—as well as inter-layer and intra-layer relationships, is completed, constructing a course knowledge graph. As an embodiment of this invention, taking the "database" course as an example, a course knowledge graph including 2368 entities and 14027 relationships is constructed.

[0049] Step S102: Obtain the learner's cognitive mastery of the knowledge points in the current course; based on the cognitive mastery and according to the teaching organization layer and knowledge representation layer, obtain the learner's cognitive development knowledge cluster at the current stage.

[0050] The learner's level of understanding and mastery of the knowledge points in the current course, which is also the learner's evaluation of the probability of mastering the knowledge characteristics, is based on relatively mature existing technology and will not be elaborated here.

[0051] As an embodiment of the present invention, cognitive development knowledge cluster The process of obtaining it includes: S201: Determine whether the learner's cognitive mastery of the knowledge points has been updated; if not, use the previously obtained cognitive development knowledge cluster as the current cognitive development knowledge cluster. Conversely, if the condition is not met, then S202 will be executed.

[0052] S202: Retrieve the learner's stored data to determine if the current learner has a stored record of a cognitive development knowledge cluster; if so, use the stored cognitive development knowledge cluster as the cognitive development knowledge cluster for the current stage. Conversely, the initial knowledge cluster of the current course is retrieved from the course knowledge graph. and the initial knowledge cluster As a knowledge cluster of cognitive development at the current stage .

[0053] S203: Retrieve the course knowledge graph to obtain the cognitive development knowledge clusters for the current stage. .

[0054] Among them, acquiring the knowledge cluster of cognitive development at the current stage ,include: If the conditions are met: If yes, then S204 will be executed; otherwise, it will be executed through... Acquire cognitive development knowledge cluster Non-dependent knowledge clusters in subclass hierarchical knowledge clusters and will not depend on knowledge clusters As a cluster of knowledge for cognitive development at the current stage ; in, Knowledge clusters representing the teaching organization level; Representing the order relation triple ; Represents a hierarchical triplet .

[0055] S204: Acquiring Knowledge Clusters for Cognitive Development The set of knowledge points contained in the knowledge point set and the degree of cognitive mastery of the knowledge points contained in the knowledge point set are used to obtain the corresponding cognitive development knowledge cluster. The level of cognitive mastery.

[0056] Among them, the corresponding knowledge clusters of cognitive development Cognitive mastery The calculation methods include: ; in, Represents a set of knowledge points. ; Indicates the learner's understanding of knowledge points The level of cognitive mastery; Set a threshold to indicate that the learner has mastered the knowledge point. ; This indicates that learners are in the knowledge cluster of cognitive development. Of the knowledge points included, those that have already been mastered The number of; This indicates that learners are in the knowledge cluster of cognitive development. The number of all knowledge points contained therein.

[0057] S205: Assessing knowledge clusters in cognitive development If the student's cognitive mastery exceeds a set threshold, then the cognitive development knowledge cluster for the next stage of learning is inferred from the course knowledge graph and used as the cognitive development knowledge cluster for the current stage. Conversely, directly output the knowledge cluster of cognitive development at the current stage. Preferably, the set threshold is 0.8.

[0058] Furthermore, the cognitive development knowledge clusters for the next stage of learning can be deduced from the course knowledge graph. And as a knowledge cluster for cognitive development at the current stage. ,include: S301: If ,but The next stage of learning will focus on the cognitive development knowledge clusters. As a cluster of knowledge for cognitive development at the current stage If the condition is not met, then execute S203; otherwise, execute S302. S302: If Then the knowledge clusters in the knowledge graph As a cluster of knowledge for cognitive development at the current stage If the program returns to its previous state, it will execute S301; otherwise, it indicates that the course has been completed and the course learning has ended.

[0059] As a specific embodiment of the present invention, the learner's cognitive mastery of the knowledge points in the current course was obtained, as shown in Table 1.

[0060] Table 1. A learner's level of understanding and mastery of certain knowledge points

[0061] First, determine if the learner's understanding of the knowledge points has been updated, and if the current learner with ID "2270" has a stored record of the cognitive development knowledge cluster in the database, then directly use the cognitive development knowledge cluster that the learner obtained last time as the cognitive development knowledge cluster of the current stage.

[0062] Then, the knowledge graph of the current course is retrieved to obtain the cognitive development knowledge clusters for the current stage. ;condition If established, then the knowledge cluster of cognitive development will be acquired. The collection of knowledge points and knowledge point collection The degree of cognitive mastery of the knowledge points contained therein. According to the formula... Calculations are performed to obtain the learner's knowledge clusters of cognitive development. Cognitive mastery =0.83.

[0063] Then, due to =0.83, If the value is greater than 0.8, then the knowledge clusters that can be learned in the next stage can be directly deduced from the course knowledge graph. Due to conditions , =2273, then Assign to It then returns to the knowledge graph of the current course to retrieve the cognitive development knowledge clusters for the current stage. ;condition If it is not valid, then proceed through infer Non-dependent knowledge clusters in subclass hierarchical knowledge clusters , =2276, and assign it to .

[0064] Finally, continue with the following steps. =0, less than 0.8; Output the learner's cognitive development knowledge cluster. .

[0065] By performing the above steps, the learner's cognitive development knowledge clusters can be derived, such as... Figure 3 As shown, this is the cognitive development knowledge cluster of the learner with the id "2276".

[0066] Step S103: Based on the current cognitive development knowledge cluster and the dependency relationships between various entities in the course knowledge graph, generate a cognitive difference subgraph according to the full dependency reasoning rule and pruning processing rule.

[0067] In other words, based on the reasoning rules and the knowledge clusters of cognitive development, the necessary prior knowledge is inferred through steps such as fully dependent subgraph reasoning and subgraph pruning. The knowledge points not mastered in the previous stage are then filled in to generate a cognitive difference subgraph for the target learner.

[0068] Furthermore, the steps for generating the cognitive difference subgraph include: Create an empty subgraph as the initial cognitive difference subgraph.

[0069] The course knowledge graph is retrieved, and a set of entities that generate full dependencies is obtained based on the full dependency reasoning rule. The entity set is then added to the initial cognitive difference subgraph to obtain the first cognitive difference subgraph.

[0070] As an example, the fully dependent inference rule includes: First rule: Retrieve cognitive development knowledge clusters from the course knowledge graph. Furthermore, the knowledge cluster of cognitive development is added to the initial cognitive difference subgraph of full dependence.

[0071] Second rule: According to ,get The collection of knowledge points included The set of knowledge points is then added to the initial cognitive difference subgraph of full dependence.

[0072] Third rule: According to Acquiring knowledge clusters related to cognitive development A set of higher-level knowledge clusters with sequential dependencies And add it to the initial cognitive difference subgraph of full dependence.

[0073] Fourth rule: According to To obtain the set of knowledge points contained in the higher-level knowledge cluster. And add it to the initial cognitive difference subgraph of full dependence.

[0074] Fifth rule: According to ,get The set of prior knowledge points it relies on And add it to the initial cognitive difference subgraph of full dependence.

[0075] Based on the full dependency reasoning rule, all the knowledge points obtained are added to the initial cognitive difference subgraph to obtain the first cognitive difference subgraph.

[0076] Based on the entity set, the dependencies between all entities in the entity set within the course knowledge graph are obtained, and these dependencies are mapped to the first cognitive difference subgraph to obtain the second cognitive difference subgraph. As an example, according to... Get entity set We obtain the relational dependencies between all entities in the first cognitive difference subgraph in the course knowledge graph.

[0077] The learner's cognitive mastery of the knowledge points in the current course is obtained, and the second cognitive difference subgraph is pruned to generate a cognitive difference subgraph.

[0078] As an example, the pruning rules include: pruning of already mastered knowledge points, pruning of free-state knowledge points, and pruning of knowledge clusters that do not exist in the knowledge clusters of cognitive development.

[0079] As a specific example, the knowledge point pruning process has been mastered, including: knowledge point entities in the cognitive difference subgraph of full dependency. If there are knowledge points Based on learners' understanding of knowledge points The level of cognitive mastery is considered when it is believed that the learner has mastered the knowledge point. Then, the entity of the knowledge point and its related relationships are pruned from the second cognitive difference subgraph.

[0080] As a specific embodiment, the detached-state knowledge point pruning process includes: after completing the resource entity, according to... We obtain the free-state knowledge point entity and prune the entity and its relationships.

[0081] As a specific embodiment, the knowledge cluster pruning process in the absence of cognitive development knowledge clusters includes: after completing the resource entities, for the leaf knowledge clusters, if the condition... If a leaf knowledge cluster is found to contain no knowledge point entities, then that entity and its relationships are pruned.

[0082] As a specific embodiment of the present invention, according to the first rule, the cognitive development knowledge cluster entity with the id "2276" is retrieved and added to the fully dependent initial cognitive difference subgraph.

[0083] According to the second rule, based on the formula Retrieve the set of knowledge points contained in the cognitive development knowledge cluster with id "2276". ={"relation", "component", "tuple", "relation instance"}, and add it to the initial cognitive difference subgraph of full dependencies.

[0084] According to the third rule, based on the formula Acquisition and cognitive development of knowledge clusters A set of higher-level knowledge clusters with sequential dependencies , ={“2270”}, and add it to the initial cognitive difference subgraph of full dependence.

[0085] According to the fourth rule, based on the formula , obtain The collection of knowledge points , ={"Data Model", "Relational Model", "Hierarchical Model", "Data Structure", "Data Manipulation", "Data Integrity Constraints"}, and add them to the initial cognitive difference subgraph of full dependency.

[0086] According to the fifth rule, based on the formula , obtain The set of prior knowledge points it relies on , ={"Cartesian product", "domain"}, and add it to the initial cognitive difference subgraph of full dependence.

[0087] Based on entity sets, according to the formula The entity set is deduced, the dependencies between all entities in the initial cognitive difference subgraph in the course knowledge graph are obtained, and these dependencies are mapped to the second cognitive difference subgraph, such as... Figure 4 As shown.

[0088] Then, the learner's mastery of the knowledge points in the cognitive difference subgraph is retrieved, and the mastered knowledge points and loose knowledge points are pruned, as are the knowledge clusters that do not contain knowledge points, to generate a cognitive difference subgraph oriented towards the target learner.

[0089] Based on pruning, the mastered knowledge points are pruned. For example, according to the partial cognitive mastery of a learner shown in Table 1 of this embodiment, the knowledge points are pruned accordingly. Figure 4 Judging from the knowledge point entities in the fully dependent cognitive difference subgraph, it can be seen that the student's cognitive diagnostic score is greater than 0.8 for the knowledge points "domain", "data model", "relational model", "hierarchical model", "data structure" and "data operation". Therefore, the knowledge point entity and its related relationships are pruned from KGraph.

[0090] Pruning of knowledge points in the free state involves pruning existing knowledge points and then, based on... If no free-state knowledge points are found during reasoning, no pruning will be performed.

[0091] Pruning leaf knowledge clusters that do not contain any knowledge points: Building upon the pruning of detached knowledge points, for leaf knowledge clusters with id="2270" and "2276", conditional... If none of these conditions are met, then no pruning will be performed. Figure 5 As shown, the pruned cognitive difference subgraph is obtained.

[0092] Step S104: Based on the knowledge points included in the cognitive difference subgraph, retrieve the related resources in the resource aggregation layer of the course knowledge graph, and map the related resources and the aggregation relationship between resources to the cognitive difference subgraph to generate adaptive learning content for the learner's current stage.

[0093] In other words, by supplementing related resources, and based on the knowledge points contained in the pruned cognitive difference subgraph, related resources in the resource aggregation layer are retrieved as supporting resources for adaptive learning.

[0094] As an embodiment of the present invention, firstly, based on the cognitive difference subgraph, according to The required supporting resources S = {"S2", "S3", "S4", "S5", "S6", "S7", "S8", "S11", "S13"} are obtained, and the supporting resources S are added to the cognitive difference subgraph.

[0095] Then, complete the aggregation relation, based on This process obtains all aggregation relationships between the supporting resource set and other entities in the cognitive difference subgraph, and maps all aggregation relationships to the cognitive difference subgraph.

[0096] This generates a three-layer cognitive difference subgraph based on the learner's current learning state, and generates adaptive learning content for the learner, providing a foundation for subsequent adaptive learning path planning, adaptive learning content recommendation, and adaptive navigation.

[0097] On the other hand, the present invention also provides an adaptive learning content generation system based on knowledge graphs, characterized in that the system includes at least a course knowledge graph construction module, a cognitive development knowledge cluster generation module, a cognitive difference subgraph generation module, and a learning content output module, for performing the steps of any of the above methods.

[0098] The technical features of the system are consistent with those of the methods described above, and will not be repeated here.

[0099] 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., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0100] 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 knowledge graph-based adaptive learning content generation method, characterized in that, The method includes: The knowledge cluster entities, knowledge point entities, and resource entities of the course are obtained, and a course knowledge graph is constructed. The course knowledge graph includes a teaching organization layer based on knowledge cluster entities, a knowledge representation layer based on knowledge point entities, and a resource aggregation layer based on resource entities. Obtain the learner's cognitive mastery of the knowledge points in the current course; based on the cognitive mastery, and according to the teaching organization layer and the knowledge representation layer, obtain the learner's cognitive development knowledge cluster at the current stage; Based on the current cognitive development knowledge cluster and the dependencies between various entities in the course knowledge graph, a cognitive difference subgraph is generated according to the full dependency reasoning rule and the pruning processing rule. Based on the knowledge points included in the cognitive difference subgraph, the associated resources in the resource aggregation layer of the course knowledge graph are retrieved, and the associated resources and the aggregation relationship between resources are mapped to the cognitive difference subgraph to generate adaptive learning content for the learner's current stage. The fully dependent reasoning rule includes: retrieving cognitive development knowledge clusters from the course knowledge graph, obtaining the set of knowledge points contained in the cognitive development knowledge clusters, obtaining the set of superior knowledge clusters that have a sequential dependency relationship with the cognitive development knowledge clusters, obtaining the set of knowledge points contained in the superior knowledge clusters, obtaining the set of prior knowledge points on which the set of knowledge points depends, and adding all the retrieved and obtained content to the initial cognitive difference subgraph of the fully dependent reasoning.

2. The adaptive learning content generation method based on knowledge graph as described in claim 1, characterized in that, The method for constructing the course knowledge graph includes: Based on the IKR three-layer course knowledge graph model, the three-layer entities—knowledge cluster entities, knowledge point entities, and resource entities—as well as their inter-layer and intra-layer relationships are extracted. Then, according to the formal representation model of the three-layer entities and their inter-layer and intra-layer relationships in the IKR three-layer course knowledge graph model, a tuple-based representation is adopted to construct the course knowledge graph. The IKR three-layer course knowledge graph model is a three-layer course knowledge graph model that includes a teaching organization layer, a knowledge representation layer, and a resource aggregation layer.

3. The adaptive learning content generation method based on knowledge graph as described in claim 1, characterized in that, The knowledge cluster of cognitive development The process of obtaining it includes: S201: Determine whether the learner's cognitive mastery of the knowledge points has been updated; if not, use the previously obtained cognitive development knowledge cluster as the current cognitive development knowledge cluster. Conversely, if the condition is not met, then execute S202. S202: Determine whether the current learner has a stored record of a cognitive development knowledge cluster; if so, use the stored cognitive development knowledge cluster as the cognitive development knowledge cluster for the current stage. Conversely, the initial knowledge cluster of the current course is retrieved from the course knowledge graph. and the initial knowledge cluster As a knowledge cluster of cognitive development at the current stage ; S203: Retrieve the course knowledge graph to obtain the cognitive development knowledge clusters for the current stage. ; S204: Acquiring Knowledge Clusters for Cognitive Development The set of knowledge points contained in the knowledge point set and the degree of cognitive mastery of the knowledge points contained in the knowledge point set are used to obtain the corresponding cognitive development knowledge cluster. Cognitive mastery level; S205: Assessing knowledge clusters in cognitive development If the level of cognitive mastery exceeds a set threshold, then the cognitive development knowledge cluster for the next stage of learning is inferred from the course knowledge graph and used as the cognitive development knowledge cluster for the current stage. Conversely, directly output the knowledge cluster of cognitive development at the current stage. .

4. The adaptive learning content generation method based on knowledge graph as described in claim 3, characterized in that, In S203, the course knowledge graph is retrieved to obtain the cognitive development knowledge clusters for the current stage. ,include: If the conditions are met: If so, then execute S204; Conversely, through Acquire cognitive development knowledge cluster Non-dependent knowledge clusters in subclass hierarchical knowledge clusters and the independent knowledge cluster As a knowledge cluster of cognitive development at the current stage ; in, This represents the knowledge clusters of the teaching organization layer; Representing the order relation triple ; Represents a hierarchical triplet .

5. The adaptive learning content generation method based on knowledge graph as described in claim 3, characterized in that, Corresponding cognitive development knowledge cluster Cognitive mastery The calculation methods include: ； in, Represents a set of knowledge points. ; Indicates the learner's understanding of knowledge points The level of cognitive mastery; Set a threshold to indicate that the learner has mastered the knowledge point. ; This indicates that learners are in the knowledge cluster of cognitive development. Of the knowledge points included, those that have already been mastered The number of; This indicates that learners are in the knowledge cluster of cognitive development. The number of all knowledge points contained therein.

6. The adaptive learning content generation method based on knowledge graph as described in claim 3, characterized in that, Infer the cognitive development knowledge clusters for the next stage of learning from the aforementioned course knowledge graph. And as a knowledge cluster for cognitive development at the current stage. ,include: S301: If ,but The next stage of learning will focus on the cognitive development knowledge clusters. As a knowledge cluster of cognitive development at the current stage If the condition is not met, then execute S203; otherwise, execute S302. S302: If Then the knowledge clusters in the knowledge graph As a knowledge cluster of cognitive development at the current stage If the condition is met, the course will return to execute S301; otherwise, it indicates that the course is complete and the course will end. in, This indicates the sequential relationship between entities in a knowledge cluster; It represents the hierarchical relationship between entities in a knowledge cluster.

7. The adaptive learning content generation method based on knowledge graph as described in claim 1, characterized in that, The steps for generating the cognitive difference submap include: Create an empty subgraph as the initial cognitive difference subgraph; Retrieve the course knowledge graph, obtain the fully dependent entity set based on the fully dependent reasoning rule, and add the entity set to the initial cognitive difference subgraph to obtain the first cognitive difference subgraph; Based on the entity set, the dependencies between all entities in the entity set in the course knowledge graph are obtained, and the dependencies are mapped to the first cognitive difference subgraph to obtain the second cognitive difference subgraph. The learner's cognitive mastery of the knowledge points in the current course is obtained, and the second cognitive difference subgraph is pruned to generate the cognitive difference subgraph.

8. The adaptive learning content generation method based on knowledge graph as described in claim 1 or 7, characterized in that, The fully dependent inference rule includes: First rule: Retrieve cognitive development knowledge clusters from the course knowledge graph. The cognitive development knowledge cluster is then added to the fully dependent initial cognitive difference subgraph; Second rule: According to ,get The collection of knowledge points included And add the set of knowledge points to the initial cognitive difference subgraph of full dependence; Third rule: According to To obtain the knowledge cluster related to cognitive development A set of higher-level knowledge clusters with sequential dependencies And add it to the initial cognitive difference subgraph of full dependence; Fourth rule: According to To obtain the set of knowledge points contained in the higher-level knowledge cluster. And add it to the initial cognitive difference subgraph of full dependence; Fifth rule: According to ,get The set of prior knowledge points it relies on And add it to the initial cognitive difference subgraph of full dependence; in, This indicates the inclusion relationship between knowledge cluster entities and knowledge point entities in the knowledge representation layer. This indicates the sequential relationship between entities in a knowledge cluster; This represents the hierarchical relationship between entities in a knowledge cluster; It represents the prior-successor relationship between knowledge point entities.

9. A knowledge graph-based adaptive learning content generation method as described in claim 1 or 7, characterized in that, The pruning rules include: pruning of mastered knowledge points, pruning of free-state knowledge points, and pruning of knowledge clusters that do not exist in the cognitive development knowledge clusters; The knowledge point pruning process I have mastered includes: knowledge point entities in the cognitive difference subgraph of full dependency. If there are knowledge points Based on learners' understanding of knowledge points The level of cognitive mastery is considered when it is believed that the learner has mastered the knowledge point. Then, the entity of the knowledge point and its related relationships are pruned from the second cognitive difference subgraph; The pruning process for detached knowledge points includes: after completing the resource entity, based on... This yields detached knowledge point entities, which are then pruned along with their associated relationships. The pruning of knowledge clusters in the absence of cognitive development knowledge clusters includes: after completing the resource entities, for leaf knowledge clusters, if the conditions are not met... If a leaf knowledge cluster is found to contain no knowledge point entities, then that entity and its relationships will be pruned.

10. A knowledge graph-based adaptive learning content generation system, characterized in that, The system includes at least a course knowledge graph construction module, a cognitive development knowledge cluster generation module, a cognitive difference subgraph generation module, and a learning content output module, for performing the steps of the method described in any one of claims 1 to 9.

Images