Method for recommending education content based on knowledge graph and related product

Abstract

The disclosure discloses a method for recommending educational content based on a knowledge graph and related products, in which the knowledge graph is constructed in a knowledge base containing information and examples corresponding to a plurality of knowledge points, such that a plurality of entities and edges of the knowledge graph represent the plurality of knowledge points and their mutual relationships, respectively. The method comprises the following steps: inputting a recommendation request by a user; identifying a starting entity representing the starting knowledge point in a knowledge graph based on the data; searching one or more additional entities in the knowledge graph along one or more edges starting from the starting entity according to a recommendation policy; and outputting information or examples corresponding to the starting knowledge point and the extended knowledge point in the knowledge base as recommended education content. According to the embodiment of the invention, the recommendation request including the recommendation strategy and the data corresponding to the starting knowledge point is input by the user, so that the user can specify the learning starting point and the recommendation strategy, and various learning requirements of the user are met.

Description

Technical Field

[0001] The present disclosure generally relates to the technical field of information processing. More specifically, the present disclosure relates to a method, system, device and storage medium for recommending educational content based on a knowledge graph. Background Art

[0002] The concept of knowledge graph was formally proposed by Google in 2012, aiming to realize a smarter search engine, and began to be popularized in academia and industry after 2013. At present, with the continuous development of intelligent information service applications, knowledge graph has been widely used in various aspects such as intelligent search, intelligent question and answer, personalized recommendation, and intelligence analysis.

[0003] The knowledge graph represents the relationship between entities in the real world in the form of a heterogeneous graph, which is used to quickly describe concepts in the physical world and their relationships, where nodes are entities and edges represent the relationships between entities. The knowledge graph effectively processes, processes, and integrates the data of complex documents, converting them into simple and clear "entity, relationship, entity" triples, and finally aggregates a large amount of knowledge, thereby achieving rapid response and reasoning of knowledge. In other words, the knowledge graph uses a graph to represent the relationship between data entities.

[0004] The example construction process of the knowledge graph is as follows Figure 1 As shown, it is mainly divided into three steps: information extraction, knowledge fusion and knowledge processing.

[0005] Information extraction refers to extracting concepts, attributes, and relationships between entity concepts from various types of data sources, and forming ontological knowledge expressions. Data sources can generally be divided into three categories: structured data, semi-structured data, and unstructured data. Structured data has clear relationship definitions and is easy to use. It is generally represented and stored in a relational database. Semi-structured data is a special form of structured data. It does not conform to the structure of a relational database, but contains relevant tags, and the structure and content of the data are mixed together. Unstructured data has no fixed structure and generally contains text, pictures, reports, and other content.

[0006] Knowledge fusion refers to the integration of knowledge after it is acquired to eliminate ambiguity and contradictory content. Knowledge fusion includes two parts: entity linking and knowledge merging. Entity linking links the objects extracted from the text to the corresponding entities in the knowledge base, and then performs entity disambiguation and coreference resolution, that is, determining whether there are entities with the same name in the knowledge base that represent the same meaning, and whether there are different entities that represent the same meaning. Knowledge merging can merge external knowledge bases and relational databases to update and optimize the knowledge graph.

[0007] The knowledge processing step evaluates the new knowledge acquired in the previous step and then adds it to the knowledge base to ensure the quality of the knowledge base. The steps include ontology construction, knowledge reasoning, and quality assessment. Ontology construction normalizes the extracted knowledge and clearly defines concepts and relationships. Knowledge reasoning starts from existing entity relationships and infers new relationships to improve the knowledge graph. Quality assessment quantifies the credibility of knowledge to ensure the quality of the knowledge base.

[0008] There are two ways to build a knowledge graph: top-down and bottom-up. The top-down construction method requires first defining the ontology (Ontology or Schema), and then completing the process of information extraction to graph construction based on the input data. This method is more suitable for the construction of professional knowledge graphs, such as enterprise knowledge graphs, which are used by professional users in specific fields. The bottom-up construction method extracts high-confidence knowledge from open linked data (Open Linked Data) or extracts knowledge from unstructured text to complete the construction of the knowledge graph. This method is more suitable for common sense knowledge, such as the construction of general knowledge graphs such as names of people and organization names.

[0009] As mentioned above, personalized recommendation is one of the many application aspects of knowledge graphs, thanks to the two major advantages of knowledge graphs: First, knowledge graphs can integrate multi-level heterogeneous relationships such as users, products, and features into one, which can effectively improve the accuracy of recommendations; second, the relationships in the knowledge graph are clear, which provides explainability for recommendations.

[0010] Generally speaking, the purpose of recommendation is to recommend one or a series of items to a given user. The process can be divided into the following steps: First, the representation of the user and the item needs to be generated, u i and v i ; Secondly, define the score function f: to represent the user's preference for the project; finally, the candidate projects are ranked by the scoring function, and the top-ranked projects are recommended to the user. Conventional recommendation methods based on knowledge graphs basically include the following three types: embedding-based methods, path-based methods, and joint methods. The idea of ​​the embedding-based method is to use the knowledge graph embedding algorithm (KGE) to enrich the representation of users and projects with the information of the graph. The basic idea of ​​the path-based method is to consider the connectivity similarity (semantic similarity / structural equivalence) between users and projects, thereby improving the recommendation effect. The joint method combines the first two methods. Its basic idea is to use embedding propagation to improve the representation of nodes with multi-hop neighbors in the graph. The propagation process can be regarded as a search for user preference patterns, similar to finding connection patterns in path-based methods. These recommendation methods have their own corresponding rules.

[0011] In the field of education, personalized recommendations based on knowledge graphs include recommending educational content based on knowledge graphs. As an example, patent application CN112685645A discloses a method and corresponding system for intelligent education recommendation based on knowledge graphs, which includes: obtaining personal information and learning information of surveyed users, obtaining at least one key node information based on the personal information and learning information; constructing an intelligent education recommendation knowledge graph based on the at least one key node information; obtaining user information of target users, obtaining at least one user node information based on the user information, and obtaining recommendation node information corresponding to the at least one user node information based on the intelligent education recommendation knowledge graph; generating recommended content based on the recommended node information, and sending the recommended content to a preset terminal. As another example, patent application CN112785140A discloses an intelligent education recommendation method and corresponding system based on a knowledge graph, the method comprising: constructing a knowledge graph; obtaining the learner's personal information, and determining matching study partner information and teacher information based on the learner's personal information; determining the learner's learning situation based on the personal information, determining the study partner's learning situation based on the study partner information, and determining the teaching situation based on the teacher information; integrating the learner's learning situation, the study partner's learning situation and the teaching situation to obtain the learner's learning improvement needs; determining the corresponding educational content based on the learning improvement needs and the knowledge graph and recommending it to the learner. Summary of the invention

[0012] The inventor of the present disclosure has found that the methods and systems for recommending educational content based on knowledge graphs in the prior art are based on the user's personal information and learning situation. The system automatically determines the learning starting point in the knowledge graph and starts to obtain recommended content from it, but the user cannot specify the learning starting point or the recommended strategy. These methods and systems are not convenient and flexible enough for users, because in many cases, users may want to specify the learning starting point and / or the recommended strategy for various learning requirements such as review, preview, divergent learning, and free learning.

[0013] The inventors of the present disclosure also discovered that the knowledge graphs in such methods and systems of the prior art do not have detailed knowledge point classifications and knowledge point relationships that fully reflect teaching logic, and therefore cannot accurately recommend educational content or widely support various recommendation strategies.

[0014] In order to at least solve one or more of the technical problems mentioned above, the present disclosure proposes a solution for recommending educational content based on knowledge graphs in multiple aspects.

[0015] In a first aspect, the present disclosure provides a method for recommending educational content based on a knowledge graph, wherein the knowledge graph is constructed in a knowledge base containing information and examples corresponding to a plurality of knowledge points, so that a plurality of entities and edges of the knowledge graph respectively represent the plurality of knowledge points and their mutual relationships, and the information corresponding to the knowledge point represented by each entity is stored as an attribute of the entity. The method comprises: a user inputs a recommendation request, the recommendation request comprising a recommendation strategy and data corresponding to a starting knowledge point; based on the data, a starting entity representing the starting knowledge point is identified in the knowledge graph; according to the recommendation strategy, one or more additional entities are searched in the knowledge graph along one or more edges starting from the starting entity; and information or examples corresponding to the starting knowledge point and one or more extended knowledge points represented by the one or more additional entities in the knowledge base are output as recommended educational content.

[0016] In some embodiments, the plurality of knowledge points are divided into the following three types: concept, knowledge, and application.

[0017] In some embodiments, the mutual relationships of the multiple knowledge points are divided into the following ten relationships: inclusion relationship, instance relationship, pre-requisite relationship, property relationship, method relationship, classification relationship, representation relationship, calculation relationship, derived concept relationship and other relationships.

[0018] In a second aspect, the present disclosure provides a system for recommending educational content based on a knowledge graph, wherein the knowledge graph is constructed in a knowledge base containing information and examples corresponding to a plurality of knowledge points, so that a plurality of entities and edges of the knowledge graph respectively represent the plurality of knowledge points and their mutual relationships, and the information corresponding to the knowledge point represented by each entity is stored as an attribute of the entity. The system comprises: an input module for a user to input a recommendation request, wherein the recommendation request comprises a recommendation strategy and data corresponding to a starting knowledge point; an identification module for identifying a starting entity representing the starting knowledge point in the knowledge graph based on the data; a search module for searching for one or more additional entities in the knowledge graph along one or more edges starting from the starting entity according to the recommendation strategy; and an output module for outputting information or examples corresponding to the starting knowledge point and one or more extended knowledge points represented by the one or more additional entities in the knowledge base as recommended educational content.

[0019] In a third aspect, the present disclosure provides a device for recommending educational content based on a knowledge graph, wherein the device comprises a processor and a memory storing instructions, wherein the instructions, when executed by the processor, cause the device to perform the method of the first aspect.

[0020] In a fourth aspect, the present disclosure provides a machine-readable storage medium having stored thereon instructions, which, when executed on a device, causes the device to perform the method of the first aspect.

[0021] Through the method, system, device and storage medium for recommending educational content based on the knowledge graph as provided above, the disclosed embodiment allows the user to input a recommendation request including a recommendation strategy and data corresponding to the starting knowledge point, so that the user can specify the learning starting point and recommendation strategy, thereby meeting their various learning requirements. Further, in some embodiments, by dividing multiple knowledge points into three types: concepts, knowledge and applications, the classification of knowledge points is more refined, which is conducive to more refined recommendation of educational content. Furthermore, in some embodiments, by dividing the mutual relationships between knowledge points into inclusion relationships, instance relationships, precondition relationships, property relationships, method relationships, classification relationships, representation relationships, calculation relationships, derived concept relationships and other relationships, the knowledge graph constructed thereby can more comprehensively reflect the teaching logic, thereby being able to widely support various recommendation strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] By reading the detailed description below with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present disclosure will become readily understood. In the accompanying drawings, several embodiments of the present disclosure are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0023] Figure 1 An example construction process of a knowledge graph is shown;

[0024] Figure 2 A flowchart of a method for recommending educational content based on a knowledge graph according to an embodiment of the present disclosure is shown;

[0025] Figure 3 A schematic diagram showing a portion of an example knowledge graph constructed according to an embodiment of the present disclosure;

[0026] Figure 4 An example of searching in a knowledge graph according to a computing knowledge point recommendation strategy according to an embodiment of the present disclosure is shown;

[0027] Figure 5 An example of outputting information corresponding to a plurality of extended knowledge points in a list and a graph is shown;

[0028] Figure 6 A schematic block diagram of a system for recommending educational content based on a knowledge graph according to an embodiment of the present disclosure is shown;

[0029] Figure 7A schematic block diagram of an apparatus for recommending educational content based on a knowledge graph according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present disclosure.

[0031] It should be understood that the terms "include" and "comprising" used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0032] It should also be understood that the terms used in this disclosure are only for the purpose of describing specific embodiments and are not intended to limit the disclosure. As used in this disclosure and claims, the singular forms of "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in this disclosure and claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations.

[0033] As used in this specification and claims, the term "if" may be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" may be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0034] In addition, the use of ordinal numbers such as "first", "second" and "third" to modify objects in this document does not itself imply any priority, precedence or order of one object relative to another object, nor does it imply the temporal order in which the actions of the method are performed, but is merely used as a label to distinguish one object with a specific name from another object with the same name to distinguish these objects.

[0035] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as generally understood. It will be further understood that the terms used herein should be interpreted as having the same meaning as in the context of this specification and the relevant technical field, and should not be interpreted in a limiting sense unless explicitly defined as such in this article.

[0036] Figure 2 It is a flowchart of a method 200 for recommending educational content based on a knowledge graph according to an embodiment of the present disclosure, wherein the knowledge graph is constructed in a knowledge base containing information and examples corresponding to a plurality of knowledge points, so that a plurality of entities and edges of the knowledge graph respectively represent the plurality of knowledge points and their mutual relations, and the information corresponding to the knowledge point represented by each entity is stored as an attribute of the entity. The method 200 comprises: a step S201 in which a user inputs a recommendation request, the recommendation request comprising a recommendation strategy and data corresponding to a starting knowledge point; a step S202 in which a starting entity representing the starting knowledge point is identified in the knowledge graph based on the data; a step S203 in which one or more additional entities are searched in the knowledge graph along one or more edges starting from the starting entity according to the recommendation strategy; and a step S204 in which information or examples corresponding to the starting knowledge point and one or more extended knowledge points represented by the one or more additional entities in the knowledge base are output as recommended educational content. In order to implement the input in the above step S201, a graphical user interface may be provided, for example, for the user to input the recommendation request. For example, on the graphical user interface, the user can select or directly input one of multiple predetermined recommendation strategies as the above-mentioned recommendation strategy, and can select one of multiple data corresponding to multiple knowledge points in advance as the data corresponding to the starting knowledge point or directly input the data corresponding to the starting knowledge point, and then initiate a recommendation request with the selected recommendation strategy and the selected data on the graphical user interface, so that the recommendation request is input through the graphical user interface. It can be understood by those skilled in the art that there may be different ways for the user to input the recommendation request than setting the graphical user interface, for example, in the form of voice interaction. The scheme disclosed herein has no restrictions on how the user inputs the recommendation request.

[0037] The knowledge graph constructed above belongs to the knowledge graph of a specific field. Since the knowledge graph of a specific field is constructed for a specific field, only knowledge containing high accuracy can provide support for upper-level applications, so it is preferred to adopt a top-down construction method. When constructing a knowledge graph according to the method of the embodiment of the present disclosure, the construction process can be implemented, for example, semi-manually and semi-automatically, and the data source used can be, for example, mainly the People's Education Press textbooks and supplemented by online encyclopedia data. Since the content of the textbook is unstructured data, and it is difficult to find the text version of the content, it is mostly based on picture content, and ordinary entity extraction methods are difficult to implement, so the relevant content is mainly manually labeled. In the knowledge fusion and knowledge processing steps, automatic detection can be mainly used to merge duplicate knowledge points and detect ambiguous content. It can be understood that those skilled in the art can completely use other methods and process implementations to construct the knowledge graph, and the scheme disclosed in this regard is not limited.

[0038] Next, various embodiments of method 200 will be combined with Figures 3 to 5 It is further described below. It is understandable that although the following description is based on junior high school mathematics knowledge points, that is only an example, and the disclosed solution can be applied to various subjects such as Chinese, English, chemistry, physics, etc. The disclosed solution is not limited in this respect.

[0039] In one embodiment, the multiple knowledge points are divided into the following three types: concepts, knowledge and applications. Concepts refer to basic definitions, such as a certain mathematical definition, such as triangles, equations, etc.; knowledge refers to the extension of the aforementioned concepts, such as mathematical inferences, properties, theorems, etc., such as the parallel axiom, which is interpreted as "through a point outside a straight line, there is only one straight line parallel to this straight line"; and applications are related to question types, which refer to question types and test points associated with concepts and knowledge that appear in textbooks, such as translation drawing, which is the application of translation knowledge points in textbooks, corresponding to related drawing question types. The scheme disclosed in this disclosure is more refined in the classification of knowledge points than the classification in the prior art, which is conducive to more refined recommendation of educational content. It can be understood that the multiple knowledge points can also adopt other classifications, and the scheme disclosed in this disclosure is not limited to this. In addition, in the scheme disclosed in this disclosure, since the knowledge points are represented by entities in the knowledge graph, the entities can also be considered to be divided into the three types.

[0040] In one embodiment, the mutual relationship of knowledge points is divided into the following ten relationships: inclusion relationship, instance relationship, precondition relationship, property relationship, method relationship, classification relationship, representation relationship, calculation relationship, derived concept relationship and other relationships. Inclusion relationship refers to semantic inclusion, which indicates the relationship from generalization to specificity, such as the relationship between the relatively abstract category of "plane figure" and the specific figure of "triangle", and "triangle" can be further specific to "isosceles triangle" and "right triangle". Instance relationship indicates the application and question type of a certain knowledge, such as the knowledge of "parallel line determination method" to the question type of "parallel determination proof". Precondition relationship indicates the precondition knowledge point of a certain knowledge, that is, the previous knowledge point that should be learned first in order to master the postcondition knowledge according to the learning order. Property relationship connects a certain concept and its corresponding knowledge point, such as the relationship between "line segment" and "line segment axiom". Method relationship indicates the idea or method of solving a problem, such as the relationship between "factorization" and the specific method of "factorization". Classification relationship represents the relationship from the whole to the part, such as the relationship from "judgment of parallel lines" to the three specific judgment theorems of parallel line judgment (equal corresponding angles, two straight lines are parallel; equal interior angles, two straight lines are parallel, etc.). Representation relationship is the relationship between a certain knowledge point and its representation form, such as the relationship between a certain concept and its knowledge point related to measurement units, and the representation of functions, such as analytical expressions and images. Calculation relationship can be particularly applicable to the knowledge system of mathematics, which represents the connection between a certain mathematical concept and its calculation-related knowledge point, such as from "polynomial" to "addition and subtraction of polynomials". Derived concept represents the related concept of a knowledge point. The relationship between the two cannot be represented by inclusion relationship. They are essentially different concepts but often appear at the same time, such as "triangle" and "incenter of triangle". The essence of "incenter of triangle" is a point. Its prior knowledge is related to the inscribed circle, and the incenter of triangle and the two knowledge points of triangle are very related. At this time, the derived concept is used to connect the two. Other relationships are used to represent other relationships that do not match the above nine relationships, and their number is relatively small. The disclosed solution uses the above ten relationships between knowledge points to construct a knowledge graph that can more comprehensively reflect the teaching logic and thus can widely support various recommendation strategies. Figure 3 It is shown in ( Figure 3 Not all of the above ten relationships are shown, because the figure only shows a part of the example knowledge graph). It is worth noting that in practice, not all of the ten relationships between the knowledge points in the knowledge graph constructed in this way necessarily exist. This is because the knowledge graph does not contain enough knowledge points or the subject targeted by the knowledge graph is not mathematics (so there is at least no computational relationship between the knowledge points), but this does not affect the construction of the knowledge graph with the above ten relationships between knowledge points.

[0041] In one embodiment, the information corresponding to the knowledge point includes one or more of the following: the ID of the knowledge point, the name of the knowledge point, the analysis of the knowledge point, the type of the knowledge point, the name of the textbook from which the knowledge point comes, the textbook chapter from which the knowledge point comes, the textbook page number from which the knowledge point comes, and the source type of the knowledge point. The ID of a knowledge point refers to the identifier of the knowledge point in the knowledge base, and the knowledge point in the knowledge base can be uniquely found with the ID of the knowledge point. The name of a knowledge point is usually an abbreviation so that users can quickly understand the meaning of the knowledge point. The analysis of a knowledge point is the detailed content of the explanation of the knowledge point. For example, the analysis corresponding to the concept "root exponent" is its definition of "the cube root of a number a, expressed in Indicates, where a is the radicand and 3 is the root exponent". The type of knowledge point is, for example, one of the three types of concepts, knowledge, and applications mentioned above, or other suitable types. The scheme disclosed herein has no restrictions on the type of knowledge points. The name, chapter, and page number of the textbook from which the knowledge point comes record the location of the textbook where the knowledge point is located, which facilitates users to locate the knowledge point in the textbook. The source type of the knowledge point records the data source from which the knowledge point is obtained, such as textbooks, web pages, expansion materials, etc. In the scheme disclosed herein, since the information corresponding to the knowledge point is stored as the attribute of the entity representing the knowledge point, the above information can also be considered as the attribute of the entity.

[0042] In one embodiment, the data corresponding to the starting knowledge point included in the recommendation request is the ID of the starting knowledge point, the name of the starting knowledge point, the condition corresponding to the starting knowledge point, or the ID of the example corresponding to the starting knowledge point. For example, when the user has previously learned the ID of a certain recommended knowledge point through method 200, the user can include the ID of the knowledge point as the ID of the starting knowledge point in the recommendation request. The user can also select or directly enter the name of the starting knowledge point through, for example, a graphical user interface. In another example, the user can set the condition corresponding to the knowledge point on, for example, a graphical user interface, such as the name and chapter of the textbook, or the name and page number of the textbook, or any appropriate condition corresponding to the knowledge point can be found based on the information corresponding to the knowledge point in the knowledge base, and then include the condition in the recommendation request. In another example, when the user has previously learned the ID of a certain recommended example (as described below) through method 200, the user can include the ID of the example as the ID of the example corresponding to the starting knowledge point in the recommendation request.

[0043] In one embodiment, identifying the starting entity representing the starting knowledge point in the knowledge graph based on the data includes: in response to the data corresponding to more than one knowledge point, prompting the user to specify one of the more than one knowledge point as the starting knowledge point. Since the scheme disclosed herein does not require that the data included in the recommendation request must correspond uniquely to the starting knowledge point, it is possible that the data corresponds to more than one knowledge point. For example, as described above, the user may enter a condition corresponding to the starting knowledge point (such as the name and chapter of the textbook, or the name and page number of the textbook) on, for example, a graphical user interface, or enter the ID of a certain example question, and more than one knowledge point may be found based on the condition or the ID of the example question. At this time, a window displaying the names of these knowledge points may pop up on the graphical user interface, prompting the user to specify one of these knowledge points on the window as the starting knowledge point. It can be understood by those skilled in the art that there may be a different way from the pop-up window to prompt the user to specify the starting knowledge point, for example, in the form of voice interaction. The scheme disclosed herein has no restrictions on how to prompt the user to specify the starting knowledge point.

[0044] In one embodiment, the recommendation strategy is one of the following strategies: recommendation of pre-knowledge points, recommendation of post-knowledge points, recommendation of calculation knowledge points, recommendation of knowledge points of the same level and recommendation of related concepts. Recommendation of pre-knowledge points is a strategy to obtain knowledge points before the starting knowledge points for users in accordance with the order of textbook learning. This strategy can facilitate users to consolidate and review the pre-content of the current knowledge points, so as to better grasp and understand the content of the current learning. Recommendation of post-knowledge points is opposite to recommendation of pre-knowledge points. It is a strategy to recommend knowledge points after the starting knowledge points for users in accordance with the order of textbook learning, so as to facilitate users to preview knowledge after the knowledge points currently being learned, for example. Recommendation of calculation knowledge points is a strategy particularly suitable for the field of mathematics learning. This recommendation strategy is mainly used to recommend relevant knowledge points about calculation of a mathematical concept as a starting knowledge point. For example, the concept of "polymorphism" in the seventh-grade textbook of the People's Education Press has related calculation knowledge points such as "polymorphism addition and subtraction", "polymorphism multiplication and division", "polymorphism simplification and evaluation", etc., so as to facilitate users to learn this type of calculation knowledge in a unified manner. The recommendation of knowledge points of the same level is a strategy for recommending knowledge points of the same level as the starting knowledge point, which can help users to master the knowledge points of the same type more comprehensively, wherein the knowledge points of the same level refer to multiple knowledge points represented by multiple entities that have the same parent entity and whose relationship belongs to a certain specific type in the knowledge graph. For example, the undetermined coefficient method and the cross multiplication method under factorization belong to the knowledge points of the same level. If the user specifies the recommendation strategy of knowledge points of the same level and the data corresponding to the undetermined coefficient method (the undetermined coefficient method is the starting knowledge point in this example) (such as its ID) in the recommendation request, then in response to the recommendation request, method 200 can output the information or examples corresponding to the cross multiplication method. Related concept recommendation is a strategy for recommending other knowledge points related to the starting knowledge point, which can enable users to deepen their understanding of the relationship between different knowledge points, help form a mind map for users, and help users sort out the knowledge they have learned. It can be understood by those skilled in the art that users can include any other appropriate recommendation strategies different from these recommendation strategies in the recommendation request. The scheme disclosed herein has no restrictions on what recommendation strategies users specifically specify.

[0045] The search in the knowledge graph according to the recommendation strategy may adopt certain rules, such as one or more of the rules in the above-mentioned embedding-based method, path-based method and joint method, or one or more of the variations of these rules, or any rule suitable for making recommendations.

[0046] In one embodiment, one or more rules used in the search are automatically determined according to the recommendation strategy. The rules determined in this way will enable the search to better meet the unique requirements of each recommendation strategy. For example, in one example, the rules used in the search are automatically determined as Rule 1, Rule 2, and Rule 3 according to the pre-knowledge point recommendation strategy, automatically determined as Rule 2 according to the post-knowledge point recommendation strategy, automatically determined as Rule 1, Rule 3, and Rule 4 according to the computing knowledge point recommendation strategy, automatically determined as Rule 2, Rule 3, and Rule 4 according to the same-level knowledge point recommendation strategy, and automatically determined as Rule 1 and Rule 3 according to the related concept recommendation strategy. Figure 4 An example of searching in a knowledge graph according to the computing knowledge point recommendation strategy according to the disclosed embodiment is shown, where the square entity (whose name is "formula") is the starting entity, and all the circular entities are multiple additional entities found (which represent multiple extended knowledge points). Figure 4 It can be seen that according to the computing knowledge point recommendation strategy, three rules (shown as "R0", "R1" and "R2" in the figure) are automatically determined and multiple additional entities are found using the three rules.

[0047] In addition, in the solution disclosed herein, the one or more rules used in the search may also be fixed rules regardless of the current recommendation strategy. As long as the rules can meet the requirements of the recommendation strategy, the solution disclosed herein has no restrictions on the use of the rules.

[0048] In one embodiment, the recommendation request further specifies whether the educational content to be recommended is information corresponding to the extended knowledge point or examples corresponding to the extended knowledge point. In other words, the recommendation request from the user can not only specify the recommendation strategy and the data corresponding to the starting knowledge point, but also specify the type of educational content to be recommended. If the recommendation request further specifies that the educational content to be recommended is information corresponding to the extended knowledge point, the method 200 will output the information; if the recommendation request further specifies that the educational content to be recommended is examples corresponding to the extended knowledge point, the method 200 will output the examples. Of course, in the scheme of the present disclosure, the recommendation request from the user may not specify the type of educational content to be recommended, and in this case, the method 200 may automatically select to output the information or examples according to, for example, a pre-configuration. The scheme of the present disclosure is not limited in this respect.

[0049] In one embodiment, in the case of outputting information corresponding to the one or more extended knowledge points, method 200 further includes: in response to a further request of the user, further outputting examples corresponding to the one or more extended knowledge points. For example, after the user obtains information corresponding to one or more extended knowledge points through the recommendation of method 200, the user may initiate a further request on, for example, a graphical user interface, and method 200 will further output examples corresponding to the one or more extended knowledge points in response to the further request. In one example, if the number of examples corresponding to the one or more extended knowledge points is too large, examples may be screened, for example, according to weights before further output (the weights may be obtained, for example, according to the recommendation strategy) to ensure that the number of examples further output does not exceed a certain number.

[0050] In one embodiment, the information corresponding to the one or more extended knowledge points is output in a list and / or a graph. Figure 5 An example of outputting the information in a list and a graph is shown, wherein the graph is shown on the left side of the figure and the list is shown on the right side. The user can switch between the output of the list and the output of the graph by himself with the help of, for example, a graphical user interface. In the graph, the user can, for example, hover the mouse over an entity to cause a small window to pop up to display the information corresponding to the knowledge point represented by the entity, or can, for example, click on an entity to automatically update the graph to display an appropriate number of entities near the entity as the center. It can be understood that the scheme of the present disclosure can also use other forms to output the information. In terms of the form of outputting the information, the scheme of the present disclosure is not limited.

[0051] In one embodiment, the knowledge base also stores one or more of the ID, answer, analysis and difficulty of the example questions, and method 200 also includes: when outputting the example questions corresponding to the one or more extended knowledge points, simultaneously outputting or outputting one or more of the ID, answer, analysis and difficulty of the example questions corresponding to the one or more extended knowledge points in response to further request from the user.

[0052] Accordingly, if Figure 6As shown, the embodiment of the present disclosure also provides a system 600 for recommending educational content based on a knowledge graph, wherein the knowledge graph is constructed in a knowledge base containing information and examples corresponding to multiple knowledge points, so that multiple entities and edges of the knowledge graph respectively represent the multiple knowledge points and their mutual relationships, and the information corresponding to the knowledge point represented by each entity is stored as the attribute of the entity. The system 600 includes: an input module 601, for inputting a recommendation request by a user, the recommendation request including a recommendation strategy and data corresponding to a starting knowledge point; an identification module 602, for identifying a starting entity representing the starting knowledge point in the knowledge graph based on the data; a search module 603, for searching one or more additional entities in the knowledge graph along one or more edges starting from the starting entity according to the recommendation strategy; and an output module 604, for outputting information or examples corresponding to the starting knowledge point and one or more extended knowledge points represented by the one or more additional entities in the knowledge base as recommended educational content. The various embodiments of the system 600 correspond to the various embodiments of the method 200, and therefore are not described repeatedly here.

[0053] It is understandable that in order to ensure the normal operation of system 600, system 600 must be able to access the knowledge base, but it does not necessarily include the knowledge base. In other words, as long as system 600 is able to access the knowledge base, the solution disclosed herein has no restrictions on whether system 600 includes the knowledge base. In addition, system 600 can also implement a simple query function (without recommendation) for knowledge points and test questions. For example, in response to the ID of a knowledge point, the name of a knowledge point, or the ID of a test question input by a user, system 600 can only output the information corresponding to the knowledge point or the test question (optionally with information such as the answer, analysis, and / or difficulty of the example question). Those skilled in the art fully understand how to implement this query function, so it will not be described in detail.

[0054] It is understandable that the system 600 described herein can be implemented by various modules, so that the system 600 that implements one or more functions described in the above embodiments can include not only Figure 6The modules shown in , may also include other modules for realizing one or more other functions thereof. In addition, system 600 may include a single module configured to perform two or more functions, or a separate module for respective corresponding functions. In addition, these modules may include hardware, firmware, software or any combination thereof. For example, the input module 601 and output module 604 of system 600 may each include an exclusive graphical user interface for the input and the output, respectively, or the input module 601 and output module 604 may share a graphical user interface for the input and the output, or the input module 601 and output module 604 may adopt a voice interaction mode for the input and the output. The scheme disclosed herein is not limited in this respect.

[0055] It should be understood that the blocks of the block diagram and / or flowchart illustration and the combination of blocks in the block diagram and / or flowchart illustration can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer and / or other programmable data processing device, so that the instructions executed via the processor of the computer and / or other programmable data processing device create components for implementing the functions / actions specified in the blocks of the block diagram and / or flowchart.

[0056] It should also be understood that the functions / actions shown in the blocks of the flow chart may not occur in the order shown in the diagram. For example, depending on the functions / actions involved, two blocks shown in succession may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Although some figures may include arrows on communication paths to illustrate the primary direction of communication, it should be understood that communication may occur in the direction opposite to the arrows depicted.

[0057] In addition, the disclosed embodiments may take the form of a computer program on a memory having a computer usable or computer readable program code implemented in a medium for use by or in conjunction with an instruction execution system. In the context of this document, a memory may be any medium that may contain, store, or be suitable for transmitting a program for use by or in conjunction with an instruction execution system, device, or apparatus.

[0058] Therefore, the present disclosure also provides an apparatus 700 including a processor 701 and a memory 702, such as Figure 7 In the device 700, the memory 702 stores instructions, which, when executed by the processor 701, enable the device 700 to perform the above-mentioned method for recommending educational content based on the knowledge graph.

[0059] The present disclosure also provides a machine-readable storage medium (not shown) having instructions stored thereon, which, when executed on a device, causes the device to execute the above-mentioned method of recommending educational content based on a knowledge graph.

[0060] Although this specification contains many specific implementation details, these details should not be interpreted as limitations on the scope of any implementation or what may be claimed, but rather as descriptions of features of specific embodiments that may be specific to a particular implementation. Certain features described in this specification in the context of various embodiments may also be implemented in combination in a single embodiment. Conversely, the various features described in the context of a single embodiment may also be implemented in multiple embodiments individually or in any suitable sub-combination. In addition, although features may be described above as working in certain combinations, even when originally claimed, in some cases, one or more features from the claimed combination may be deleted from the combination, and the claimed combination may be directed to a sub-combination or a variant of a sub-combination.

[0061] Although multiple embodiments of the present disclosure have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art may think of many changes, modifications, and alternatives without departing from the thought and spirit of the present disclosure. It should be understood that in the process of practicing the present disclosure, various alternatives to the embodiments of the present disclosure described herein may be adopted. The attached claims are intended to define the scope of protection of the present disclosure, and therefore cover equivalents or alternatives within the scope of these claims. The collection and acquisition of various data in this application complies with relevant laws and regulations and is authorized by the data provider. Any organization or individual that needs to obtain external data must obtain authorization in accordance with the law and ensure data security. It is not allowed to illegally collect, use, process, or transmit unauthorized or unprotected data, or to illegally buy, sell, provide, or disclose unauthorized or unprotected data.

Claims

1. A method for recommending educational content based on a knowledge graph, wherein the knowledge graph is constructed in a knowledge base containing information and examples corresponding to a plurality of knowledge points, so that a plurality of entities and edges of the knowledge graph respectively represent the plurality of knowledge points and their mutual relationships, and the information corresponding to the knowledge point represented by each entity is stored as an attribute of the entity, the method comprising: A user inputs a recommendation request, wherein the recommendation request includes a recommendation strategy and data corresponding to a starting knowledge point; Identify a starting entity representing the starting knowledge point in the knowledge graph based on the data; According to the recommendation strategy, starting from the starting entity, searching for one or more additional entities along one or more edges in the knowledge graph; as well as The information or examples in the knowledge base corresponding to the starting knowledge point and the one or more extended knowledge points represented by the one or more additional entities are output as recommended educational content.

2. The method of claim 1, wherein: The multiple knowledge points are divided into the following three types: concept, knowledge and application.

3. The method according to any one of claims 1 to 2, wherein: The mutual relationships of the multiple knowledge points are divided into the following ten relationships: inclusion relationship, instance relationship, pre-requisite relationship, property relationship, method relationship, classification relationship, representation relationship, calculation relationship, derived concept relationship and other relationships.

4. The method according to any one of claims 1 to 3, wherein: The information corresponding to the knowledge point includes one or more of the following: the ID of the knowledge point, the name of the knowledge point, the analysis of the knowledge point, the type of the knowledge point, the name of the textbook from which the knowledge point comes, the chapter of the textbook from which the knowledge point comes, the page number of the textbook from which the knowledge point comes, and the source type of the knowledge point.

5. The method according to any one of claims 1 to 4, wherein: The data is the ID of the starting knowledge point, the name of the starting knowledge point, the condition corresponding to the starting knowledge point, or the ID of the example question corresponding to the starting knowledge point.

6. The method according to any one of claims 1 to 5, wherein: Identifying a starting entity representing the starting knowledge point in the knowledge graph based on the data includes: in response to the data corresponding to more than one knowledge point, prompting the user to designate one of the more than one knowledge point as the starting knowledge point.

7. The method according to any one of claims 1 to 6, wherein: The recommendation strategy is one of the following strategies: previous knowledge point recommendation, subsequent knowledge point recommendation, computing knowledge point recommendation, same-level knowledge point recommendation and related concept recommendation.

8. The method according to any one of claims 1 to 7, wherein: One or more rules used in the search are automatically determined according to the recommendation strategy.

9. The method according to any one of claims 1 to 8, wherein: The recommendation request further specifies whether the educational content to be recommended is information corresponding to the extended knowledge point or examples corresponding to the extended knowledge point.

10. The method according to any one of claims 1 to 9, wherein: In the case of outputting information corresponding to the one or more extended knowledge points, the method further includes: in response to a further request of the user, further outputting example questions corresponding to the one or more extended knowledge points.

11. The method according to any one of claims 1 to 10, wherein: The information corresponding to the one or more extended knowledge points is output in a list and / or a graph.

12. The method according to any one of claims 1 to 11, wherein: The knowledge base also stores one or more of the ID, answer, analysis and difficulty of the example questions, and the method also includes: when outputting the example questions corresponding to the one or more extended knowledge points, simultaneously outputting or outputting one or more of the ID, answer, analysis and difficulty of the example questions corresponding to the one or more extended knowledge points in response to further requests from the user.

13. A system for recommending educational content based on a knowledge graph, wherein the knowledge graph is constructed in a knowledge base containing information and examples corresponding to a plurality of knowledge points, so that a plurality of entities and edges of the knowledge graph respectively represent the plurality of knowledge points and their mutual relationships, and the information corresponding to the knowledge point represented by each entity is stored as an attribute of the entity, the system comprising: An input module, used for a user to input a recommendation request, wherein the recommendation request includes a recommendation strategy and data corresponding to a starting knowledge point; An identification module, configured to identify a starting entity representing the starting knowledge point in the knowledge graph based on the data; A search module, configured to search for one or more additional entities along one or more edges in the knowledge graph starting from the starting entity according to the recommendation strategy; as well as An output module is used to output information or examples in the knowledge base corresponding to the starting knowledge point and one or more extended knowledge points represented by the one or more additional entities as recommended educational content.

14. The system of claim 13, wherein: The multiple knowledge points are divided into the following three types: concept, knowledge and application.

15. The system of any one of claims 13-14, wherein: The mutual relationships of the multiple knowledge points are divided into the following ten relationships: inclusion relationship, instance relationship, pre-requisite relationship, property relationship, method relationship, classification relationship, representation relationship, calculation relationship, derived concept relationship and other relationships.

16. The system of any one of claims 13-15, wherein: The information corresponding to the knowledge point includes one or more of the following: the ID of the knowledge point, the name of the knowledge point, the analysis of the knowledge point, the type of the knowledge point, the name of the textbook from which the knowledge point comes, the chapter of the textbook from which the knowledge point comes, the page number of the textbook from which the knowledge point comes, and the source type of the knowledge point.

17. The system of any one of claims 13 to 16, wherein: The data is the ID of the starting knowledge point, the name of the starting knowledge point, the condition corresponding to the starting knowledge point, or the ID of the example question corresponding to the starting knowledge point.

18. The system of any one of claims 13-17, wherein: Identifying a starting entity representing the starting knowledge point in the knowledge graph based on the data includes: in response to the data corresponding to more than one knowledge point, prompting the user to designate one of the more than one knowledge point as the starting knowledge point.

19. The system of any one of claims 13 to 18, wherein: The recommendation strategy is one of the following strategies: previous knowledge point recommendation, subsequent knowledge point recommendation, computing knowledge point recommendation, same-level knowledge point recommendation and related concept recommendation.

20. The system of any one of claims 13-19, wherein: One or more rules used in the search are automatically determined according to the recommendation strategy.

21. The system of any one of claims 13-20, wherein: The recommendation request further specifies whether the educational content to be recommended is information corresponding to the extended knowledge point or examples corresponding to the extended knowledge point.

22. The system of any one of claims 13-21, wherein: In the case of outputting information corresponding to the one or more extended knowledge points, the output module is further used to: in response to a further request from the user, further output example questions corresponding to the one or more extended knowledge points.

23. The system of any one of claims 13-22, wherein: The information corresponding to the one or more extended knowledge points is output in a list and / or a graph.

24. The system of any one of claims 13-23, wherein: The knowledge base also stores one or more of the ID, answer, analysis and difficulty of the example questions, and the output module is also used to: when outputting the example questions corresponding to the one or more extended knowledge points, simultaneously output or output one or more of the ID, answer, analysis and difficulty of the example questions corresponding to the one or more extended knowledge points in response to further requests from the user.

25. A device for recommending educational content based on a knowledge graph, the device comprising a processor and a memory storing instructions, wherein the instructions, when executed by the processor, cause the device to execute a method as described in any one of claims 1-12.

26. A machine-readable storage medium storing instructions, wherein when the instructions are executed on a device, the device performs the method according to any one of claims 1 to 12.

Images