Knowledge graph-based learning path analysis method, apparatus and device, and medium

Abstract

The invention discloses a knowledge graph-based learning path analysis method, device and equipment and a medium, and relates to the field of online learning platform analysis, and the method comprises the steps: determining a knowledge behavior event and a learning behavior event according to user operation information and a knowledge graph of an online learning platform, thereby determining structured user operation record data, determining a knowledge point learning sequence and a learning behavior sequence by adopting a timestamp sequence analysis method so as to determine a learning path basic sequence; extracting behavior pattern features according to the learning behavior event; according to the relationship among the nodes in the knowledge graph, a hidden Markov model is adopted to calculate the knowledge correlation degree of the learning path; and determining a learning path representation model according to the behavior pattern characteristics, the learning path knowledge association degree and the learning path basic sequence, so that the learning behavior pattern of the user can be combined with the knowledge point learning sequence, and the association degree between the learning behavior and the knowledge points and the association degree between the knowledge points before and after can be evaluated.

Description

Technical Field

[0001] The present application relates to the field of online learning platform analysis, and in particular to a learning path analysis method, device, equipment and medium based on knowledge graph. Background Art

[0002] In relevant online learning platforms, the learning path of users (for example, students) is usually limited to a simple record of access logs, lacking comprehensive consideration and in-depth analysis of knowledge point associations and learning behaviors. When analyzing users' learning behaviors, relevant systems pay more attention to surface data such as the number of behaviors and page access time, making it difficult to accurately portray the learner's knowledge acquisition process. Most platforms currently lack suitable technical solutions for effectively combining users' learning behavior patterns with the order of knowledge point learning. In addition, relevant learning path analysis methods are usually unable to dynamically evaluate the correlation between learning behaviors and knowledge points, resulting in limited effects of path optimization and personalized teaching recommendations based on this. Summary of the invention

[0003] The purpose of this application is to provide a learning path analysis method, device, equipment and medium based on knowledge graph, which can combine the user's learning behavior pattern with the knowledge point learning order, and can evaluate the correlation between learning behavior and knowledge points and the correlation between previous and subsequent knowledge points, so as to improve the user's learning efficiency and learning effect.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a learning path analysis method based on a knowledge graph, comprising:

[0006] Obtaining user operation information of an online learning platform; the online learning platform has a knowledge graph; the nodes in the knowledge graph represent knowledge points in the online learning platform; the lines between the nodes in the knowledge graph represent the relationships between the knowledge points in the online learning platform;

[0007] Determine knowledge behavior events and learning behavior events according to the user operation information and the knowledge graph, and determine structured user operation record data according to the knowledge behavior events and the learning behavior events;

[0008] According to the structured user operation record data, the time stamp sequence analysis method is used to determine the knowledge point learning sequence and learning behavior sequence;

[0009] Determine a basic sequence of learning paths according to the knowledge point learning sequence and the learning behavior sequence;

[0010] Extracting behavior pattern features according to the learning behavior events; the behavior pattern features include: behavior frequency, behavior duration and behavior conversion rate;

[0011] Calculate the knowledge relevance of the learning path using a hidden Markov model based on the relationship between nodes in the knowledge graph;

[0012] A learning path representation model is determined based on the behavior pattern characteristics, the learning path knowledge association and the learning path basic sequence; the learning path representation model is used to characterize the correspondence between the user's learning behavior pattern and the knowledge point learning order, the association between the learning behavior and the knowledge points, and the association between the knowledge points.

[0013] In a second aspect, the present application provides a learning path analysis device based on a knowledge graph, comprising:

[0014] An information acquisition module, used to acquire user operation information of an online learning platform; the online learning platform has a knowledge graph; the nodes in the knowledge graph represent knowledge points in the online learning platform; the lines between the nodes in the knowledge graph represent the relationship between the knowledge points in the online learning platform;

[0015] A structured data determination module, used to determine knowledge behavior events and learning behavior events according to the user operation information and the knowledge graph, and determine structured user operation record data according to the knowledge behavior events and the learning behavior events;

[0016] A knowledge point and learning behavior sequence extraction module is used to determine the knowledge point learning sequence and learning behavior sequence using a timestamp sequence analysis method based on structured user operation record data;

[0017] A learning path basic sequence determination module, used to determine a learning path basic sequence according to the knowledge point learning sequence and the learning behavior sequence;

[0018] A behavior pattern feature extraction module, used to extract behavior pattern features according to the learning behavior events; the behavior pattern features include: behavior frequency, behavior duration and behavior conversion rate;

[0019] A knowledge relevance calculation module, used to calculate the knowledge relevance of the learning path using a hidden Markov model according to the relationship between the nodes in the knowledge graph;

[0020] A learning path characterization module is used to determine a learning path characterization model based on the behavior pattern characteristics, the learning path knowledge association and the learning path basic sequence; the learning path characterization model is used to characterize the correspondence between the user's learning behavior pattern and the knowledge point learning sequence, the association between the learning behavior and the knowledge points, and the association between the knowledge points.

[0021] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the above-mentioned learning path analysis methods based on knowledge graphs.

[0022] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned knowledge graph-based learning path analysis methods.

[0023] According to the specific embodiments provided in this application, this application has the following technical effects:

[0024] The present application provides a learning path analysis method, apparatus, device and medium based on knowledge graph, which is based on user operation information and introduces knowledge graph to determine the basic sequence of learning path, realizes the mining of learning path, extracts the behavior pattern characteristics in learning behavior events, and calculates the knowledge association of learning path based on the relationship between nodes in the knowledge graph using hidden Markov model, determines the learning path representation model based on the behavior pattern characteristics, the knowledge association of learning path and the basic sequence of learning path, and realizes the representation of learning path. The present application can not only combine the user's learning behavior pattern with the learning order of knowledge points, but also evaluate the association between learning behavior and knowledge points and the association between knowledge points, guides user learning according to the learning path representation model, and helps to improve the user's learning efficiency and learning effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 This is an application environment diagram of a learning path analysis method based on a knowledge graph in one embodiment of the present application;

[0027] Figure 2 A flowchart of a learning path analysis method based on a knowledge graph provided in one embodiment of the present application;

[0028] Figure 3 A diagram of the implementation process of a learning path analysis method based on a knowledge graph in a practical application provided in an embodiment of the present application;

[0029] Figure 4A schematic diagram of functional modules of a learning path analysis device based on a knowledge graph provided in another embodiment of the present application;

[0030] Figure 5 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0031] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0032] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0033] In this application, all actions to obtain signals, information or data are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0034] Based on the problem that the data analysis of the current online learning platform is superficial and lacks personalized path analysis, the embodiment of the present application proposes a learning path analysis method based on the knowledge graph, and realizes learning path mining and learning path characterization based on the knowledge graph. With the rapid development of smart education technology, how to use knowledge graphs and advanced data mining technology to deeply analyze the correlation between students' learning behaviors and knowledge points has become an important direction of personalized education research. Under this premise, the embodiment of the present application proposes a complete set of learning path analysis methods based on the node relationship of the user's learning behavior data and the knowledge graph, and solves the problem that the correlation of knowledge points in the learning path cannot be accurately evaluated by introducing the hidden Markov model (HMM) and the transition probability matrix. By conducting a multi-dimensional quantitative analysis of learning behavior, not only the accuracy of the learning path is improved, but also a basis is provided for subsequent path optimization and personalized teaching strategies.

[0035] The learning path analysis method based on knowledge graph provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send user operation information to the server 104. After the server 104 receives the user operation information, for the user operation information, the server 104 determines the knowledge behavior event and the learning behavior event based on the user operation information and the knowledge graph, and determines the structured user operation record data based on the knowledge behavior event and the learning behavior event; according to the structured user operation record data, the timestamp sequence analysis method is used to determine the knowledge point learning sequence and the learning behavior sequence; according to the knowledge point learning sequence and the learning behavior sequence, the basic sequence of the learning path is determined; according to the learning behavior event, the behavior pattern features are extracted; according to the learning behavior event, the hidden Markov model is used to calculate the knowledge association of the learning path based on the relationship between the nodes in the knowledge graph; according to the behavior pattern features, the knowledge association of the learning path and the basic sequence of the learning path, the learning path representation model is determined.

[0036] The server 104 may feed back the obtained learning path representation model to the terminal 102. In addition, in some embodiments, the learning path analysis method based on the knowledge graph may also be implemented by the server 104 or the terminal 102 alone, such as the terminal 102 may directly process the user operation information, or the server 104 may obtain the user operation information from the data storage system and process the user operation information.

[0037] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0038] In an exemplary embodiment, Figure 2 As shown, a learning path analysis method based on knowledge graph is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following steps 201 to 207. Among them:

[0039] Step 201, obtaining user operation information of an online learning platform; the online learning platform has a knowledge graph.

[0040] Specifically, the online learning platform may be a smart education online learning platform with a course knowledge graph or a professional knowledge graph learning environment, and may be capable of monitoring and storing user operation information.

[0041] Wherein, the nodes in the knowledge graph represent the knowledge points in the online learning platform; the lines between the nodes in the knowledge graph represent the relationships between the knowledge points in the online learning platform. Wherein, the relationships between knowledge points include: prerequisite relationships, subsequent relationships, inclusion relationships and similarity relationships. Prerequisite relationship: Node A is the prerequisite knowledge of Node B, and learning Node A helps to understand Node B. Subsequent relationship: Node A is the subsequent knowledge of Node B, and Node A depends on the understanding of Node B to learn. Inclusion relationship: Some knowledge nodes are included in the content of other nodes. Similarity relationship: There are similarities between knowledge nodes, but they do not completely overlap.

[0042] Step 202, determining knowledge behavior events and learning behavior events based on the user operation information and the knowledge graph, and determining structured user operation record data based on the knowledge behavior events and the learning behavior events.

[0043] Step 203: Determine the knowledge point learning sequence and the learning behavior sequence by using a timestamp sequence analysis method according to the structured user operation record data.

[0044] Step 204: Determine a basic sequence of a learning path according to the knowledge point learning sequence and the learning behavior sequence.

[0045] Step 205: extracting behavior pattern features according to the learning behavior events.

[0046] The behavior pattern characteristics include: behavior frequency, behavior duration and behavior conversion rate.

[0047] Step 206, using a hidden Markov model to calculate the knowledge association of the learning path according to the relationship between the nodes in the knowledge graph.

[0048] Step 207: determining a learning path representation model according to the behavior pattern characteristics, the learning path knowledge association degree and the learning path basic sequence.

[0049] The learning path representation model is used to represent the correspondence between the user's learning behavior pattern and the knowledge point learning sequence, the correlation between the learning behavior and the knowledge points, and the correlation between the knowledge points.

[0050] When analyzing user learning behaviors, related systems focus more on surface data such as the number of behaviors and page access time, lacking in-depth integration with professional knowledge graphs, making it difficult to accurately characterize the learner's knowledge acquisition process. By implementing the above steps 201 to 207, the learning path representation model determined based on the knowledge graph can combine the user's learning behavior pattern with the order of knowledge point learning, and can evaluate the correlation between learning behavior and knowledge points. It can guide user learning according to the learning path representation model, which helps to improve the user's learning efficiency and learning effect.

[0051] In another exemplary embodiment of the present application, step 202 specifically includes:

[0052] (1) Determine the operation events in the user learning process according to the user operation information. The event is defined as: based on the functional characteristics of the online learning platform, define the operation events such as node access, function use, login and logout in the user learning process.

[0053] (2) Determine the knowledge behavior event and the learning behavior event based on whether the operation event directly acts on a node in the knowledge graph.

[0054] (3) Collecting the basic attributes of the knowledge behavior event to obtain knowledge behavior operation data, and collecting the basic attributes of the learning behavior event to obtain learning behavior operation data. The basic attributes include: user ID, access page, page type, and access time.

[0055] (4) Performing data structuring processing on the knowledge behavior operation data and the learning behavior operation data respectively to obtain structured knowledge behavior operation data and structured learning behavior operation data; the structured user operation record data includes: structured knowledge behavior operation data and structured learning behavior operation data.

[0056] The main purpose of the structured processing of user operation records is to clean irrelevant data and calculate the duration of each operation behavior. The data structured processing operation process can be: export user operation data record files (knowledge behavior operation data and learning behavior operation data) - filter irrelevant page data based on page type data - remove blank and duplicate data - sort data by user ID and access time - calculate the duration of the previous behavior through the time difference between adjacent data - match data with events according to the definition of events - obtain structured user operation record data and store it in CSV format.

[0057] In another exemplary embodiment of the present application, step 203 specifically includes:

[0058] (1) Extracting a knowledge point set and a learning behavior set based on structured user operation record data; the knowledge point set includes: the knowledge point information accessed by the user in the knowledge graph; the learning behavior set includes: the operation behavior information performed by the user during the learning process.

[0059] Extraction of knowledge point information: Based on the node access information in the knowledge graph, extract the set of knowledge points accessed by the user P (N i ), this knowledge point set represents the knowledge content experienced by the user, and its basic identification is the page information of the node type in the "page type" label.

[0060] Operation behavior information extraction: Based on the event definition formed by the time difference of adjacent data, the operation behavior information (such as exercises, data downloads, learning interruptions, etc.) performed by the user during the learning process is extracted to obtain the learning behavior set Q (A i ).

[0061] (2) The knowledge point set is sorted according to the access time points using a timestamp sequence analysis method to obtain a knowledge point learning sequence.

[0062] Sequential analysis based on timestamp: First, the set of knowledge points visited by the user P(N i ) are sorted according to the access time point t to form a knowledge point learning sequence. i The access nodes at time points t1, t2, and t3 are N1, N2, and N3, and the order of their knowledge points is: N1→N2→N3, recorded as a vector: K=[N1 N2 N3], and the knowledge point learning sequence is expressed as K(N i ).

[0063] (3) Using a timestamp sequence analysis method, the learning behavior set is sorted according to the access time point to obtain a learning behavior sequence.

[0064] Sequential analysis based on timestamps: First, the learning behavior set Q(A i ) are sorted by access time point t to form a learning behavior sequence. i The behaviors performed at time points t1, t2, and t3 are A1, A2, and A3. The learning behavior sequence is: A1→A2→A3, recorded as a vector: T=[A1 A2 A3]. The learning behavior sequence is expressed as T(A1→A2→A3). i ).

[0065] In another exemplary embodiment of the present application, step 204 specifically includes:

[0066] (1) Integrating the knowledge point learning sequence and the learning behavior sequence in the form of a matrix to determine an integration matrix; the rows of the integration matrix represent the temporal relationship between each event, and the columns of the integration matrix represent the corresponding relationship between the learning behavior and the knowledge point.

[0067] Specifically, a 2×n matrix is ​​used to integrate the knowledge point learning sequence and the learning behavior sequence. The rows represent the temporal relationship of each event in the learning path, and the columns represent the corresponding relationship between the behavior and the knowledge point. The expression of the integration matrix is:

[0068] (2) The knowledge points that do not correspond to the learning behaviors in the integration matrix are completed to obtain a basic sequence of the learning path; the completed knowledge points in the basic sequence of the learning path correspond one-to-one to the learning behaviors.

[0069] Specifically, from a vertical perspective, all knowledge behaviors can correspond to specific knowledge node information, but non-knowledge behaviors (i.e., learning behaviors) cannot find corresponding knowledge nodes. In this embodiment, the vector can be padded with 0. When the same behavior interval corresponds to multiple knowledge points, the behavior is copied synchronously based on the number of knowledge points to extend the behavior sequence to correspond to the knowledge point sequence, and the learning path represented by the above integrated matrix is ​​corrected. The basic sequence representation of the learning path obtained after correction is:

[0070] In another exemplary embodiment of the present application, in step 205, feature extraction is performed on learning behavior events, and the behavior pattern is defined and quantified. The behavior frequency f represents the number of times a specific learning behavior occurs in the entire learning path, which is quantified by counting the frequencies of various behaviors in the learning behavior sequence. The behavior duration Δt represents the time a user stays on a certain learning behavior, which is equal to the time difference Δt between adjacent operation events. By matching the time difference with the behavior sequence, the duration of each behavior is obtained. The behavior conversion rate C represents the frequency of users switching between different behavior types, which is defined as the operation conversion frequency between discontinuous behaviors of the same type, and the calculation formula is as follows: n x Indicates the number of conversions of a certain behavior, and num indicates the total number of behaviors.

[0071] In another exemplary embodiment of the present application, step 206 specifically includes:

[0072] (1) Construct a transition probability matrix between knowledge points based on the relationship between nodes in the knowledge graph.

[0073] Specifically, suppose there are m nodes in the knowledge graph, and the weight parameter P for the mutual conversion between different nodes is defined according to the closeness of the relationship set in the knowledge graph. ij, and normalize it to form a (m+1)×(m+1) transition probability matrix, ensuring that the sum of the probabilities of each row and column is 1. The expression of the transition probability matrix is ​​as follows:

[0074]

[0075] (2) Based on the transition probability matrix, a hidden Markov model is used to calculate the strongest correlation transfer path between knowledge points.

[0076] Specifically, based on the transition probability matrix, the hidden Markov model is used to estimate the strongest associated transfer path of the knowledge point sequence actually learned by the user, and identify the knowledge path with the highest correlation in the knowledge graph. max (U i ), and use the transition probability matrix to calculate the overall transition probability NUM of the entire path max :

[0077] NUM max =P 12 +P 23 +……+P ij .

[0078] (3) Calculate the knowledge association degree of the learning path based on the strongest associated transfer path.

[0079] Specifically, the actual knowledge point path relevance is calculated: first, the user's actual knowledge point path Path (U i )’s overall transition probability NUM, and then calculates the knowledge path with the highest correlation Path max (U i ) Overall transition probability and user’s actual learning path Path(U i ) The difference Δd between the overall transition probabilities NUM. In actual situations, knowledge points may appear repeatedly. Therefore, when calculating the sum of transition probabilities, the repeated knowledge points need to calculate the maximum transition probability related to them before, rather than only considering the transfer relationship with the previous knowledge point. The final learning path knowledge association S is expressed as: S = Δd = NUM max -NUM.

[0080] In another exemplary embodiment of the present application, step 207 specifically includes:

[0081] (1) Integrate the behavior pattern features and the learning path knowledge correlation to obtain a feature vector. Specifically, normalize the behavior pattern features and the learning path knowledge correlation to obtain normalized behavior pattern features and normalized learning path knowledge correlation; integrate the normalized behavior pattern features and the normalized learning path knowledge correlation to obtain a feature vector.

[0082] The purpose of normalization in this step is to ensure that the learning paths and behavior characteristics of different users are comparable during analysis. The knowledge relevance of the learning path and the behavior pattern characteristics (behavior frequency, behavior duration, and behavior conversion rate) are normalized. The normalization formula is:

[0083]

[0084] After normalization, the values ​​of all behavior pattern features and learning path knowledge associations are limited to between 0 and 1, which facilitates the comparison and analysis of learning paths between different users. X represents the parameters before normalization, which can be f, Δt, C, S, X norm represents the normalized parameter, X min Indicates the minimum value of the parameter, X max Indicates the maximum value of the parameter.

[0085] The feature vector λ(t) formed by integrating the normalized behavior pattern features and the normalized learning path knowledge association corresponds to the timestamp data to form data that changes over time.

[0086] (2) Integrate the feature vector with the basic sequence of the learning path to obtain a learning path representation model.

[0087] Specifically, after the feature vector is integrated with the basic sequence Path of the learning path, the following learning path representation model is formed:

[0088] Combine the following Figure 3 , an implementation process of the above-mentioned knowledge graph-based learning path analysis method in practical applications is explained.

[0089] This method mines and characterizes students' learning paths based on professional knowledge maps, calculates the correlation between behavioral pattern characteristics and learning path knowledge based on a smart education platform based on professional knowledge maps, and comprehensively constructs a learning path representation model.

[0090] Step 1: Structuring user online operation data. 1) Define user operation events, where events are divided into knowledge behavior events and non-knowledge behavior events (i.e. learning behavior events) based on whether the events directly act on knowledge graph nodes, to achieve event classification. 2) User operation record collection: record the basic attributes of user online operation events. 3) Structural processing of user operation record data.

[0091] Step 2: Extract the user's knowledge point learning sequence and user learning behavior sequence information. 1) Extract knowledge point and behavior data information: According to the user's operation record on the smart education platform, the user's learning path is extracted according to the pre-defined events to obtain knowledge point information and operation behavior information, thereby determining the knowledge point set and learning behavior set. 2) Generation of knowledge point learning sequence and learning behavior sequence.

[0092] Step 3: Construct the basic sequence of learning paths. 1) Integrate the knowledge point learning sequence and the learning behavior sequence to obtain an integrated matrix. 2) Fill in non-knowledge behaviors to obtain the basic sequence of learning paths.

[0093] Step 4: Calculate the learning path behavior pattern characteristics and knowledge relevance. 1) Define and quantify behavior pattern characteristics. 2) Define the relationship between knowledge graph nodes (i.e., the relationship between knowledge points), including: prerequisite relationship, subsequent relationship, inclusion relationship, and similarity relationship. 3) Define the learning path knowledge relevance. Knowledge relevance measures the strength of the logical relationship between the knowledge points learned by the user. Construct a transition probability matrix between knowledge points based on the node relationship in the knowledge graph, use the hidden Markov model to calculate the strongest correlation transfer path between knowledge points, and calculate the correlation of the existing path (i.e., the learning path knowledge relevance) based on the strongest correlation transfer matrix. 4) Data standardization.

[0094] Step 5: Construct a learning path representation model, which serves as the result of learning path mining and representation.

[0095] The learning path analysis method based on knowledge graph provided in the embodiment of the present application mainly includes five steps: 1. Structuring user online operation data; 2. Extracting user knowledge point learning order and user learning behavior sequence information; 3. Constructing the basic sequence of learning path; 4. Calculating the learning path behavior pattern characteristics and knowledge correlation; 5. Constructing a learning path representation model. This method uses a 3×n matrix to represent the objective logical relationship between the user's subjective learning behavior and knowledge nodes around the professional knowledge graph, and includes three types of information: actions, knowledge, and path characteristics, as well as the time series of each type of information data. This method promotes the research on student learning path analysis and learning process monitoring, and solves two problems: 1) How to comprehensively represent the student's subjective behavior and knowledge object organization sequence; 2) How to measure the characteristics of students' learning paths from multiple dimensions.

[0096] The above method is mainly reflected in the path mining stage and the path characterization stage: 1) In the learning path mining stage, user data is obtained through a specific operation record interface, and the time series analysis method is used to extract the user knowledge point learning and behavior sequence based on the timestamp data, and integrated to form the basic sequence of the learning path; 2) In the learning path characterization stage, firstly, the node transfer matrix is ​​constructed based on the directed edge weights of the professional knowledge graph, and then the hidden Markov model analysis method is used to generate the strongest correlation path as the knowledge sequence evaluation standard, so as to generate the learning path knowledge correlation; in addition, the behavior pattern analysis can be performed based on the Apriori algorithm and the PrefixSpan method to extract the learning behavior sequence characteristics.

[0097] The above-mentioned learning path analysis method based on knowledge graph has the following advantages:

[0098] (1) Learning path representation based on knowledge graph: This application uses professional knowledge graph to accurately extract the knowledge nodes and corresponding behavior patterns experienced by users in the learning process, forming a clear learning path representation model, making the relationship between learning behavior and knowledge points clearer.

[0099] (2) Quantitative analysis of multi-dimensional behavioral characteristics: By conducting quantitative analysis on characteristics such as the frequency, duration, and behavior conversion rate of user learning behaviors, we can comprehensively evaluate the user's learning path behaviors, deeply understand the user's learning habits and behavior patterns, and provide support for personalized teaching strategies.

[0100] (3) Knowledge relevance calculation: By constructing a transition probability matrix based on the knowledge graph and using the hidden Markov model to calculate the strongest association path between knowledge points, the difference between the user's actual learning path and the optimal path can be effectively evaluated, providing a theoretical basis for optimizing the learning path.

[0101] (4) Accurate behavior pattern extraction and quantification: Through accurate learning behavior pattern extraction and quantification processing, the present invention can effectively identify the key features of learners' different operational behaviors during the learning process, providing reliable data support for behavior pattern analysis.

[0102] (5) Combination of knowledge graph and learning behavior: This application innovatively links user learning behavior with knowledge nodes in the knowledge graph, and effectively improves the interpretability and analysis accuracy of the learning path by quantitatively evaluating the relationship between knowledge points and learning behavior in multiple dimensions.

[0103] (6) Calculation of the strongest association path based on the hidden Markov model: The hidden Markov model is used to deduce the strongest association path of knowledge nodes, ensuring the scientificity and rationality of the learning path and enhancing the rigor and accuracy of the knowledge association calculation.

[0104] (7) Flexible path optimization support: By calculating the difference between the user's actual learning path and the strongest associated path, this application can provide the education platform with optimization suggestions for the learning path, thereby effectively improving learning efficiency and the depth of knowledge acquisition.

[0105] Based on the same inventive concept, the embodiment of the present application also provides a knowledge graph-based learning path analysis device for implementing the above-mentioned knowledge graph-based learning path analysis method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more knowledge graph-based learning path analysis device embodiments provided below can refer to the limitations of the knowledge graph-based learning path analysis method above, and will not be repeated here.

[0106] In an exemplary embodiment, Figure 4 As shown, a learning path analysis device based on a knowledge graph is provided, comprising:

[0107] The information acquisition module 401 is used to obtain user operation information of the online learning platform; the online learning platform has a knowledge graph; the nodes in the knowledge graph represent knowledge points in the online learning platform; the lines between the nodes in the knowledge graph represent the relationship between the knowledge points in the online learning platform.

[0108] The structured data determination module 402 is used to determine knowledge behavior events and learning behavior events based on the user operation information and the knowledge graph, and determine structured user operation record data based on the knowledge behavior events and the learning behavior events.

[0109] The knowledge point and learning behavior sequence extraction module 403 is used to determine the knowledge point learning sequence and the learning behavior sequence by using a timestamp sequence analysis method according to the structured user operation record data.

[0110] The learning path basic sequence determination module 404 is used to determine the learning path basic sequence according to the knowledge point learning sequence and the learning behavior sequence.

[0111] The behavior pattern feature extraction module 405 is used to extract behavior pattern features according to the learning behavior events; the behavior pattern features include: behavior frequency, behavior duration and behavior conversion rate.

[0112] The knowledge association calculation module 406 is used to calculate the knowledge association of the learning path using a hidden Markov model according to the relationship between the nodes in the knowledge graph.

[0113] The learning path characterization module 407 is used to determine the learning path characterization model based on the behavior pattern characteristics, the learning path knowledge association and the learning path basic sequence; the learning path characterization model is used to characterize the correspondence between the user's learning behavior pattern and the knowledge point learning order, the association between the learning behavior and the knowledge points, and the association between the knowledge points.

[0114] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store user operation information. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a learning path analysis method based on a knowledge graph is implemented.

[0115] Those skilled in the art will understand that Figure 5The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0116] In an exemplary embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0117] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0118] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0119] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, etc., but is not limited thereto.

[0120] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0121] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A learning path analysis method based on knowledge graph, characterized in that: The learning path analysis method based on knowledge graph includes: Obtaining user operation information of an online learning platform; the online learning platform has a knowledge graph; the nodes in the knowledge graph represent knowledge points in the online learning platform; the lines between the nodes in the knowledge graph represent the relationships between the knowledge points in the online learning platform; Determine knowledge behavior events and learning behavior events according to the user operation information and the knowledge graph, and determine structured user operation record data according to the knowledge behavior events and the learning behavior events; According to the structured user operation record data, the time stamp sequence analysis method is used to determine the knowledge point learning sequence and learning behavior sequence; Determine a basic sequence of learning paths according to the knowledge point learning sequence and the learning behavior sequence; Extracting behavior pattern features according to the learning behavior events; the behavior pattern features include: behavior frequency, behavior duration and behavior conversion rate; Calculate the knowledge relevance of the learning path using a hidden Markov model based on the relationship between nodes in the knowledge graph; A learning path representation model is determined based on the behavior pattern characteristics, the learning path knowledge association and the learning path basic sequence; the learning path representation model is used to characterize the correspondence between the user's learning behavior pattern and the knowledge point learning order, the association between the learning behavior and the knowledge points, and the association between the knowledge points.

2. The learning path analysis method based on knowledge graph according to claim 1 is characterized in that: Determine knowledge behavior events and learning behavior events according to the user operation information and the knowledge graph, and determine structured user operation record data according to the knowledge behavior events and the learning behavior events, specifically including: Determining an operation event in a user's learning process according to the user operation information; Determine a knowledge behavior event and a learning behavior event according to whether the operation event acts on a node in the knowledge graph; Collecting the basic attributes of the knowledge behavior event to obtain knowledge behavior operation data, and collecting the basic attributes of the learning behavior event to obtain learning behavior operation data; The knowledge behavior operation data and the learning behavior operation data are respectively subjected to data structuring processing to obtain structured knowledge behavior operation data and structured learning behavior operation data; the structured user operation record data includes: structured knowledge behavior operation data and structured learning behavior operation data.

3. The learning path analysis method based on knowledge graph according to claim 1 is characterized in that: Based on the structured user operation record data, the timestamp sequence analysis method is used to determine the knowledge point learning sequence and learning behavior sequence, including: Extracting a knowledge point set and a learning behavior set according to structured user operation record data; the knowledge point set includes: the knowledge point information accessed by the user in the knowledge graph; the learning behavior set includes: the operation behavior information performed by the user during the learning process; The knowledge point set is sorted according to the access time points using a timestamp sequence analysis method to obtain a knowledge point learning sequence; The learning behavior set is sorted according to the access time points using a timestamp sequence analysis method to obtain a learning behavior sequence.

4. The learning path analysis method based on knowledge graph according to claim 1 is characterized in that: Determining a basic sequence of a learning path according to the knowledge point learning sequence and the learning behavior sequence specifically includes: Integrating the knowledge point learning sequence and the learning behavior sequence in the form of a matrix to determine an integration matrix; the rows of the integration matrix represent the temporal relationship of each event, and the columns of the integration matrix represent the corresponding relationship between the learning behavior and the knowledge point; The knowledge points that do not correspond to the learning behaviors in the integration matrix are supplemented to obtain a basic sequence of learning paths; the supplemented knowledge points in the basic sequence of learning paths correspond one to one with the learning behaviors.

5. The learning path analysis method based on knowledge graph according to claim 1 is characterized in that: According to the relationship between nodes in the knowledge graph, a hidden Markov model is used to calculate the knowledge association of the learning path, specifically including: Constructing a transition probability matrix between knowledge points according to the relationship between nodes in the knowledge graph; Based on the transition probability matrix, a hidden Markov model is used to calculate the strongest correlation transition path between knowledge points; The learning path knowledge association degree is calculated based on the strongest association transfer path.

6. The learning path analysis method based on knowledge graph according to claim 1 is characterized in that: Determining a learning path representation model according to the behavior pattern characteristics, the learning path knowledge association and the learning path basic sequence specifically includes: Integrate the behavior pattern features and the learning path knowledge association to obtain a feature vector; The feature vector is integrated with the learning path basic sequence to obtain a learning path representation model.

7. The method for analyzing learning paths based on knowledge graph according to claim 6, characterized in that: The behavior pattern features and the learning path knowledge association are integrated to obtain a feature vector, which specifically includes: Normalizing the behavior pattern characteristics and the learning path knowledge association degree respectively to obtain normalized behavior pattern characteristics and normalized learning path knowledge association degree; The normalized behavior pattern features and the normalized learning path knowledge association are integrated to obtain a feature vector.

8. A learning path analysis device based on knowledge graph, characterized in that: The learning path analysis device based on knowledge graph includes: An information acquisition module, used to acquire user operation information of an online learning platform; the online learning platform has a knowledge graph; the nodes in the knowledge graph represent knowledge points in the online learning platform; the lines between the nodes in the knowledge graph represent the relationship between the knowledge points in the online learning platform; A structured data determination module, used to determine knowledge behavior events and learning behavior events according to the user operation information and the knowledge graph, and determine structured user operation record data according to the knowledge behavior events and the learning behavior events; A knowledge point and learning behavior sequence extraction module is used to determine the knowledge point learning sequence and learning behavior sequence using a timestamp sequence analysis method based on structured user operation record data; A learning path basic sequence determination module, used to determine a learning path basic sequence according to the knowledge point learning sequence and the learning behavior sequence; A behavior pattern feature extraction module, used to extract behavior pattern features according to the learning behavior events; the behavior pattern features include: behavior frequency, behavior duration and behavior conversion rate; A knowledge relevance calculation module, used to calculate the knowledge relevance of the learning path using a hidden Markov model according to the relationship between the nodes in the knowledge graph; A learning path characterization module is used to determine a learning path characterization model based on the behavior pattern characteristics, the learning path knowledge association and the learning path basic sequence; the learning path characterization model is used to characterize the correspondence between the user's learning behavior pattern and the knowledge point learning sequence, the association between the learning behavior and the knowledge points, and the association between the knowledge points.

9. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the knowledge graph-based learning path analysis method described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the learning path analysis method based on knowledge graph described in any one of claims 1 to 7.

Images