Intelligent course recommendation method and related device based on knowledge mastery data

Abstract

The present invention relates to the field of data processing technology, and discloses an intelligent course recommendation method based on knowledge mastery data and related devices, including: obtaining employee basic data and employee knowledge point scoring matrix; locating similar groups according to the employee knowledge point scoring matrix to obtain similar knowledge point data; integrating employee basic data, employee knowledge point scoring matrix and similar knowledge point data to obtain an intelligent course recommendation model training set; training a neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain an intelligent course recommendation model; and performing intelligent course recommendation on target employees through the intelligent course recommendation model to obtain course recommendation results. By implementing the method provided by the present application, the accuracy of the intelligent course recommendation process can be effectively improved, which is conducive to achieving the purpose of more accurate and efficient intelligent course recommendation.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an intelligent course recommendation method based on knowledge mastery data and related devices. Background Art

[0002] At present, online training has been widely used in various fields. When conducting online training, users can choose videos to be learned based on the video catalog provided by the online training; or the system can recommend videos to be learned based on the user tag set in the user portrait. The existing intelligent course recommendation method cannot accurately and quickly screen out training courses for the training population, nor can it make personalized course recommendations based on the user's knowledge mastery, learning ability, and job differences, resulting in low accuracy in the intelligent course recommendation process. Summary of the invention

[0003] The present invention provides an intelligent course recommendation method based on knowledge mastery data and a related device to solve the technical problem of low accuracy in the intelligent course recommendation process.

[0004] In the first aspect, an intelligent course recommendation method based on knowledge mastery data is provided, comprising:

[0005] Obtain employee basic data and employee knowledge point scoring matrix;

[0006] Locate similar groups of people based on the employee knowledge point scoring matrix and obtain similar knowledge point data;

[0007] Integrate employee basic data, employee knowledge point scoring matrix and similar knowledge point data to obtain the intelligent course recommendation model training set;

[0008] The neural network model based on auxiliary information is trained through the intelligent course recommendation model training set to obtain the intelligent course recommendation model;

[0009] Through the intelligent course recommendation model, intelligent courses are recommended to target employees to obtain course recommendation results.

[0010] In a second aspect, an intelligent course recommendation device based on knowledge mastering data is provided, comprising:

[0011] The acquisition module is used to obtain basic employee data and employee knowledge point scoring matrix;

[0012] The first processing module is used to locate similar groups of people based on the employee knowledge point scoring matrix and obtain similar knowledge point data;

[0013] The second processing module is used to integrate the basic data of employees, the employee knowledge point scoring matrix and the similar knowledge point data to obtain the intelligent course recommendation model training set;

[0014] The third processing module is used to train the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain the intelligent course recommendation model;

[0015] The fourth processing module is used to make intelligent course recommendations to target employees through an intelligent course recommendation model to obtain course recommendation results.

[0016] In a third aspect, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the above-mentioned intelligent course recommendation method based on knowledge mastery data are implemented.

[0017] In a fourth aspect, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the steps of the above-mentioned intelligent course recommendation method based on knowledge mastery data are implemented.

[0018] In the scheme implemented by the above-mentioned intelligent course recommendation method based on knowledge mastery data and related devices, by obtaining employee basic data and employee knowledge point scoring matrix, and locating similar people according to the employee knowledge point scoring matrix, similar knowledge point data is obtained, and the employee basic data, employee knowledge point scoring matrix and similar knowledge point data are further integrated to obtain an intelligent course recommendation model training set, so that the neural network model based on auxiliary information can be trained through the intelligent course recommendation model training set to obtain an intelligent course recommendation model, and then the intelligent course recommendation model can be used to make intelligent course recommendations for target employees, so as to obtain more accurate course recommendation results, which can improve the accuracy of the intelligent course recommendation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0020] Figure 1 Schematic diagram of an application environment of an intelligent course recommendation method based on knowledge mastering data in one embodiment of the present invention;

[0021] Figure 2is a flow chart of an intelligent course recommendation method based on knowledge mastering data in one embodiment of the present invention;

[0022] Figure 3 is a structural diagram of an intelligent course recommendation model in one embodiment of the present invention;

[0023] Figure 4 is a structural diagram of an intelligent course recommendation device based on knowledge mastering data in one embodiment of the present invention;

[0024] Figure 5 is a schematic diagram of a structure of a computer device in one embodiment of the present invention;

[0025] Figure 6 It is another structural schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] The intelligent course recommendation method based on knowledge mastery data provided by the embodiment of the present invention can be applied in the following aspects: Figure 1 In an application environment, the client communicates with the server through a network. Exemplarily, taking the scenario of intelligent course recommendation for target users as an example, R&D personnel or managers can upload employee basic data and employee knowledge point scoring matrix through the client. Correspondingly, the server can obtain employee basic data and employee knowledge point scoring matrix through the client, and locate similar people according to the employee knowledge point scoring matrix to obtain similar knowledge point data, so as to further integrate employee basic data, employee knowledge point scoring matrix and similar knowledge point data to obtain an intelligent course recommendation model training set, so that the neural network model based on auxiliary information can be trained through the intelligent course recommendation model training set to obtain an intelligent course recommendation model, and then the intelligent course recommendation model can be used to perform intelligent course recommendations for target employees, obtain more accurate course recommendation results, and feed back the course recommendation results to the client. Correspondingly, the client can receive the course recommendation results from the server, and can display the course recommendation results on the client for the target user to query or browse. By adopting the intelligent course recommendation method based on knowledge mastering data provided by the present application, the accuracy of the intelligent course recommendation process can be effectively improved.

[0028] The client may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and portable wearable devices. The server may be implemented by an independent server or a server cluster consisting of multiple servers. The present invention is described in detail below through specific embodiments.

[0029] See also Figure 2 As shown, Figure 2 A flow chart of an intelligent course recommendation method based on knowledge mastering data provided by an embodiment of the present invention includes the following steps:

[0030] S10: Obtain employee basic data and employee knowledge point scoring matrix.

[0031] Employee basic data may include but is not limited to employee age, gender, years of service, job level and other related data. The server can obtain employee basic data from the enterprise's human resources management system. Employee basic data can reflect the basic characteristics of employees, which may affect their learning needs and course preferences.

[0032] The employee knowledge point scoring matrix can record the employee's score on the degree of mastery of each knowledge point. The employee knowledge point scoring matrix can be constructed based on the employee's historical assessment results, training evaluation data, job resource knowledge graph and other data, and this application does not limit this.

[0033] S20: locate similar groups of people based on the employee knowledge point scoring matrix to obtain similar knowledge point data.

[0034] By analyzing the employee knowledge point scoring matrix, similar groups of people can be further found. Optionally, the server can use clustering algorithms, similarity calculation methods, etc. to locate similar groups of people. For example, the server can use cosine similarity to calculate the similarity between two employees in terms of their knowledge point mastery, and group employees with higher similarity into similar groups.

[0035] It is understandable that similar people may have similarities in their knowledge mastery, so similar knowledge points can be determined, and the similar knowledge point data can reflect the commonalities in the knowledge structure of this similar group.

[0036] It should be understood that locating similar people based on the employee knowledge point scoring matrix and obtaining similar knowledge point data refers to the process of determining similar knowledge point data. Among them, step S20, that is, locating similar people based on the employee knowledge point scoring matrix and obtaining similar knowledge point data, includes the following steps:

[0037] S21: Based on the employee knowledge point scoring matrix, the employee similarity is calculated using modified cosine similarity to obtain an employee similarity set;

[0038] S22: construct a similarity matrix based on the employee similarity set;

[0039] S23: Determine similar employee groups based on the similarity matrix;

[0040] S24: Determine similar knowledge point data corresponding to the similar employee groups based on the knowledge point scoring data of the similar employee groups.

[0041] In the course recommendation scenario, the employee knowledge point rating matrix can record each employee's rating of different knowledge points. By calculating the modified cosine similarity, the similarity of different employees in terms of knowledge mastery can be compared. Among them, the modified cosine similarity is an improved cosine similarity calculation method that takes into account the impact of the average value of users' ratings on different items and can more accurately measure the similarity between users.

[0042] Optionally, assume that the employee knowledge point scoring matrix is ​​R, R u,i where represents the score of employee u on knowledge point i. The server can use the following formula to calculate the similarity of employees based on the employee knowledge point score matrix and use the modified cosine similarity to obtain the steps of the employee similarity set:

[0043]

[0044] Among them, sim(u, v) represents the similarity between employee u and employee v; I u，v represents the set of knowledge points that employees u and v have learned and rated together; i represents the index of a knowledge point in the knowledge point set; R u，i represents the score of knowledge point i of employee u; represents the average rating of employee u; R v,i represents the score of employee v’s knowledge point i; represents the average rating of employee v.

[0045] The similarity matrix can be a two-dimensional matrix, and the elements in the similarity matrix can represent the similarities between different employees. By constructing the similarity matrix, the similarity relationship between employees can be intuitively displayed, which is convenient for subsequent analysis and processing. For example, assuming that there are n employees in total, the similarity matrix S can be an n×n matrix, where the elements S uv It can represent the similarity between employee u and employee v, and this application does not impose any limitation on this.

[0046] Furthermore, a similarity threshold may be set, and if the similarity between two employees is greater than or equal to the threshold, it can be determined that the two employees belong to a similar employee group. For example, assuming that the similarity threshold is set to 0.8, if the similarity between employee u and employee v is greater than 0.8, employee u and employee v can be determined as a similar employee group. Optionally, a clustering algorithm, such as hierarchical clustering, K-means clustering algorithm (K-Means), etc., may also be used to cluster employees according to the similarity matrix to obtain different similar employee groups, and this application does not limit this.

[0047] For each similar employee group, the server can analyze their ratings of each knowledge point to screen the knowledge points and select the knowledge points with high ratings or common interests among similar employee groups as similar knowledge point data. For example, for a similar employee group, if multiple employees have high ratings for a certain knowledge point, then the knowledge point can be determined as a similar knowledge point for the group. Through the above steps, similar employee groups can be effectively determined based on the employee knowledge point rating matrix, and similar knowledge point data that the similar employee groups need or are interested in can be found, which is helpful to provide an important basis for subsequent course recommendations.

[0048] In a possible implementation, the server can determine the target employee's mastered knowledge (such as determining the mastered knowledge point data) and unmastered knowledge (such as determining the unmastered knowledge point data) based on similar employee groups, so as to better understand the difference between the target employee's current learning situation and the similar employee group, so as to better recommend courses for the target employee. In other words, the similar knowledge point data may include mastered knowledge point data and unmastered knowledge point data, and this application does not limit this.

[0049] Exemplarily, the server can further calculate and estimate the learning ability of employees by finding similar employees, such as predicting the knowledge points that employees have not mastered through the latent semantic model LFM. The similarity between employees can be calculated based on the employee knowledge point scoring matrix to achieve the search for similar employees, and the modified cosine similarity can be used to calculate the similarity of employees so as to recommend employees with higher similarity in knowledge mastery to each other. The employee knowledge point scoring matrix can be used to comprehensively consider the employee knowledge point scoring matrix using LFM, and the target employee's mastery of the unlearned knowledge points can be calculated by combining the method of predictive scoring, so as to select the N knowledge points with the worst mastery as the set of unmastered knowledge points of the target employee, and then select the weighted average calculation method based on the employee score mean to predict the mastery of the unmastered knowledge points of the employee. The predicted score of the unlearned knowledge points can be obtained by calculation, and the predicted scores can be arranged to obtain the N unmastered knowledge points of the employee. The knowledge points with a predicted score of 0 can be defined as courses that have not been learned, and the knowledge points with lower predicted scores can be classified as unmastered knowledge points. Furthermore, the N unmastered knowledge points can be combined with the job resource knowledge graph to obtain the unmastered knowledge points of the job, and this application does not impose any restrictions on this.

[0050] For steps S21-S24, based on the employee knowledge point scoring matrix, modified cosine similarity is used to calculate the employee similarity to obtain an employee similarity set, and based on the employee similarity set, a similarity matrix is ​​constructed, so that similar employee groups are determined based on the similarity matrix, and then based on the knowledge point scoring data of the similar employee groups, similar knowledge point data corresponding to the similar employee groups are determined, which is conducive to obtaining a more accurate intelligent course recommendation model based on the similar knowledge point data in subsequent steps.

[0051] S30: Integrate employee basic data, employee knowledge point scoring matrix and similar knowledge point data to obtain an intelligent course recommendation model training set.

[0052] Further integration of employee basic data, employee knowledge point scoring matrix and similar knowledge point data can construct a training set for the intelligent course recommendation model. Employee basic data provides personal characteristic information of employees, knowledge point scoring matrix reflects employees' knowledge mastery, and similar knowledge point data reflects the commonality of knowledge among groups. After being sorted and integrated, the above data can provide rich training materials for the course recommendation model.

[0053] In one possible implementation, the server can integrate employee basic data, employee knowledge point scoring matrix and similar knowledge point data to construct an employee knowledge mastery map, thereby training and optimizing the intelligent course recommendation model based on the employee knowledge mastery map, which is not limited in this application. Optionally, the construction process of the employee knowledge mastery map can be:

[0054] 1. Obtain training courseware resources, cluster the content of the training courseware, perform Chinese word segmentation on the training courseware, remove used words to create a stop file, build a bag-of-words space, count the word sets of all documents, and build a vector for each document. The value of the vector is the number of times the word appears in the document. Convert the number of occurrences of each word into a weight, generate a feature model, and cluster the training courseware using the K-means algorithm.

[0055] 2. Match the training courseware after clustering with the job classification to generate a job resource map, match the feature model after clustering with each job one by one, and describe the job with the corresponding model features to generate a job resource map.

[0056] 3. Combine the job resource map with the employee knowledge mastery to generate the employee knowledge mastery map. Extract the job resource map data after semi-structured clustering and the structured employee knowledge mastery information. Perform knowledge representation on the extracted entity, relationship and attribute triples. The knowledge representation uses the entity link model to express the knowledge elements more accurately, and disambiguates the heterogeneous data through knowledge fusion. The entities, relationships and attributes of knowledge are fused to eliminate the ambiguity of knowledge information. Finally, the knowledge model is constructed, and the employee knowledge mastery map is visualized through the graph database management system (Neo4jGraph Database Management System, Nco4j).

[0057] S40: Train the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain the intelligent course recommendation model.

[0058] This application uses a neural network model based on auxiliary information as the initial model. The neural network model based on auxiliary information can be composed of a generalized matrix decomposition model and a multi-layer perceptron model, so as to construct a model by analyzing the employee's knowledge-related data. The employee and the courseware are the smallest unit classification functions.

[0059] The neural network model based on auxiliary information can make full use of auxiliary information such as basic employee data, and learn in combination with data such as knowledge point scoring matrix. During the training process, the initial model can fit the data in the training set of the intelligent course recommendation model by continuously adjusting parameters, so that the model can learn the relationship between employee characteristics, knowledge mastery and course recommendations, and then obtain an intelligent course recommendation model.

[0060] It should be understood that training the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain the intelligent course recommendation model refers to the process of training the initial model. Among them, step S40, that is, training the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain the intelligent course recommendation model, includes the following steps:

[0061] S41: extracting basic information of training employees, training position course information, and auxiliary information from the intelligent course recommendation model training set through the input layer of the neural network model based on auxiliary information;

[0062] S42: flattening the basic information of the training employees and the training position course information through the embedding layer of the neural network model based on the auxiliary information to obtain a one-dimensional feature vector;

[0063] S43: adding the auxiliary information to the one-dimensional feature vector for expansion through an embedding layer of a neural network model based on the auxiliary information to obtain an expanded feature vector;

[0064] S44: performing feature decomposition processing on the one-dimensional feature vector and the extended feature vector through a generalized matrix decomposition sub-model in an output layer of the neural network model based on auxiliary information to obtain a potential feature vector;

[0065] S45: performing feature enrichment processing on the one-dimensional feature vector and the extended feature vector through a multi-layer perceptron sub-model in the output layer of the neural network model based on auxiliary information to obtain a multi-dimensional feature vector;

[0066] S46: Through the output layer of the neural network model based on auxiliary information, the potential feature vector and the multi-dimensional feature vector are subjected to information fusion processing to obtain the reference course recommendation result;

[0067] S47: According to the reference course recommendation results, the neural network model based on the auxiliary information is iteratively trained multiple times to obtain an intelligent course recommendation model.

[0068] like Figure 3 As shown, Figure 3The schematic diagram of an intelligent course recommendation model is shown as an example. Data extraction operations can be performed through the input layer of the neural network model based on auxiliary information. Among them, the basic information of the training employees may include but is not limited to the employee's age, work experience, education, etc.; the training position course information may include but is not limited to the position-related course content, course difficulty, etc., and is encoded using one-hot encoding.

[0069] Among them, auxiliary information can be supplementary data that helps model learning but does not belong to the first two categories of information, such as the company's business direction, industry trends, and employees' interests, hobbies, areas of expertise, etc. This application does not impose any restrictions on this.

[0070] Through the embedding layer of the neural network model based on auxiliary information, flattening and feature expansion operations can be performed. Flattening can convert the information extracted in the above process into a format that is convenient for neural network processing. For example, multiple attributes in employee basic information and multiple aspects of job course information will be converted into a linear vector representation, so that the model can process these data uniformly.

[0071] Furthermore, the one-dimensional feature vector is expanded by adding auxiliary information to obtain an extended feature vector, which can consider all relevant information more comprehensively so that the auxiliary information can be combined with the main features related to employees and job courses, thereby enriching the representation of features and enabling the model to learn with more comprehensive data.

[0072] In the output layer, the generalized matrix decomposition sub-model can perform point multiplication on the flattened results as the linear output results; the multi-layer perceptron sub-model can connect the flattened results end to end to form a new vector as the input of the neural network, and then the outputs of the generalized matrix decomposition and multi-layer perceptron models can be further connected end to end and mapped to [0, 1.0] through an activation function (such as the Sigmoid activation function) as the course recommendation result.

[0073] Among them, the generalized matrix decomposition sub-model can decompose a high-dimensional matrix into the product of two or more low-dimensional matrices based on the basic idea of ​​matrix decomposition. Through generalized matrix decomposition, a high-dimensional matrix can be decomposed into the product of an employee feature matrix and a course feature matrix, and each row vector in the employee feature matrix and the course feature matrix can represent the representation of employees and courses in the latent feature space respectively. It can be understood that the latent feature is usually a feature hidden in the original data that can reflect the essential attributes of employees and courses. For example, latent features may include employees' learning style, interest preferences, and the difficulty level and knowledge field of the course.

[0074] In other words, generalized matrix decomposition can be used to discover hidden and potential feature relationships in the data. For example, it can be used to discover the matching relationship between certain potential learning ability characteristics of employees and specific job courses. This potential feature vector helps the model understand the data more deeply.

[0075] It should be noted that traditional matrix decomposition methods (such as singular value decomposition, SVD) usually only consider numerical data, while generalized matrix decomposition can handle a wider range of data types, including discrete and mixed data. Generalized matrix decomposition can also introduce additional information, such as the above-mentioned auxiliary information (such as basic information of employees, attribute information of courses, etc.), to improve the accuracy of course recommendations.

[0076] Multilayer perceptrons can perform nonlinear transformations on input features through calculations in multiple neuron layers, thereby increasing the richness and expressiveness of features. For example, a multilayer perceptron submodel can generate some new and more discriminative features from the original employee and job course features, thereby forming a multidimensional feature vector.

[0077] By synthesizing the feature vectors obtained by different processing methods, a reference course recommendation result can be generated. For example, potential features and multi-dimensional features are combined by weighted summation, splicing, etc. to determine reference course recommendations suitable for employees. The reference course recommendation result can be the course recommendation result obtained during the model training process, and this application does not limit this.

[0078] According to the reference course recommendation results, the neural network model based on auxiliary information can be trained multiple times to obtain an intelligent course recommendation model. In each iteration, the model can adjust the model parameters, such as the weight of the neural network, based on the difference between the current recommendation results and the actual situation (which may be measured by the real course recommendation data in the training set), so that the model can be continuously optimized through multiple iterations to make course recommendations more accurately.

[0079] Optionally, the neural network model based on auxiliary information is trained through the intelligent course recommendation model training set to obtain the intelligent course recommendation model expressed as follows:

[0080]

[0081]

[0082] in, represents the predicted value, i.e., the course recommendation result of the intelligent course recommendation model; f out Represents the activation function weight of the output layer, such as the first activation function weight; W out Represents the activation function weight of the output layer, such as the second activation function weight; XGMF represents the potential feature vector of the GMF model; X MLP represents the latent feature vector of the MLP model; Represents X GMF The user latent feature vector in ; Represents X GMF The latent feature vector of the items in ; Indicates employee information, such as the first employee information; Indicates employee information, such as the second employee information; Represents employee information, such as the m+1th employee information; Represents employee information, such as the m+1th employee information; Indicates auxiliary information, such as being called the first auxiliary information; Indicates auxiliary information, such as being called the second auxiliary information; represents auxiliary information, such as the m+1th auxiliary information; f L represents the activation function weight of the Lth layer, such as the activation function weight of the Lth layer is 1; W out represents the activation function weight of the Lth layer, such as the activation function weight of the Lth layer is 2; f 1 Represents the activation function weight of the first layer, such as the first layer activation function weight 1; W 1 Represents the activation function weight of the first layer, such as the activation function weight of the second layer is 2; Represents X MLP The latent feature vector of the user in ; Represents X MLP The latent feature vector of the item in b 1 represents the bias of the first layer; b L represents the bias of the Lth layer.

[0083] For steps S41-S47, the basic information of training employees, training position course information, and auxiliary information are extracted from the intelligent course recommendation model training set through the input layer of the neural network model based on auxiliary information, and the basic information of training employees and the training position course information are flattened through the embedding layer of the neural network model based on auxiliary information to obtain a one-dimensional feature vector, and the one-dimensional feature vector is added with auxiliary information through the embedding layer of the neural network model based on auxiliary information to expand it to obtain an extended feature vector, and the one-dimensional feature vector and the extended feature vector are feature decomposed through the generalized matrix decomposition sub-model in the output layer of the neural network model based on auxiliary information to obtain a potential feature vector, so that the one-dimensional feature vector and the extended feature vector can be feature enriched through the multi-layer perceptron sub-model in the output layer of the neural network model based on auxiliary information to obtain a multi-dimensional feature vector, and the potential feature vector and the multi-dimensional feature vector can be further information-fused through the output layer of the neural network model based on auxiliary information to obtain a reference course recommendation result, and then the neural network model based on auxiliary information can be iteratively trained multiple times according to the reference course recommendation result to obtain a more accurate intelligent course recommendation model, which is conducive to obtaining a more accurate course recommendation result based on the intelligent course recommendation model in subsequent steps.

[0084] S50: Through the intelligent course recommendation model, intelligent courses are recommended to the target employees to obtain course recommendation results.

[0085] Among them, the target employees may be employees who need intelligent course recommendations. Through the trained intelligent course recommendation model, intelligent course recommendations can be made to target employees. The intelligent course recommendation model can recommend appropriate courses to target employees based on the basic data and knowledge point scores of the target employees, combined with the patterns learned from the training set, and finally obtain course recommendation results. The course recommendation results can help enterprises provide training courses for employees in a more targeted manner, thereby improving the knowledge level and work ability of employees.

[0086] It can be seen that in the above scheme, the server obtains employee basic data and employee knowledge point scoring matrix, and locates similar people according to the employee knowledge point scoring matrix to obtain similar knowledge point data, so as to further integrate employee basic data, employee knowledge point scoring matrix and similar knowledge point data to obtain the intelligent course recommendation model training set, so that the neural network model based on auxiliary information can be trained through the intelligent course recommendation model training set to obtain the intelligent course recommendation model, and then the intelligent course recommendation model can be used to make intelligent course recommendations for target employees, so as to obtain more accurate course recommendation results, which can improve the accuracy of the intelligent course recommendation process.

[0087] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0088] In one embodiment, an intelligent course recommendation device based on knowledge mastering data is provided, and the intelligent course recommendation device based on knowledge mastering data corresponds to the intelligent course recommendation method based on knowledge mastering data in the above embodiment. Figure 4 As shown, the intelligent course recommendation device based on knowledge mastering data includes an acquisition module 101, a first processing module 102, a second processing module 103, a third processing module 104 and a fourth processing module 105. The detailed description of each functional module is as follows:

[0089] Acquisition module 101, used for acquiring the basic data of employees and the employee knowledge point scoring matrix;

[0090] The first processing module 102 is used to locate similar groups of people according to the employee knowledge point scoring matrix to obtain similar knowledge point data;

[0091] The second processing module 103 is used to integrate the employee basic data, the employee knowledge point scoring matrix and the similar knowledge point data to obtain the intelligent course recommendation model training set;

[0092] The third processing module 104 is used to train the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain the intelligent course recommendation model;

[0093] The fourth processing module 105 is used to perform intelligent course recommendation for the target employee through the intelligent course recommendation model to obtain a course recommendation result.

[0094] In one embodiment, the third processing module 104 is used to train the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain the intelligent course recommendation model, specifically for:

[0095] Through the input layer of the neural network model based on auxiliary information, basic information of training employees, training position course information, and auxiliary information are extracted from the intelligent course recommendation model training set;

[0096] Through the embedding layer of the neural network model based on auxiliary information, the basic information of the training employees and the training position course information are flattened to obtain a one-dimensional feature vector;

[0097] Through the embedding layer of the neural network model based on auxiliary information, the one-dimensional feature vector is expanded by adding auxiliary information to obtain an extended feature vector;

[0098] Through the generalized matrix decomposition sub-model in the output layer of the neural network model based on auxiliary information, the one-dimensional feature vector and the extended feature vector are subjected to feature decomposition processing to obtain a potential feature vector;

[0099] Through the multi-layer perceptron sub-model in the output layer of the neural network model based on auxiliary information, the one-dimensional feature vector and the extended feature vector are processed with feature enrichment to obtain a multi-dimensional feature vector;

[0100] Through the output layer of the neural network model based on auxiliary information, the potential feature vector and the multi-dimensional feature vector are fused to obtain the reference course recommendation result;

[0101] According to the reference course recommendation results, the neural network model based on auxiliary information is trained iteratively multiple times to obtain an intelligent course recommendation model.

[0102] In one embodiment, the third processing module 104 is used to train the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain the intelligent course recommendation model, specifically for:

[0103] The neural network model based on auxiliary information is trained through the intelligent course recommendation model training set to obtain the intelligent course recommendation model expressed as follows:

[0104]

[0105]

[0106] in, represents the predicted value, i.e., the course recommendation result of the intelligent course recommendation model; f out Represents the activation function weight of the output layer, such as the first activation function weight; W out Represents the activation function weight of the output layer, such as the second activation function weight; X GMF represents the potential feature vector of the GMF model; X MLP represents the latent feature vector of the MLP model; Represents X GMF The user latent feature vector in ; Represents X GMF The latent feature vector of the items in ; Indicates employee information, such as the first employee information; Indicates employee information, such as the second employee information; Represents employee information, such as the m+1th employee information; Represents employee information, such as the m+1th employee information; Indicates auxiliary information, such as being called the first auxiliary information; Indicates auxiliary information, such as being called the second auxiliary information; represents auxiliary information, such as the m+1th auxiliary information; f L represents the activation function weight of the Lth layer, such as the activation function weight of the Lth layer is 1; W out represents the activation function weight of the Lth layer, such as the activation function weight of the Lth layer is 2; f 1 Represents the activation function weight of the first layer, such as the first layer activation function weight 1; W 1 Represents the activation function weight of the first layer, such as the activation function weight of the second layer is 2; Represents X MLP The latent feature vector of the user in ; Represents X MLP The latent feature vector of the item in b 1 represents the bias of the first layer; b L represents the bias of the Lth layer.

[0107] In one embodiment, the first processing module 102 is used to locate similar groups of people according to the employee knowledge point scoring matrix to obtain similar knowledge point data, specifically for:

[0108] According to the employee knowledge point scoring matrix, the employee similarity is calculated using modified cosine similarity to obtain the employee similarity set;

[0109] Construct a similarity matrix based on the employee similarity set;

[0110] Based on the similarity matrix, identify similar employee groups;

[0111] Based on the knowledge point scoring data of similar employee groups, similar knowledge point data corresponding to the similar employee groups are determined.

[0112] In one embodiment, the first processing module 102 is used to calculate the similarity of employees using modified cosine similarity according to the employee knowledge point scoring matrix to obtain an employee similarity set, which is specifically used to:

[0113] The following formula is used to calculate the similarity of employees based on the employee knowledge point scoring matrix and use the modified cosine similarity to obtain the steps of the employee similarity set:

[0114]

[0115] Among them, sim(u, v) represents the similarity between employee u and employee v; I u，v represents the set of knowledge points that employees u and v have learned together; i represents the index of a knowledge point in the set of knowledge points; R u，irepresents the score of knowledge point i of employee u; represents the average rating of employee u; R v,i represents the score of employee v’s knowledge point i; represents the average rating of employee v.

[0116] The present invention provides an intelligent course recommendation device based on knowledge mastery data, which obtains employee basic data and employee knowledge point scoring matrix, locates similar people according to the employee knowledge point scoring matrix, obtains similar knowledge point data, and further integrates the employee basic data, employee knowledge point scoring matrix and similar knowledge point data to obtain an intelligent course recommendation model training set, so that a neural network model based on auxiliary information can be trained through the intelligent course recommendation model training set to obtain an intelligent course recommendation model, and then the intelligent course recommendation model can be used to perform intelligent course recommendations on target employees, so as to obtain more accurate course recommendation results, and the accuracy of the intelligent course recommendation process can be improved.

[0117] For the specific limitations of the intelligent course recommendation device based on knowledge mastery data, please refer to the limitations of the intelligent course recommendation method based on knowledge mastery data in the above text, which will not be repeated here. Each module in the above-mentioned intelligent course recommendation device based on knowledge mastery data can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0118] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. 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 and / or 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 network interface of the computer device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the server side of an intelligent course recommendation method based on knowledge mastery data.

[0119] In one embodiment, a computer device is provided. The computer device may be a client, and its internal structure diagram may be as follows: Figure 6As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. 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 and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the client side of an intelligent course recommendation method based on knowledge mastering data.

[0120] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0121] Obtain employee basic data and employee knowledge point scoring matrix;

[0122] Locate similar groups of people based on the employee knowledge point scoring matrix and obtain similar knowledge point data;

[0123] Integrate employee basic data, employee knowledge point scoring matrix and similar knowledge point data to obtain the intelligent course recommendation model training set;

[0124] The neural network model based on auxiliary information is trained through the intelligent course recommendation model training set to obtain the intelligent course recommendation model;

[0125] Through the intelligent course recommendation model, intelligent courses are recommended to target employees to obtain course recommendation results.

[0126] The present invention provides a computer device, which obtains employee basic data and an employee knowledge point scoring matrix, locates similar people according to the employee knowledge point scoring matrix, obtains similar knowledge point data, and further integrates the employee basic data, the employee knowledge point scoring matrix and the similar knowledge point data to obtain an intelligent course recommendation model training set, so that a neural network model based on auxiliary information can be trained through the intelligent course recommendation model training set to obtain an intelligent course recommendation model, and then the intelligent course recommendation model can be used to perform intelligent course recommendations on target employees, thereby obtaining more accurate course recommendation results, and improving the accuracy of the intelligent course recommendation process.

[0127] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0128] Obtain employee basic data and employee knowledge point scoring matrix;

[0129] Locate similar groups of people based on the employee knowledge point scoring matrix and obtain similar knowledge point data;

[0130] Integrate employee basic data, employee knowledge point scoring matrix and similar knowledge point data to obtain the intelligent course recommendation model training set;

[0131] The neural network model based on auxiliary information is trained through the intelligent course recommendation model training set to obtain the intelligent course recommendation model;

[0132] Through the intelligent course recommendation model, intelligent courses are recommended to target employees to obtain course recommendation results.

[0133] The present invention provides a computer-readable storage medium, which obtains employee basic data and an employee knowledge point scoring matrix, locates similar people according to the employee knowledge point scoring matrix, obtains similar knowledge point data, and further integrates the employee basic data, the employee knowledge point scoring matrix and the similar knowledge point data to obtain an intelligent course recommendation model training set, so that a neural network model based on auxiliary information can be trained through the intelligent course recommendation model training set to obtain an intelligent course recommendation model, and then the intelligent course recommendation model can be used to perform intelligent course recommendations on target employees, thereby obtaining more accurate course recommendation results, and improving the accuracy of the intelligent course recommendation process.

[0134] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0135] Those skilled 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 memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0136] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0137] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. An intelligent course recommendation method based on knowledge mastery data, characterized in that: The method comprises: Obtain employee basic data and employee knowledge point scoring matrix; Locate similar groups of people based on the employee knowledge point scoring matrix to obtain similar knowledge point data; Integrate the employee basic data, the employee knowledge point scoring matrix and the similar knowledge point data to obtain an intelligent course recommendation model training set; The neural network model based on auxiliary information is trained by using the intelligent course recommendation model training set to obtain an intelligent course recommendation model; Through the intelligent course recommendation model, intelligent courses are recommended to target employees to obtain course recommendation results.

2. The intelligent course recommendation method based on knowledge mastering data according to claim 1 is characterized in that: The method of training the neural network model based on auxiliary information by using the intelligent course recommendation model training set to obtain the intelligent course recommendation model includes: Extracting basic information of training employees, training position course information, and auxiliary information from the training set of the intelligent course recommendation model through the input layer of the neural network model based on auxiliary information; The basic information of the training employees and the training position course information are flattened through the embedding layer of the neural network model based on auxiliary information to obtain a one-dimensional feature vector; By using the embedding layer of the neural network model based on auxiliary information, the one-dimensional feature vector is added to the auxiliary information for expansion to obtain an expanded feature vector; Performing feature decomposition processing on the one-dimensional feature vector and the extended feature vector through a generalized matrix decomposition sub-model in the output layer of the neural network model based on auxiliary information to obtain a potential feature vector; The one-dimensional feature vector and the extended feature vector are subjected to feature enrichment processing by a multi-layer perceptron sub-model in an output layer of the neural network model based on auxiliary information to obtain a multi-dimensional feature vector; Through the output layer of the neural network model based on auxiliary information, the potential feature vector and the multi-dimensional feature vector are subjected to information fusion processing to obtain a reference course recommendation result; According to the reference course recommendation results, the neural network model based on auxiliary information is iteratively trained for multiple times to obtain an intelligent course recommendation model.

3. The intelligent course recommendation method based on knowledge mastering data according to claim 2 is characterized in that: The method of training the neural network model based on auxiliary information by using the intelligent course recommendation model training set to obtain the intelligent course recommendation model includes: The neural network model based on auxiliary information is trained by the intelligent course recommendation model training set to obtain the intelligent course recommendation model expressed as follows: in, represents the predicted value, i.e., the course recommendation result of the intelligent course recommendation model; f out Represents the activation function weight of the output layer, such as the first activation function weight; W out Represents the activation function weight of the output layer, such as the second activation function weight; X GMF represents the potential feature vector of the GMF model; X MLP represents the latent feature vector of the MLP model; Represents X GMF The user latent feature vector in ; Represents X GMF The latent feature vector of the items in ; Indicates employee information, such as the first employee information; Indicates employee information, such as the second employee information; Represents employee information, such as the m+1th employee information; Represents employee information, such as the m+1th employee information; Indicates auxiliary information, such as being called the first auxiliary information; Indicates auxiliary information, such as being called the second auxiliary information; represents auxiliary information, such as the m+1th auxiliary information; f L represents the activation function weight of the Lth layer, such as the activation function weight of the Lth layer is 1; W out represents the activation function weight of the Lth layer, such as the activation function weight 2 of the Lth layer; f1 represents the activation function weight of the first layer, such as the activation function weight 1 of the first layer; W1 represents the activation function weight of the first layer, such as the activation function weight 2 of the second layer; Represents X MLP The latent feature vector of the user in ; Represents X MLP The latent feature vector of the item in, b1 represents the bias of the first layer; b L represents the bias of the Lth layer.

4. The intelligent course recommendation method based on knowledge mastering data according to claim 3 is characterized in that: The method of locating similar groups of people based on the employee knowledge point scoring matrix to obtain similar knowledge point data includes: According to the employee knowledge point scoring matrix, the employee similarity is calculated using modified cosine similarity to obtain an employee similarity set; Constructing a similarity matrix according to the employee similarity set; Determining similar employee groups according to the similarity matrix; Similar knowledge point data corresponding to the similar employee group is determined according to the knowledge point scoring data of the similar employee group.

5. The intelligent course recommendation method based on knowledge mastering data according to claim 4 is characterized in that: The employee similarity calculation is performed based on the employee knowledge point scoring matrix using modified cosine similarity to obtain an employee similarity set, including: The step of calculating employee similarity using modified cosine similarity based on the employee knowledge point scoring matrix to obtain an employee similarity set is implemented by the following formula: Among them, sim(u,v) represents the similarity between employee u and employee v; I u,v represents the set of knowledge points that employees u and v have learned together; i represents the index of a knowledge point in the set of knowledge points; R u,i represents the score of knowledge point i of employee u; represents the average rating of employee u; R v,i represents the score of employee v’s knowledge point i; represents the average rating of employee v.

6. An intelligent course recommendation device based on knowledge mastery data, characterized in that: The intelligent course recommendation device based on knowledge mastering data comprises: The acquisition module is used to obtain basic employee data and employee knowledge point scoring matrix; A first processing module is used to locate similar groups of people according to the employee knowledge point scoring matrix to obtain similar knowledge point data; The second processing module is used to integrate the employee basic data, the employee knowledge point scoring matrix and the similar knowledge point data to obtain an intelligent course recommendation model training set; A third processing module is used to train the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain an intelligent course recommendation model; The fourth processing module is used to perform intelligent course recommendation for the target employee through the intelligent course recommendation model to obtain a course recommendation result.

7. The intelligent course recommendation device based on knowledge mastering data according to claim 6 is characterized in that: The third processing module is used to train the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain the intelligent course recommendation model, specifically for: Extracting basic information of training employees, training position course information, and auxiliary information from the training set of the intelligent course recommendation model through the input layer of the neural network model based on auxiliary information; The basic information of the training employees and the training position course information are flattened through the embedding layer of the neural network model based on auxiliary information to obtain a one-dimensional feature vector; By using the embedding layer of the neural network model based on auxiliary information, the one-dimensional feature vector is added to the auxiliary information for expansion to obtain an expanded feature vector; Performing feature decomposition processing on the one-dimensional feature vector and the extended feature vector through a generalized matrix decomposition sub-model in the output layer of the neural network model based on auxiliary information to obtain a potential feature vector; The one-dimensional feature vector and the extended feature vector are subjected to feature enrichment processing by a multi-layer perceptron sub-model in an output layer of the neural network model based on auxiliary information to obtain a multi-dimensional feature vector; Through the output layer of the neural network model based on auxiliary information, the potential feature vector and the multi-dimensional feature vector are subjected to information fusion processing to obtain a reference course recommendation result; According to the reference course recommendation results, the neural network model based on auxiliary information is iteratively trained for multiple times to obtain an intelligent course recommendation model.

8. The intelligent course recommendation device based on knowledge mastering data according to claim 7 is characterized in that: The third processing module is used to train the neural network model based on auxiliary information through the intelligent course recommendation model training set to obtain the intelligent course recommendation model, specifically for: The neural network model based on auxiliary information is trained by the intelligent course recommendation model training set to obtain the intelligent course recommendation model expressed as follows: in, represents the predicted value, i.e., the course recommendation result of the intelligent course recommendation model; f out Represents the activation function weight of the output layer, such as the first activation function weight; W out Represents the activation function weight of the output layer, such as the second activation function weight; X GMF represents the potential feature vector of the GMF model; X MLP represents the latent feature vector of the MLP model; Represents X GMF The user latent feature vector in ; Represents X GMF The latent feature vector of the items in ; Indicates employee information, such as the first employee information; Indicates employee information, such as the second employee information; Represents employee information, such as the m+1th employee information; Represents employee information, such as the m+1th employee information; Indicates auxiliary information, such as being called the first auxiliary information; Indicates auxiliary information, such as being called the second auxiliary information; represents auxiliary information, such as the m+1th auxiliary information; f L represents the activation function weight of the Lth layer, such as the activation function weight of the Lth layer is 1; W out represents the activation function weight of the Lth layer, such as the activation function weight 2 of the Lth layer; f1 represents the activation function weight of the first layer, such as the activation function weight 1 of the first layer; W1 represents the activation function weight of the first layer, such as the activation function weight 2 of the second layer; Represents X MLP The latent feature vector of the user in ; Represents X MLP The latent feature vector of the item in, b1 represents the bias of the first layer; b L represents the bias of the Lth layer.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the intelligent course recommendation method based on knowledge mastering data as described in any one of claims 1 to 5 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the intelligent course recommendation method based on knowledge mastering data as described in any one of claims 1 to 5 is implemented.

Images