Method and related device for automatically recommending training courses

Abstract

The present disclosure provides a method and related equipment for automatically recommending training courses, which obtains multiple course recall candidate sets by analyzing and processing user personal information and course information collected by an education platform from multiple aspects, and selects a corresponding number of courses from the multiple course recall candidate sets according to a predetermined weight, thereby obtaining a multi-channel course recall candidate set for recommending courses to users. It realizes grouping users for analysis and calculation according to user job classification, thereby improving calculation efficiency and data analysis accuracy, and the analysis results are more in line with the training needs of enterprises and individuals, thereby achieving accurate course recommendation.

Description

Technical Field

[0001] The present disclosure relates to the field of education and training technology, and in particular to a method for automatically recommending training courses and related equipment. Background Art

[0002] In today's context of Internet information overload, most people, in addition to actively obtaining information, have higher and higher requirements for the quality and characteristics of passively obtained information. Therefore, software platforms based on scenario services urgently need to improve the efficiency of information distribution. When the efficiency is improved to a certain level, it will bring users a brand new experience.

[0003] Large-scale enterprise online education platforms urgently need to change from the traditional means of manual maintenance or simple traffic sorting to push resource information to personalized, automated, and intelligent push methods. The reason for using recommendation models is that different users have different needs for training and learning resources, and users' preferences for training courses are becoming more and more diverse, and unified recommendations for all users cannot meet users' personalized needs. Usually, operation and maintenance personnel will only recommend content that most people like, while recommendation models can push personalized course resources based on the individual needs of users. However, the mainstream Internet recommendation model is not suitable for training course recommendations on enterprise-level education platforms. In order to meet the needs of enterprise-level education platforms for recommending courses, a new method for automatically recommending training courses is needed. Summary of the invention

[0004] In view of this, the purpose of the present disclosure is to provide a method for automatically recommending training courses and related devices.

[0005] Based on the above purpose, the present disclosure provides a method for automatically recommending training courses, which is applied to an enterprise-level education platform, comprising:

[0006] Collecting personal data and behavioral data of users using the education platform, and extracting user feature vectors from the personal data and the behavioral data;

[0007] Sort the courses on the education platform according to the maximum number of times the user has studied and the highest evaluation score, and obtain a first course recall candidate set based on the number of times the user has studied and the evaluation score;

[0008] For any user, determine the user group to which the user belongs; calculate the user similarity between the single user and the remaining users in the same user group according to the user feature vector, sort the remaining users from large to small according to the user similarity, and obtain the second course recall candidate set based on the user similarity according to the sorting result and the behavior data; wherein the user group is pre-divided according to the user job category;

[0009] Calculate the course similarity between the course studied by the user and the remaining courses studied by other users who have studied the same course, sort the remaining courses from large to small according to the course similarity, and obtain a third course recall candidate set based on the course similarity;

[0010] A certain number of the courses are selected from the first course recall alternative set, the second course recall alternative set and the third course recall alternative set according to predetermined weights, and a multi-channel course recall alternative set is generated and recommended to the user.

[0011] Based on the same invention purpose, the present disclosure also provides a device for automatically recommending training courses, comprising:

[0012] A data collection module, configured to collect personal data and behavior data of users using the education platform, and extract user feature vectors from the personal data and the behavior data;

[0013] A first course recall candidate set generation module is configured to sort the courses on the education platform according to the maximum number of times the user has studied and the highest evaluation score, and obtain the first course recall candidate set based on the number of times the user has studied and the evaluation score;

[0014] The second course recall candidate set generation module is configured to determine, for any user, the user group to which the user belongs; calculate the user similarity between the single user and the remaining users in the same user group according to the user feature vector, sort the remaining users from large to small according to the user similarity, and obtain the second course recall candidate set based on the user similarity according to the sorting result and the behavior data; wherein the user group is pre-divided according to the user job category;

[0015] A third course recall candidate set generation module is configured to calculate the course similarity between the course studied by the user and the remaining courses studied by other users who have studied the same course, sort the remaining courses from large to small according to the course similarity, and obtain a third course recall candidate set based on the course similarity;

[0016] The module for generating a multi-course recall alternative set is configured to select a certain number of the courses from the first course recall alternative set, the second course recall alternative set and the third course recall alternative set according to predetermined weights, and generate a multi-course recall alternative set for recommendation to the user.

[0017] Based on the same invention purpose, the present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements a method for automatically recommending training courses when executing the program.

[0018] Based on the same invention purpose, the present disclosure also provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute a method for automatically recommending training courses.

[0019] From the above description, it can be seen that the method and related equipment for automatically recommending training courses provided by the present disclosure collect user information and course information in various ways, and after analysis, accurately recommend training courses that better meet enterprise requirements and user needs based on the user's learning situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the present disclosure or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 A flowchart of a method for automatically recommending training courses provided by an embodiment of the present disclosure;

[0022] Figure 2 A structural diagram of an automatic training course recommendation device provided in an embodiment of the present disclosure;

[0023] Figure 3 A structural diagram of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0024] In order to make the objectives, technical solutions and advantages of the present disclosure more clearly understood, the present disclosure is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0025] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present disclosure should be understood by people with ordinary skills in the field to which the present disclosure belongs. The "first", "second" and similar words used in the embodiments of the present disclosure do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprising" and similar words mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0026] In the existing technology, the mainstream course recommendation systems can be roughly divided into two categories: content-based recommendation and user-based recommendation. The algorithms used are mainly collaborative filtering and machine learning, that is, the model is trained through historical data and the model indicators are continuously improved through object behavior, so as to ensure the accuracy of the recommendation results.

[0027] However, there are significant differences between the recommendation model of enterprise-level online education platforms and the mainstream recommendation model of the Internet, including:

[0028] The enterprise-level education platform is built on the enterprise-level information platform, which can obtain complete and accurate user portraits, and use user portraits and feature information as the data source for user similarity analysis. However, open Internet platforms are limited by security and other factors, and usually can only obtain less user information, which may lead to large deviations in the analysis process based on collaborative filtering.

[0029] The number of enterprise-level users is usually relatively stable, and the expected total amount of data, computing resources, and analysis efficiency are easy to evaluate with high accuracy, while indicators related to Internet platforms are usually difficult to estimate.

[0030] The course resources on enterprise-level education platforms are usually directly related to the nature of the enterprise, tend to focus on some professional courses, and have clear classifications, which are directly related to users' jobs. Internet platforms have more diverse course resources, and the classifications are sometimes not very clear.

[0031] To solve the above problems, the present invention provides a method for automatically recommending training courses, which collects user information of an enterprise-level education platform, processes the information using different methods to obtain multiple course recall alternative sets, and extracts a corresponding number of courses from each course recall alternative set according to different weights to form a multi-channel course recall alternative set recommended to the user, so as to achieve the purpose of accurately recommending training courses to users using the education platform.

[0032] As an optional embodiment, refer to Figure 1 The method for automatically recommending training courses provided by the present disclosure includes:

[0033] Step S101 : collecting personal data and behavior data of users using the education platform, and extracting user feature vectors from the personal data and the behavior data.

[0034] In this step, the personal data includes: personnel number, name, gender, age, education, major, job classification, years of work experience, affiliated unit and other attributes. The behavioral data refers to the user's search, learning, evaluation and collection of courses on the education platform. Among them, the user's evaluation of the course refers to the evaluation of the course as poor, general, good, excellent, and recommended. When obtaining the user's behavioral data, it is necessary to deduplicate the obtained behavioral data and save the user's most recent 100 behavioral records at most.

[0035] In this step, the characteristics of education level, job level, years of work, cumulative learning time and number of courses are extracted from personal data and behavioral data as the five dimensions of the user feature vector, and the above characteristics are scored according to the preset standards. The scoring results are the eigenvalues ​​of the feature vector corresponding to each dimension. Among them, education level, job level, and years of work constitute the static portrait of the user, and the cumulative learning time and number of courses constitute the dynamic portrait of the user.

[0036] Step S102, sorting the courses on the education platform according to the maximum number of times the user has studied and the highest evaluation score, and obtaining a first course recall candidate set based on the number of times the user has studied and the evaluation score.

[0037] In this step, all courses are sorted in descending order according to the number of learning times in the collected user information, and the result obtained by sorting is the first course recall alternative set based on the user's learning times; similarly, the total evaluation score of the course is obtained according to the user's evaluation of the course, and all courses are sorted in descending order according to the total evaluation score, and the result obtained is the first course recall alternative set based on the evaluation score.

[0038] The course score is calculated based on the different levels of user evaluation of the course, including: 0 points for poor, 1 point for average, 2 points for good, 3 points for excellent, and 5 points for recommended. The total evaluation score of a course is obtained by adding up the scores corresponding to the different evaluations of the course by users.

[0039] Step S103, calculating the user similarity between the single user and the remaining users in the same user group based on the user feature vector, sorting the remaining users from large to small according to the user similarity, and obtaining a second course recall alternative set based on the user similarity based on the sorting result and the behavior data.

[0040] In this step, different users need to learn different courses due to the work requirements of different positions. Users with very different positions need to learn very different courses, while users with similar positions need to learn similar courses. Therefore, users are divided into four user groups according to their job categories, including management group, operation group, technical group and service group. This ensures that there is no obvious difference in the courses learned by different users in the same user group, making the calculated user similarity more representative. The calculated cosine similarity cosθ is used as the user similarity between user A and user B; wherein A i represents the i-th eigenvalue of the user feature vector of the user A, B i The i-th eigenvalue of the user feature vector representing the user B, where n=5.

[0041] Dividing users into management group, operation group, technical group and service group is only a rough division based on the responsibilities corresponding to user positions. When a higher degree of accuracy is required for user similarity, the above user groups can be further divided according to the responsibilities corresponding to different positions, such as further dividing the technical group into wind power group, hydropower group and thermal power group.

[0042] Step S104, calculate the course similarity between the course studied by the user and the remaining courses studied by other users who have studied the same course, sort the remaining courses from large to small according to the course similarity, and obtain a third course recall alternative set based on the course similarity.

[0043] In this step, we will follow The calculated W ab It is regarded as the course similarity between course a and course b; wherein N(a) represents the number of users who have studied course a, and N(b) represents the number of users who have studied course b.

[0044] Step S105, selecting a certain number of the courses from the first course recall alternative set, the second course recall alternative set and the third course recall alternative set according to predetermined weights, and generating a multi-channel course recall alternative set to recommend to the user.

[0045] In this step, before selecting courses from each course recall candidate set, the course recall candidate set needs to be deduplicated to ensure that there are no duplicate courses; and the courses that the user of the recommended course has already studied are removed from the course recall candidate set.

[0046] In this step, when generating the multi-channel recall candidate set, the business management personnel are also required to add or remove specific courses according to the job category of the user to whom the courses are to be recommended.

[0047] As an optional embodiment, in step S101, the user feature vector is scored according to a preset standard, and the standard includes:

[0048] Educational level: 1 point for high school, 2 points for college, 3 points for undergraduate, 4 points for master, and 5 points for doctorate.

[0049] Job level: junior 1 point, intermediate 2 points, senior 3 points.

[0050] Years of working experience: 1 point for 1-3 years, 2 points for 3-5 years, 3 points for 5-10 years, 4 points for 10-20 years, 5 points for more than 20 years.

[0051] Cumulative study time: 1 point for 1-3 hours, 2 points for 3-10 hours, 3 points for 10-30 hours, 4 points for 30-60 hours, and 5 points for more than 60 hours.

[0052] Number of courses studied: 1 point for 1-5, 2 points for 5-10, 3 points for 10-20, 4 points for 20-40, 5 points for more than 40.

[0053] Suppose a user has a master's degree, a senior position, 3 years of work experience, 3 hours of cumulative learning time, and has studied 3 courses. According to the above standards, the user feature vector of this user can be obtained as (4, 3, 1, 1, 1).

[0054] As an optional embodiment, in step S103, the user similarity is calculated and the second course recall candidate set is obtained, wherein, referring to Table 1, the calculated user similarity is:

[0055] Table 1. User similarity calculation results

[0056]

[0057] Sort users 2-9 in Table 1 in descending order of user similarity, select the top three users with the greatest user similarity, and obtain the three courses that these three users have studied most recently. After removing duplicates, the obtained nine courses are used as the second course recall candidate set.

[0058] As an optional embodiment, in step S104, the course similarity is calculated and the third course recall candidate set is obtained, wherein, referring to Table 2, the calculated course similarity is:

[0059] Table 2. Course similarity calculation results

[0060]

[0061]

[0062] Table 2 calculates the similarity between the remaining courses and course a. When a user studies multiple courses, it is necessary to calculate the similarity between the remaining courses and each course separately, and sort them uniformly according to the calculation results to obtain the third course recall candidate set.

[0063] As an optional embodiment, in step S105, courses are selected from each course recall alternative set according to predetermined weights to generate multiple course recall alternative sets. The predetermined weights are set by the staff of the education platform. For example, the weights of the two first course recall alternative sets are set to 20%, the weight of the second course recall alternative set is set to 30%, and the weight of the third course recall alternative set is set to 30%.

[0064] The first two courses ranked high are selected from the first course recall candidate set based on the number of user learning times and course evaluation scores after deduplication according to the above weights, the first three courses ranked high are selected from the second course recall candidate set based on user similarity after deduplication, and the first three courses are selected from the third course recall candidate set based on course similarity after deduplication. The resulting 10 courses are recommended to users as a multi-course recall candidate set.

[0065] The method for automatically recommending training courses provided by the present disclosure collects various user data, analyzes the user data from different angles and generates corresponding course recall candidate sets, selects courses from each course recall candidate set according to preset weights to generate a multi-channel course recall candidate set recommended to the user. It realizes grouping users according to user job classification for analysis and calculation, thereby improving calculation efficiency and data analysis accuracy, and the analysis results are more in line with the training needs of enterprises and individuals, thereby achieving accurate course recommendation.

[0066] It should be noted that the method of the embodiment of the present disclosure can be performed by a single device, such as a computer or a server. The method of the present embodiment can also be applied in a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present disclosure, and the multiple devices will interact with each other to complete the described method.

[0067] It should be noted that the above describes some embodiments of the present disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0068] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present disclosure also provides an automatic training course recommendation device.

[0069] refer to Figure 2 , the automatic training course recommendation device comprises:

[0070] A data collection module 201 is configured to collect personal data and behavior data of users using the education platform, and extract user feature vectors from the personal data and the behavior data;

[0071] The first course recall candidate set generation module 202 is configured to sort the courses on the education platform according to the maximum number of times the user has studied and the highest evaluation score, and obtain the first course recall candidate set based on the number of times the user has studied and the evaluation score respectively;

[0072] The second course recall candidate set generation module 203 is configured to determine, for any user, the user group to which the user belongs; calculate the user similarity between the single user and the remaining users in the same user group according to the user feature vector, sort the remaining users from large to small according to the user similarity, and obtain the second course recall candidate set based on the user similarity according to the sorting result and the behavior data; wherein the user group is pre-divided according to the user job category;

[0073] The third course recall candidate set generating module 204 is configured to calculate the course similarity between the course studied by the user and the remaining courses studied by other users who have studied the same course, sort the remaining courses from large to small according to the course similarity, and obtain the third course recall candidate set based on the course similarity;

[0074] The multiple course recall alternative set generation module 205 is configured to select a certain number of the courses from the first course recall alternative set, the second course recall alternative set and the third course recall alternative set according to predetermined weights, and generate a multiple course recall alternative set for recommendation to the user.

[0075] For the convenience of description, the above device is described by dividing it into various modules according to its functions. Of course, when implementing the present disclosure, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0076] The device of the above embodiment is used to implement the corresponding automatic training course recommendation method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0077] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments, the present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for automatically recommending training courses described in any of the above embodiments is implemented.

[0078] Figure 3 A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 in the device.

[0079] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0080] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0081] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0082] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).

[0083] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0084] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.

[0085] The electronic device of the above embodiment is used to implement the corresponding method for automatically recommending training courses in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0086] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present disclosure also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method for automatically recommending training courses as described in any of the above embodiments.

[0087] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0088] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the method for automatically recommending training courses as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0089] It should be noted that the embodiments of the present disclosure may be further described in the following manner:

[0090] A method for automatically recommending training courses, applied to an enterprise-level education platform, includes:

[0091] Collecting personal data and behavioral data of users using the education platform, and extracting user feature vectors from the personal data and the behavioral data;

[0092] Sort the courses on the education platform according to the maximum number of times the user has studied and the highest evaluation score, and obtain a first course recall candidate set based on the number of times the user has studied and the evaluation score;

[0093] For any user, determine the user group to which the user belongs; calculate the user similarity between the single user and the remaining users in the same user group according to the user feature vector, sort the remaining users from large to small according to the user similarity, and obtain the second course recall candidate set based on the user similarity according to the sorting result and the behavior data; wherein the user group is pre-divided according to the user job category;

[0094] Calculate the course similarity between the course studied by the user and the remaining courses studied by other users who have studied the same course, sort the remaining courses from large to small according to the course similarity, and obtain a third course recall candidate set based on the course similarity;

[0095] A certain number of the courses are selected from the first course recall alternative set, the second course recall alternative set and the third course recall alternative set according to predetermined weights, and a multi-channel course recall alternative set is generated and recommended to the user.

[0096] As an optional embodiment, extracting the user feature vector includes: extracting the educational level, job level, years of work experience, cumulative learning time and number of learning courses from the personal data and the behavioral data and scoring them respectively, and regarding the results of each scoring as the eigenvalues ​​of the five-dimensional feature vectors of the user feature vector.

[0097] As an optional embodiment, the user groups pre-divided according to the job categories include: a management group, an operation group, a technical group and a service group.

[0098] As an optional embodiment, calculating the user similarity includes: The calculated cosine similarity cosθ is used as the user similarity between user A and user B; wherein A i represents the i-th eigenvalue of the user feature vector of the user A, B i The i-th eigenvalue of the user feature vector representing the user B, where n=5.

[0099] As an optional embodiment, the obtaining of the recall alternative set based on the second course includes: sorting the remaining users of the same user group from large to small according to their respective user similarities, selecting the three courses most recently studied by the first three users, and considering the obtained nine courses as the recall alternative set based on the second course.

[0100] As an optional embodiment, calculating the course similarity includes: The calculated W ab It is regarded as the course similarity between course a and course b; wherein N(a) represents the number of users who have studied course a, and N(b) represents the number of users who have studied course b.

[0101] As an optional embodiment, generating the multiple recall course alternative sets includes: performing deduplication operations on each of the course recall alternative sets to remove duplicate courses and courses that have been studied by a single user using the education platform; and adding or removing specific courses based on actual conditions.

[0102] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present disclosure (including the claims) is limited to these examples. Based on the concept of the present disclosure, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present disclosure as described above, which are not provided in detail for the sake of simplicity.

[0103] In addition, to simplify the description and discussion, and in order not to make the embodiments of the present disclosure difficult to understand, the known power / ground connections to the integrated circuit (IC) chips and other components may or may not be shown in the provided figures. In addition, the device can be shown in the form of a block diagram to avoid making the embodiments of the present disclosure difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform on which the embodiments of the present disclosure will be implemented (that is, these details should be fully within the scope of understanding of those skilled in the art). Where specific details (e.g., circuits) are set forth to describe exemplary embodiments of the present disclosure, it is apparent to those skilled in the art that the embodiments of the present disclosure can be implemented without these specific details or with changes in these specific details. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0104] Although the present disclosure has been described in conjunction with specific embodiments of the present disclosure, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.

[0105] The embodiments of the present disclosure are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present disclosure should be included in the scope of protection of the present disclosure.

Claims

1. A method for automatically recommending training courses, applied to an enterprise-level education platform, comprising: Collecting personal data and behavioral data of users using the education platform, and extracting user feature vectors from the personal data and the behavioral data; Sort the courses on the education platform according to the maximum number of times the user has studied and the highest evaluation score, and obtain a first course recall candidate set based on the number of times the user has studied and the evaluation score; For any user, determine the user group to which the user belongs; calculate the user similarity between the single user and the remaining users in the same user group according to the user feature vector, sort the remaining users from large to small according to the user similarity, and obtain the second course recall candidate set based on the user similarity according to the sorting result and the behavior data; wherein the user group is pre-divided according to the user job category; Calculate the course similarity between the course studied by the user and the remaining courses studied by other users who have studied the same course, sort the remaining courses from large to small according to the course similarity, and obtain a third course recall candidate set based on the course similarity; A certain number of the courses are selected from the first course recall alternative set, the second course recall alternative set and the third course recall alternative set according to predetermined weights, and a multi-channel course recall alternative set is generated and recommended to the user.

2. The method according to claim 1, wherein: Extracting the user feature vector includes: The educational level, job grade, years of work experience, cumulative learning time and number of learning courses are extracted from the personal data and the behavioral data and scored respectively, and the scoring results are regarded as the eigenvalues ​​of the five-dimensional feature vectors of the user feature vector.

3. The method according to claim 1, wherein: The user groups pre-divided according to the job categories include: a management group, an operation group, a technical group and a service group.

4. The method according to claim 2, wherein: Calculating the user similarity includes: Will follow The calculated cosine similarity cosθ is used as the user similarity between user A and user B; Among them, A i represents the i-th eigenvalue of the user feature vector of the user A, B i The i-th eigenvalue of the user feature vector representing the user B, where n=5.

5. The method according to claim 1, wherein: The obtaining of the recall alternative set based on the second course includes: The remaining users in the same user group are sorted from large to small according to their respective user similarities, and the three courses most recently studied by the first three users are respectively selected, and the obtained nine courses are regarded as the recall candidate set based on the second course.

6. The method according to claim 1, wherein: Calculating the course similarity, including: Will be based on The calculated W ab The similarity of the courses considered as course a and course b; Among them, N(a) represents the number of users who have studied the course a, and N(b) represents the number of users who have studied the course b.

7. The method according to claim 1, wherein: Generate the multi-way recall course alternative set, including: Performing a deduplication operation on each of the recalled alternative courses to remove duplicate courses and courses that have been studied by a single user using the education platform; Specific courses mentioned above may be added or removed based on actual circumstances.

8. An automatic training course recommendation device, comprising: A data collection module, configured to collect personal data and behavior data of users using the education platform, and extract user feature vectors from the personal data and the behavior data; A first course recall candidate set generation module is configured to sort the courses on the education platform according to the maximum number of times the user has studied and the highest evaluation score, and obtain the first course recall candidate set based on the number of times the user has studied and the evaluation score; A second course recall alternative set generation module is configured to determine, for any user, a user group to which the user belongs; Calculating the user similarity between a single user and the remaining users in the same user group based on the user feature vector, sorting the remaining users from large to small according to the user similarity, and obtaining a second course recall candidate set based on the user similarity based on the sorting result and the behavior data; wherein the user group is pre-divided based on the user job category; A third course recall candidate set generation module is configured to calculate the course similarity between the course studied by the user and the remaining courses studied by other users who have studied the same course, sort the remaining courses from large to small according to the course similarity, and obtain a third course recall candidate set based on the course similarity; The module for generating a multi-course recall alternative set is configured to select a certain number of the courses from the first course recall alternative set, the second course recall alternative set and the third course recall alternative set according to predetermined weights, and generate a multi-course recall alternative set for recommendation to the user.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 7 when executing the program. 10 . A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the method according to claim 1 .

Images