WO2022142043A1 - Course recommendation method and apparatus, device, and storage medium - Google Patents

Abstract

Provided are a course recommendation method and apparatus, a device, and a storage medium. The method comprises: performing word segmentation on a course name data set to be updated, and updating a course name segmentation library; according to a user preference label data set to be updated, updating a user preference label library; on the basis of a data update request, obtaining a course name segmentation result from a course name segmentation library and generating a word vector to obtain a target course name segmentation word vector set, and obtaining user preference label data from the user preference label library and generating a word vector to obtain a target user preference label word vector set; according to the target course name segmentation word vector set and the target user preference label word vector set, generating a similarity matrix and updating a similarity matrix library; and according to the similarity matrix library, obtaining a target course recommendation result corresponding to each user identifier. After the course and/or the user preference label is updated, dynamic and in-depth mining of a learning course of real interest to the user is achieved.

Description

Course recommendation method, device, equipment and storage medium

This application claims the priority of the Chinese patent application with the application number 2020116094875 and the invention titled "Course Recommendation Method, Apparatus, Equipment and Storage Medium", which was filed with the China Patent Office on December 30, 2020, the entire contents of which are incorporated by reference in in this application.

technical field

The present application relates to the field of artificial intelligence technology, and in particular, to a course recommendation method, apparatus, device and storage medium.

Background technique

The current online teaching platform does not have the function of intelligently recommending learning courses for users. Users only monotonically study the courses manually pushed by the platform system administrator in the background. The inventor realizes that this recommendation method makes employees passive and difficult to find their own feelings. Interesting courses cannot mobilize users' enthusiasm for learning.

technical problem

The purpose is to solve the technical problem that users of the existing online teaching platform can only monotonically study the courses manually pushed by the platform system administrator in the background, and cannot mobilize users' enthusiasm for learning.

technical solutions

The main purpose of this application is to provide a course recommendation method, device, equipment and storage medium, which aims to solve the problem that users of the existing online teaching platform can only monotonically study the courses manually pushed by the platform system administrator in the background, and cannot Technical issues to mobilize users' enthusiasm for learning.

In order to achieve the above purpose of the invention, the present application proposes a course recommendation method, which includes:

Get data update request;

When the data update request is a recommendation request after course update, obtain the course name data set to be updated, and use the preset course batch processing parameters to segment each course name in the to-be-updated course name data set, and obtain The word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated, according to the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated Update the course name thesaurus;

When the data update request is a recommendation request after the user preference label is updated, the user preference label data set to be updated is obtained, and the user preference label database is updated according to the user preference label data set to be updated using preset user batch processing parameters ;

Based on the data update request, the course name word segmentation result is obtained from the course name word segmentation database, and a target course name word segmentation result set is obtained, and each of the course name word segmentation results in the target course name word segmentation result set is performed generating a word vector, obtaining a target course name word segmentation word vector set corresponding to the target course name word segmentation result set;

Obtain user preference tag data from the user preference tag library based on the data update request, obtain a target user preference tag data set, and perform word vector generation on each of the user preference tag data in the target user preference tag data set respectively. , obtain the target user preference label word vector set corresponding to the target user preference label data set;

The similarity matrix is generated according to the target course name word segmentation word vector set and the target user preference label word vector set, and the similarity matrix between the courses to be updated and the preference labels is obtained. Matrix update similarity matrix library;

Obtain a course recommendation update request based on the data update request, use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain the user to be recommended. Each user of the identification set identifies the corresponding target course recommendation result.

The present application also proposes a course recommendation device, which includes:

The request acquisition module is used to acquire data update requests;

The course name word segmentation database update module is used to obtain the data set of the course name to be updated when the data update request is a recommendation request after the course update, and use the preset course batch processing parameters to respectively update the course name to be updated. Perform word segmentation for each course name in the data set, and obtain the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated. The word segmentation result of the course name corresponding to each name updates the course name word segmentation database;

The user preference tag library update module is used to obtain the user preference tag data set to be updated when the data update request is a recommendation request after the user preference tag is updated, and use preset user batch processing parameters according to the user preference tag to be updated. The preference tag dataset updates the user preference tag library;

The target course name word segmentation word vector set determination module is used to obtain the course name word segmentation result from the course name word segmentation database based on the data update request, and obtain the target course name word segmentation result set, respectively for the target course name. The word vector generation is performed on each of the course name word segmentation results in the word segmentation result set, and the target course name word segmentation word vector set corresponding to the target course name word segmentation result set is obtained;

The target user preference tag word vector set determination module is configured to obtain user preference tag data from the user preference tag library based on the data update request, and obtain a target user preference tag data set, and respectively analyze the target user preference tag data set in the target user preference tag data set. Perform word vector generation on each of the user preference tag data to obtain a target user preference tag word vector set corresponding to the target user preference tag data set;

The similarity matrix generation module is used to generate the similarity matrix according to the target course name word segmentation word vector set and the target user preference tag word vector set, and obtain the similarity matrix of the course to be updated and the preference tag, according to the to-be-updated course and preference tag similarity matrix. Updated course and preference label similarity matrix Update similarity matrix library;

A course recommendation module, configured to obtain a course recommendation update request based on the data update request, and use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain Each user identifier of the user identifier set to be recommended corresponds to a target course recommendation result.

The present application also proposes a computer device, comprising a memory and a processor, wherein the memory stores a computer program, wherein the processor implements the following method steps when executing the computer program:

Get data update request;

When the data update request is a recommendation request after course update, obtain the course name data set to be updated, and use the preset course batch processing parameters to segment each course name in the to-be-updated course name data set, and obtain The word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated, according to the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated Update the course name thesaurus;

When the data update request is a recommendation request after the user preference label is updated, the user preference label data set to be updated is obtained, and the user preference label database is updated according to the user preference label data set to be updated using preset user batch processing parameters ;

Based on the data update request, the course name word segmentation result is obtained from the course name word segmentation database, and a target course name word segmentation result set is obtained, and each of the course name word segmentation results in the target course name word segmentation result set is performed generating a word vector, obtaining a target course name word segmentation word vector set corresponding to the target course name word segmentation result set;

Obtain user preference tag data from the user preference tag library based on the data update request, obtain a target user preference tag data set, and perform word vector generation on each of the user preference tag data in the target user preference tag data set respectively. , obtain the target user preference label word vector set corresponding to the target user preference label data set;

The similarity matrix is generated according to the target course name word segmentation word vector set and the target user preference label word vector set, and the similarity matrix between the courses to be updated and the preference labels is obtained. Matrix update similarity matrix library;

Obtain a course recommendation update request based on the data update request, use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain the user to be recommended. Each user of the identification set identifies the corresponding target course recommendation result.

The present application also proposes a computer-readable storage medium on which a computer program is stored, wherein when the computer program is executed by a processor, the following method steps are implemented:

Get data update request;

When the data update request is a recommendation request after course update, obtain the course name data set to be updated, and use the preset course batch processing parameters to segment each course name in the to-be-updated course name data set, and obtain The word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated, according to the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated Update the course name thesaurus;

When the data update request is a recommendation request after the user preference label is updated, the user preference label data set to be updated is obtained, and the user preference label database is updated according to the user preference label data set to be updated using preset user batch processing parameters ;

Based on the data update request, the course name word segmentation result is obtained from the course name word segmentation database, and a target course name word segmentation result set is obtained, and each of the course name word segmentation results in the target course name word segmentation result set is performed generating a word vector, obtaining a target course name word segmentation word vector set corresponding to the target course name word segmentation result set;

Obtain user preference tag data from the user preference tag library based on the data update request, obtain a target user preference tag data set, and perform word vector generation on each of the user preference tag data in the target user preference tag data set respectively. , obtain the target user preference label word vector set corresponding to the target user preference label data set;

The similarity matrix is generated according to the target course name word segmentation word vector set and the target user preference label word vector set, and the similarity matrix between the courses to be updated and the preference labels is obtained. Matrix update similarity matrix library;

Obtain a course recommendation update request based on the data update request, use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain the user to be recommended. Each user of the identification set identifies the corresponding target course recommendation result.

beneficial effect

In the course recommendation method, device, device and storage medium of the present application, firstly, when the data update request is a recommendation request after course update, using the preset course batch processing parameters to separately perform word segmentation for each course name in the course name data set to be updated and analyze Update the course name thesaurus database. When the data update request is a recommendation request after the user preference label is updated, the preset user batch processing parameters are used to update the user preference label database according to the user preference label data set to be updated. Obtain the result of the course name word segmentation from the course name word segmentation database and generate the word vector to obtain the target course name word segmentation word vector set. Based on the data update request, obtain the user preference label data from the user preference label database and generate the word vector to obtain the target course name word segmentation word vector Then, the similarity matrix is generated according to the target course name word segmentation word vector set and the target user preference label word vector set to obtain the similarity matrix of the course to be updated and the preference label, and the similarity matrix library is updated. Finally, by adopting the preset recommendation rules Recommend courses according to the similarity matrix library and the user identification set carried in the course recommendation update request, and obtain the target course recommendation results corresponding to each user identification of the user identification set to be recommended. After the tag is updated, it realizes dynamic and in-depth mining of learning courses that users are really interested in, which is conducive to improving the timeliness of course recommendation, improving the accuracy of course recommendation, and mobilizing users' enthusiasm for learning.

Description of drawings

1 is a schematic flowchart of a course recommendation method according to an embodiment of the application;

FIG. 2 is a schematic block diagram of the structure of a course recommendation apparatus according to an embodiment of the application;

FIG. 3 is a schematic structural block diagram of a computer device according to an embodiment of the present application.

The realization, functional features and advantages of the present application will be further described with reference to the accompanying drawings in conjunction with the embodiments.

Embodiments of the present invention

In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application.

In order to solve the technical problem that the users of the online teaching platform in the prior art can only monotonically study the courses manually pushed by the platform system administrator in the background and cannot mobilize the users' enthusiasm for learning, the present application proposes a course recommendation method. Applied in the field of artificial intelligence technology. The course recommendation method updates the course name word segmentation database after the updated course name is segmented after the course update, updates the user preference label database according to the user preference label data set to be updated after the user preference label is updated, and according to the updated type ( That is, course update, user preference label update) to obtain the word segmentation result of the course name and the user preference label data, respectively generate the word vector, generate a similarity matrix according to the word vector generation result, and perform course recommendation according to the similarity matrix, so that after the course update And/or the user preference tag is updated to realize dynamic and in-depth mining of the learning courses that the user is really interested in, which is conducive to improving the timeliness of course recommendation, improving the accuracy of course recommendation, and helping to mobilize the enthusiasm of users for learning.

Referring to FIG. 1, an embodiment of the present application provides a course recommendation method, the method includes:

S1: Get data update request;

S2: When the data update request is a recommendation request after course update, obtain the course name data set to be updated, and use preset course batch processing parameters to segment each course name in the to-be-updated course name data set. , to obtain the course name word segmentation results corresponding to each of the course names corresponding to the course name data set to be updated, and according to the course name corresponding to each of the course names corresponding to the to-be-updated course name data set. The word segmentation results update the course name word segmentation database;

S3: When the data update request is a recommendation request after the user preference label is updated, obtain the user preference label data set to be updated, and update the user preference according to the user preference label data set to be updated using preset user batch processing parameters tag library;

S4: Obtain the course name word segmentation result from the course name word segmentation database based on the data update request, obtain a target course name word segmentation result set, and separately perform word segmentation for each course name in the target course name word segmentation result set As a result, word vector generation is performed to obtain the target course name word segmentation word vector set corresponding to the target course name word segmentation result set;

S5: Acquire user preference tag data from the user preference tag library based on the data update request, obtain a target user preference tag data set, and perform a word analysis on each of the user preference tag data in the target user preference tag data set. vector generation, to obtain the target user preference tag word vector set corresponding to the target user preference tag data set;

S6: Generate a similarity matrix according to the target course name word segmentation word vector set and the target user preference tag word vector set to obtain a similarity matrix between the courses to be updated and the preference labels, and according to the to-be-updated courses and preference labels The similarity matrix updates the similarity matrix library;

S7: Obtain a course recommendation update request based on the data update request, use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain the to-be-recommended course recommendation Each user ID of the set of user IDs corresponds to the target course recommendation result.

In this embodiment, when the data update request is a recommendation request after course update, the preset course batch processing parameters are used to separate words for each course name in the updated course name data set and update the course name word segmentation database. When the data update request It is the recommendation request after the user preference label is updated, using the preset user batch processing parameters to update the user preference label database according to the user preference label data set to be updated, and then obtain the course name word segmentation result from the course name word segmentation database based on the data update request. Perform word vector generation to obtain the target course name word segmentation word vector set, obtain the user preference label data from the user preference label library based on the data update request, and perform word vector generation to obtain the target course name word segmentation word vector set, and then segment the word vector according to the target course name. Set and target user preference tag word vector set to generate similarity matrix to obtain the similarity matrix of courses and preference tags to be updated and update the similarity matrix library, and finally adopt preset recommendation rules according to similarity matrix library and course recommendation update request carry The user identification set is recommended for course recommendation, and the target course recommendation results corresponding to each user identification of the user identification set to be recommended are obtained, so that dynamic and in-depth mining is realized after the course is updated and/or the user preference tag is updated. Learning courses that users are really interested in can help improve the timeliness of course recommendation, improve the accuracy of course recommendation, and help mobilize users' enthusiasm for learning.

For S1, it can obtain the data update request input by the user, or obtain the data update request sent by the third-party application system (for example, use the Apache Spark Streaming technology to obtain the data update request sent by the Kafka message middleware), or realize this Data update request triggered by the applied program file.

Kafka is a high-throughput distributed publish-subscribe messaging system.

The data update request includes: recommendation request after course update and recommendation request after user preference tag update.

The recommendation request after course update refers to the request for course recommendation after the course name is updated.

The recommendation request after the user preference tag is updated refers to a request for course recommendation after the user preference tag is updated.

For S2, when the data update request is a recommendation request after course update, the data set of the course name input by the user to be updated can be obtained, or the data set to be updated sent by a third-party application system (for example, Kafka message middleware) can be obtained. The course name dataset.

The course name data set to be updated includes: course name identifiers, course names, course name data to be added, and course name data to be deleted, each course name identifier corresponds to a course name, and each course name identifier corresponds to a new course name Added course title data and/or a course title data that needs to be deleted.

The course name word segmentation database includes: course name identification, course name word segmentation results, each course name identification corresponds to a course name word segmentation result. Each of the course title word segmentation results includes at least one word.

Optionally, the course name thesaurus is stored in the local database.

The preset course batch processing parameters include: batch interval data, block interval data, sliding window size, and sliding interval data.

The sliding window size refers to the size of the sliding window.

The sliding interval data refers to the time interval of the sliding window.

For example, if the number of CPU cores of each computer is 10, then the batch interval data is set to 2S, and the block interval data is set to 200ms, so that the number of tasks corresponding to each batch is 10 (that is, the batch interval data set 2S divided by the block interval data setting 200ms), so as to make full use of each CPU core without losing the computing performance of the computer, which is not specifically limited in this example.

Among them, the sliding window size and the sliding interval data are set as integer multiples of the batch interval data. Integer multiples include but are not limited to: 1 times, 2 times, 3 times, 4 times, and 5 times.

Wherein, the word segmentation tool is used to segment each course name in the course name data to be added in the to-be-updated course name data set according to the preset course batch processing parameters, and all the words obtained by word segmentation of a course name are used as A course name word segmentation result, adding all the course name word segmentation results to the course name word segmentation database; directly deleting the course name data that needs to be deleted in the to-be-updated course name data set from the course name word segmentation database.

For S3, when the data update request is a recommendation request after the user preference label is updated, the user preference label data set to be updated input by the user may be obtained, or the user preference label data to be updated may be obtained from a third-party software system set.

Optionally, the tag word segmentation library is stored in the local database.

The user preference tag data set to be updated includes: user identification, preference tags to be added, and preference tags to be deleted. The user identifier may be an identifier that uniquely identifies a user, such as a user name, a user ID, or the like.

The tag word segmentation library includes: user IDs and user preference tag data, each user ID corresponds to a user preference tag data. User preference tag data includes one or more preference tags. Each user preference tag data corresponds to one user.

The preset user batch processing parameters include: batch interval data, block interval data, sliding window size, and sliding interval data.

Wherein, using preset user batch processing parameters to add preference labels to be added in the user preference label data set to be updated to the user preference label database, and add preference labels to be deleted in the user preference label data set to be updated Removed from user preference tag library.

For S4, when the data update request is a recommendation request after course update, the word segmentation result of the updated course name is obtained from the course name word segmentation database, and when the data update request is a recommendation after the user preference tag is updated Obtain all course name word segmentation results from the course name word segmentation database when requesting, and use all the obtained course name word segmentation results as the target course name word segmentation result set; respectively, for each of the courses in the target course name word segmentation result set The name word segmentation result is used to generate word vectors, and all the generated word vectors are used as the target course name word segmentation word vector set corresponding to the target course name word segmentation result set. That is to say, each of the course title word segmentation results corresponds to a course title word segmentation word vector, and each course name word segmentation word vector corresponds to one of the course names.

For S5, when the data update request is a recommendation request after course update, obtain all user preference label data from the user preference label library, and when the data update request is a recommendation request after user preference label update, from all user preference labels Obtain the updated user preference label data from the user preference label database, and use all the obtained user preference label data as the target user preference label data set; respectively, for each of the user preference label data in the target user preference label data set Perform word vector generation, and use all the generated word vectors as the target user preference label word vector set corresponding to the target user preference label data set. That is to say, each user preference label data corresponds to a user preference label word vector, and each user preference label word vector corresponds to a user.

For S6, a similarity matrix is generated for each course name word vector in the target course name word segmentation word vector set and each user preference label word vector in the target user preference label word vector set, and the generated similarity matrix is used as The similarity matrix of courses to be updated and preference labels.

Among them, each element of the similarity matrix of the course to be updated and the preference label represents the similarity between a word vector of a course name and a word vector of a user preference label. The row of the similarity matrix between the course to be updated and the preference label corresponds to the word vector of the course name, and the column corresponds to the word vector of the user's preference label, or the column of the similarity matrix of the course to be updated and the preference label corresponds to the word vector of the course name and the row. User preference tag word vector.

Optionally, a cosine similarity algorithm is used to calculate the cosine similarity of each course name word vector in the target course name word segmentation word vector set and each user preference label word vector in the target user preference label word vector set, to obtain the cosine similarity. Updated course and preference label similarity matrix. That is to say, the element in the similarity matrix of the course to be updated and the preference label is the cosine similarity.

The similarity matrix library includes: user IDs, course and preference tag similarity matrix, and each user ID corresponds to a course and preference tag similarity matrix.

It can be understood that, updating operations in the similarity matrix library according to the similarity matrix of the courses to be updated and preference tags includes but not limited to: covering, adding, partial covering, partial deletion, partial covering and partial deletion.

Optionally, the similarity matrix library is stored in a local Hbase database.

For S7, the course recommendation update request sent by the administrator can be obtained based on the data update request, or the course recommendation generated based on the data update request after the program file implementing the present application receives the end signal for updating the similarity matrix library The update request may also be a course recommendation update request generated by the program files of the present application at preset recommendation intervals.

A course recommendation update request refers to a request for recommending courses to users.

Wherein, the step of generating a course recommendation update request based on the data update request after the program file implementing the present application receives an end signal for updating the similarity matrix library includes: when the data update request is a recommendation after course update When requesting, obtain a preset user identification set, and use the preset user identification set as the user identification set to be recommended carried in the course recommendation update request; when the data update request is a recommendation after the user preference tag is updated When requested, all user identifiers corresponding to the user preference tag data set to be updated are used as the user identifier set to be recommended carried in the course recommendation update request.

The set of user IDs to be recommended, that is, the set of user IDs of users who need to recommend courses. The user identification set to be recommended includes at least one user identification.

Wherein, searching the user identification set to be recommended carried in the course recommendation update request from the similarity matrix library to obtain the similarity matrix of courses and preference tags corresponding to each user identification of the user identification set to be recommended; Use preset recommendation rules to recommend courses according to the similarity matrix of courses and preference tags corresponding to each user ID of the user ID set to be recommended, and obtain the target course recommendation result corresponding to each user ID of the user ID set to be recommended. .

The preset recommendation rules include but are not limited to: the similarity is greater than the preset similarity threshold.

The target course recommendation result includes: a user ID and a set of course names, wherein each user ID corresponds to a set of course names.

The set of course names in the target course recommendation result is a set of course names of courses that the user is really interested in corresponding to the user ID in the target course recommendation result.

In one embodiment, the above-mentioned step of obtaining a data update request includes:

S11: Obtain the data update notification sent by the Kafka message middleware;

S12: when the data update notification is a course name update notification, generate a recommended request after the course update of the data update request;

S13: When the data update notification is a user preference tag update notification, generate a recommendation request after the user preference tag of the data update request is updated.

In this embodiment, the data update notification is obtained from the Kafka message middleware, and the data update request is generated according to the data update notification, so that the present application is suitable for distributed application scenarios; the Kafka message middleware is beneficial for updating the course names that need to be updated. and efficient management of user preference tags that need to be updated.

For S11, use Apache Spark Streaming technology to obtain data update notifications sent by Kafka message middleware.

Data update notifications are notifications of updates to course names and/or user preference tags.

For S12, when the data update notification is a course name update notification, it means that the course name needs to be updated at this time, so a recommendation request after the course update is generated.

For S13, when the data update notification is a user preference tag update notification, it means that the user's preference tag needs to be updated at this time, so a recommendation request after the user preference tag is updated is generated.

In one embodiment, the preset batch processing parameters are used to perform word segmentation on each course name in the to-be-updated course-name data set, to obtain each of the course names corresponding to the to-be-updated course name data set. Corresponding course title word segmentation results, the step of updating the course name word segmentation database according to the course name word segmentation results corresponding to each of the course names corresponding to the to-be-updated course name data sets, including:

S21: Obtain the preset course batch processing parameters;

S22: when the to-be-updated course title data set includes course title data that needs to be deleted, delete the course title thesaurus according to the course title data to be deleted;

S23: When the to-be-updated course title data set includes course title data that needs to be added, use the preset course batch processing parameters to perform a separate processing on each of the course names in the course title data to be added. word segmentation, obtain the word segmentation results of the course names corresponding to each of the course names in the course name data to be added, and divide the courses corresponding to each of the course names in the course name data to be added. The name segmentation result is stored in the course name segmentation database.

This embodiment realizes that each course name in the to-be-updated course name data set is word-segmented and then the course name word segmentation database is updated, which provides a data basis for subsequent similarity matrix generation; and the word segmentation result is updated to the course name word segmentation database, It is beneficial to reuse and improves the efficiency of generating the similarity matrix.

For S21, the preset course batch processing parameters input by the user may be obtained, the preset course batch processing parameters may also be obtained from a database, or the preset course batch processing parameters sent by a third-party application system. It can be understood that, the preset course batch processing parameters can also be written into a program file for realizing the present application.

For S22, when the to-be-updated course title data set includes the course title data that needs to be deleted, it means that the course corresponding to the course title data to be deleted no longer provides services, and at this time, the course title data to be deleted needs to be deleted from all The above-mentioned course name thesaurus is deleted, so as to avoid the situation that the courses recommended to users cannot be learned, and improve the user satisfaction.

For S23, when the to-be-updated course name data set includes the course name data that needs to be added, it means that the course corresponding to the new course name data needs to be online to provide services, and the word segmentation tool is used according to the preset The course batch processing parameter uses a sliding window to segment each of the course names in the to-be-added course name data, and uses a set of words obtained by segmenting a course name as a result of the course name segmentation.

Optionally, the word segmentation tool selects the Chinese word segmentation tool of Jaba. It can be understood that, the word segmentation tool can also select other tools from the prior art, which is not specifically limited here.

Jieba Chinese word segmentation tool has also become jieba Chinese word segmentation tool. Jieba Chinese word segmentation tool realizes efficient word graph scanning based on prefix dictionary, generates a directed acyclic graph (DAG) composed of all possible word formations of Chinese characters in a sentence, and uses dynamic The planning algorithm finds the path with the maximum probability, and finds the maximum segmentation combination based on the word frequency. In this application, the precise mode in the word segmentation of "jump" is used to obtain the most accurate word segmentation result.

In one embodiment, the preset batch processing parameters are used to segment each of the course names in the to-be-added course name data, to obtain each of the course names in the to-be-added course name data. Before the step of describing the result of the word segmentation of the course name corresponding to each of the course names, include:

S0231: Obtain batch processing duration monitoring results and the preset course batch processing parameters;

S0232: When the batch processing duration monitoring result is greater than the batch processing interval data of the preset course batch processing parameters, determine the batch processing parameters of the courses to be updated according to the batch processing duration monitoring results;

S0233: Update the preset course batch processing parameters according to the to-be-updated course batch processing parameters.

This embodiment realizes that the preset course batch processing parameters are updated before the step of using the preset course batch processing parameters to segment each of the course names in the to-be-added course name data respectively, which is beneficial to Adjust the batch processing parameters of the preset courses according to the monitoring results of the batch processing time, which avoids the problem that the course name dataset to be updated is not completely updated to the course name word segmentation database due to the unreasonable settings of the batch processing parameters of the preset courses.

For S0231, the batch processing duration monitoring result can be obtained from the database, or the batch processing duration monitoring result can be obtained from the cache.

The preset course batch processing parameters input by the user can be obtained, the preset course batch processing parameters can also be obtained from a database, or the preset course batch processing parameters sent by a third-party application system.

For S0232, when the batch processing duration monitoring result is greater than the batch processing interval data of the preset course batch processing parameters, it means that the last batch processing duration exceeds the expected batch processing interval data, and the preset batch processing interval data needs to be adjusted at this time. Set the batch interval data of the batch processing parameters of the course to avoid the occurrence of more than expected batch interval data in the next batch, and use the preset batch interval adjustment rules to perform batch interval according to the batch processing duration monitoring results. Data calculation, using the calculated batch interval data as the batch interval data of the batch processing parameters of the course to be updated.

The preset batch interval adjustment rules include but are not limited to: preset ratios.

For S0233, the preset course batch processing parameters are replaced by the to-be-updated course batch processing parameters.

It can be understood that, the methods of steps S0231 to 0233 can also be used to update the preset user batch processing parameters, which will not be repeated here.

In one embodiment, the above-mentioned step of determining the batch processing parameters of the courses to be updated according to the batch processing duration monitoring result includes:

S02331: Obtain the number of CPU cores, block interval data, and preset ratio;

S02332: Multiply the batch processing duration monitoring result and the preset ratio to obtain the batch processing interval data corresponding to the batch processing parameters of the courses to be updated;

S02333: Perform sliding window size calculation and sliding interval data calculation according to the batch interval data corresponding to the batch processing parameters of the courses to be updated, and obtain the size of the sliding window and the data corresponding to the batch processing parameters of the courses to be updated. the sliding interval data.

This embodiment realizes that the batch processing parameters of the courses to be updated are determined according to the batch processing duration monitoring result, the number of CPU cores, and the block interval data, so that the batch processing parameters of the courses to be updated obtained through calculation are in accordance with the program of the present application. Adjustment is made under the condition of the service performance of the server where the file is located, which is beneficial to improve the stability of the course name thesaurus.

For S02331, the number of CPU (central processing unit) cores and block interval data can be obtained from the database, and the number of CPU cores and block interval data can also be obtained from the cache.

The number of CPU cores refers to the number of CPU cores of the server loaded when the program file of the present application provides the course recommendation service.

The block interval data is the duration of each data block processed by the server loaded when the program file of the present application provides the course recommendation service.

Optionally, the value range of the preset ratio is greater than 0.9 and less than 1.1.

For S02332, multiply the batch processing duration monitoring result and the preset ratio, and use the multiplication result as the batch processing interval data corresponding to the batch processing parameter of the course to be updated.

For S02333, obtain a first preset multiple and a second preset multiple; multiply the batch interval data corresponding to the batch processing parameters of the courses to be updated by the first preset multiple to obtain the courses to be updated The size of the sliding window corresponding to the batch processing parameters; multiply the batch interval data corresponding to the batch processing parameters of the courses to be updated by a second preset multiple to obtain the batch processing parameters of the courses to be updated. the sliding interval data.

The first preset multiple is an integer. The second preset multiple is an integer.

In one embodiment, the above-mentioned steps of using preset user batch processing parameters to update the user preference tag database according to the to-be-updated user preference tag data set include:

S31: Acquire a piece of the user preference label data from the user preference label data set to be updated by using the preset user batch processing parameters, and obtain the user preference label data to be processed;

S32: according to the user identification corresponding to the user preference label data to be processed, the preference label that needs to be added in the to-be-processed user preference label data is added to the user preference label library;

S33: according to the user identification corresponding to the to-be-processed user preference label data, delete the preference label that needs to be deleted in the to-be-processed user preference label data from the user preference label library;

S34: Repeat the step of obtaining a piece of user preference label data from the user preference label data set to be updated by using the preset user batch processing parameters to obtain the user preference label data to be processed, until the completion of the The user preference label database is updated by all the user preference label data in the user preference label data set to be updated.

This embodiment implements updating the user preference tag library, which provides a data basis for subsequent similarity matrix generation; and updates the user preference tag data to the user preference tag library, which is beneficial to reuse and improves the efficiency of generating the similarity matrix.

For S31, use the preset user batch processing parameters to extract a plurality of user preference label data from the user preference label data set to be updated in a sliding window manner, and obtain a piece of the user preference label data from the extracted plurality of user preference label data Favorite tag data, the acquired user favorite tag data is used as the user favorite tag data to be processed.

For S32, the user identification corresponding to the user preference tag data to be processed is used to search in the user preference tag library, and when the search is successful, the user identification found in the user preference tag library is used as the to-be-added user identification, otherwise, add the user preference tag data to be processed and the user identification corresponding to the user preference tag data to be processed into the user preference tag database; add the user preference tag data to be processed to all in the user preference label data corresponding to the user identification to be added in the user preference label library.

For S33, it can be understood that the preference tags to be deleted may be all preference tags in a user preference tag data, or part of preference tags in a user preference tag data.

For S34, steps S31 to S34 are repeatedly executed until the user preference label database is updated with all the user preference label data in the to-be-updated user preference label data set.

In an embodiment, the above-mentioned preset recommendation rules are used to perform course recommendation according to the similarity matrix library and the user identification set to be recommended carried in the course recommendation update request, and each user of the user identification set to be recommended is obtained. The steps to identify the respective target course recommendation results, including:

S71: Search the set of user identifiers to be recommended carried in the course recommendation update request in the similarity matrix library, and obtain the respective courses and corresponding user identifiers of the user identifiers to be recommended in the set of user identifiers to be recommended. Like label similarity matrix;

S72: Obtain a preset similarity threshold;

S73: Acquire one of the user identifiers from the set of user identifiers to be recommended as the user identifier to be recommended;

S74: Find out the similarity between courses and preference tags that are greater than the preset similarity threshold from the similarity matrix of the courses and preference tags corresponding to the user IDs to be recommended, and compare all the courses and preference tags found. All course names corresponding to the label similarity are used as the target course recommendation result corresponding to the user ID to be recommended;

S75: Repeat the step of obtaining one of the user IDs from the set of user IDs to be recommended as the user ID to be recommended, until it is determined that each user ID in the set of user IDs to be recommended corresponds to each user ID The target course recommendation result.

This embodiment implements course recommendation according to the preset similarity threshold, the similarity matrix library, and the set of user identifiers to be recommended carried in the course recommendation update request, so that after the course is updated and/or the user preference tag is updated, the Dynamic and in-depth mining of the learning courses that users are really interested in is conducive to improving the timeliness of course recommendation, improving the accuracy of course recommendation, and mobilizing users' enthusiasm for learning.

For S71, search for each user identifier in the set of user identifiers to be recommended carried in the course recommendation update request in the similarity matrix library, and obtain each of the user identifiers in the set of user identifiers to be recommended. User IDs correspond to courses and preference tags similarity matrix.

For S72, the preset similarity threshold input by the user may be obtained, the preset similarity threshold may also be obtained from a database, or the preset similarity threshold sent by a third-party application system may be obtained. It can be understood that, the preset similarity threshold can also be written into a program file for realizing the present application.

For S73, acquire one user identifier from the set of user identifiers to be recommended, and use the acquired user identifier as the user identifier to be recommended.

For S74, find out the similarity between the course and the preference label that is greater than the preset similarity threshold from the similarity matrix of the course and the preference label corresponding to the user ID to be recommended, that is, the preset recommendation rule adopts the similarity It is greater than the preset similarity threshold, indicating that all the course names corresponding to the similarity between the courses and the preference tag and the user preference tag data corresponding to the user ID to be recommended are relatively similar, which is greater than the preset similarity threshold. The courses corresponding to the similarity threshold of the courses and all the course names corresponding to the similarity of the preference tag are the learning courses that the users corresponding to the user ID to be recommended are excavated and are really interested in, so it can be greater than the preset similarity. All the course names corresponding to the similarity between the courses of the threshold and the preference tag are used as the target course recommendation result corresponding to the user ID to be recommended.

For S75, steps S73 to S75 are repeatedly performed until the target course recommendation result corresponding to each of the user identifiers in the set of user identifiers to be recommended is determined.

Referring to FIG. 2, an embodiment of the present application provides a course recommendation device, the device includes:

a request obtaining module 100 for obtaining a data update request;

The course name thesaurus updating module 200 is used to obtain the data set of the course name to be updated when the data update request is a recommendation request after the course update, and use the preset course batch processing parameters for the courses to be updated respectively. Perform word segmentation for each course name in the name data set, and obtain the word segmentation result of each course name corresponding to each of the course names corresponding to the course name data set to be updated. The word segmentation result of the said course name corresponding to each course name updates the course name word segmentation database;

The user preference tag database update module 300 is configured to obtain the user preference tag data set to be updated when the data update request is a recommendation request after the user preference tag is updated, and use preset user batch processing parameters according to the to-be-updated data set. The user preference tag dataset updates the user preference tag library;

The target course name word segmentation word vector set determination module 400 is used to obtain the course name word segmentation result from the course name word segmentation database based on the data update request, and obtain a target course name word segmentation result set, respectively. The word vector generation is performed on each of the course name word segmentation results in the name word segmentation result set, and the target course name word segmentation word vector set corresponding to the target course name word segmentation result set is obtained;

The target user preference tag word vector set determination module 500 is configured to obtain user preference tag data from the user preference tag library based on the data update request, and obtain a target user preference tag data set, and respectively determine the target user preference tag data. Collecting each of the user preference tag data to generate a word vector to obtain a target user preference tag word vector set corresponding to the target user preference tag data set;

The similarity matrix generation module 600 is used to generate a similarity matrix according to the target course name word segmentation word vector set and the target user preference tag word vector set, and obtain the similarity matrix of the course to be updated and the preference tag, according to the Update the similarity matrix library of the similarity matrix of the course to be updated and the preference label;

A course recommendation module 700, configured to obtain a course recommendation update request based on the data update request, and use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, A target course recommendation result corresponding to each user ID of the user ID set to be recommended is obtained.

In this embodiment, when the data update request is a recommendation request after course update, the preset course batch processing parameters are used to separate words for each course name in the updated course name data set and update the course name word segmentation database. When the data update request It is the recommendation request after the user preference label is updated, using the preset user batch processing parameters to update the user preference label database according to the user preference label data set to be updated, and then obtain the course name word segmentation result from the course name word segmentation database based on the data update request. Perform word vector generation to obtain the target course name word segmentation word vector set, obtain the user preference label data from the user preference label library based on the data update request, and perform word vector generation to obtain the target course name word segmentation word vector set, and then segment the word vector according to the target course name. Set and target user preference tag word vector set to generate similarity matrix to obtain the similarity matrix of courses and preference tags to be updated and update the similarity matrix library, and finally adopt preset recommendation rules according to similarity matrix library and course recommendation update request carry The user identification set is recommended for course recommendation, and the target course recommendation results corresponding to each user identification of the user identification set to be recommended are obtained, so that dynamic and in-depth mining is realized after the course is updated and/or the user preference tag is updated. Learning courses that users are really interested in can help improve the timeliness of course recommendation, improve the accuracy of course recommendation, and help mobilize users' enthusiasm for learning.

Referring to FIG. 3 , an embodiment of the present application further provides a computer device. The computer device may be a server, and its internal structure may be as shown in FIG. 3 . The computer equipment includes a processor, memory, a network interface and a database connected by a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium, an internal memory. The nonvolatile storage medium stores an operating system, a computer program, and a database. The memory provides an environment for the execution of the operating system and computer programs in the non-volatile storage medium. The database of the computer equipment is used to store data such as course recommendation methods. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer program, when executed by the processor, implements a course recommendation method. The method for recommending courses includes: obtaining a data update request; when the data update request is a recommendation request after course update, obtaining a data set of course names to be updated, and using preset course batch processing parameters for the to-be-updated courses respectively The word segmentation is performed for each course name in the course name data set, and the word segmentation result of each course name corresponding to each of the course names corresponding to the course name data set to be updated is obtained. The course name word segmentation results corresponding to each of the course names update the course name word segmentation database; when the data update request is a recommendation request after the user preference label is updated, the user preference label data set to be updated is obtained, and a preset The user batch processing parameters update the user preference label database according to the user preference label data set to be updated; obtain the course name word segmentation result from the course name word segmentation database based on the data update request, and obtain the target course name word segmentation result to generate a word vector for each of the course name word segmentation results in the target course name word segmentation result set, respectively, to obtain the target course name word segmentation word vector set corresponding to the target course name word segmentation result set; based on the data update request Obtain user preference label data from the user preference label database, obtain a target user preference label data set, and perform word vector generation on each of the user preference label data in the target user preference label data set to obtain the target user The target user preference label word vector set corresponding to the preference label data set; the similarity matrix is generated according to the target course name segmentation word vector set and the target user preference label word vector set, and the similarity between the course to be updated and the preference label is obtained. matrix, update the similarity matrix library according to the similarity matrix of the courses to be updated and the preference tags; obtain a course recommendation update request based on the data update request, and adopt preset recommendation rules according to the similarity matrix library and the course recommendation The user identification set to be recommended carried in the update request is used for course recommendation, and the target course recommendation result corresponding to each user identification of the user identification set to be recommended is obtained.

In this embodiment, when the data update request is a recommendation request after course update, the preset course batch processing parameters are used to separate words for each course name in the updated course name data set and update the course name word segmentation database. When the data update request It is the recommendation request after the user preference label is updated, using the preset user batch processing parameters to update the user preference label database according to the user preference label data set to be updated, and then obtain the course name word segmentation result from the course name word segmentation database based on the data update request. Perform word vector generation to obtain the target course name word segmentation word vector set, obtain the user preference label data from the user preference label library based on the data update request, and perform word vector generation to obtain the target course name word segmentation word vector set, and then segment the word vector according to the target course name. Set and target user preference tag word vector set to generate similarity matrix to obtain the similarity matrix of courses and preference tags to be updated and update the similarity matrix library, and finally adopt preset recommendation rules according to similarity matrix library and course recommendation update request carry The user identification set is recommended for course recommendation, and the target course recommendation results corresponding to each user identification of the user identification set to be recommended are obtained, so that dynamic and in-depth mining is realized after the course is updated and/or the user preference tag is updated. Learning courses that users are really interested in can help improve the timeliness of course recommendation, improve the accuracy of course recommendation, and help mobilize users' enthusiasm for learning.

An embodiment of the present application further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements a course recommendation method, including the steps of: acquiring a data update request; When it is a recommendation request after course update, obtain the course name data set to be updated, use preset course batch processing parameters to segment each course name in the to-be-updated course name data set, and obtain the to-be-updated course name The word segmentation result of the course name corresponding to each of the course names corresponding to the name data set, and the course name word segmentation database is updated according to the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated. When the data update request is a recommendation request after the user preference label is updated, obtain the user preference label data set to be updated, and use the preset user batch processing parameters to update the user preference label according to the user preference label data set to be updated. library; based on the data update request, obtain the course name word segmentation result from the course name word segmentation database, obtain a target course name word segmentation result set, and respectively segment each of the course names in the target course name word segmentation result set. As a result, word vector generation is performed, and the target course name word segmentation word vector set corresponding to the target course name word segmentation result set is obtained; based on the data update request, the user preference label data is obtained from the user preference label library, and the target user preference label is obtained. data set, respectively perform word vector generation on each of the user preference label data in the target user preference label data set, and obtain the target user preference label word vector set corresponding to the target user preference label data set; according to the target curriculum The similarity matrix is generated between the name segment word vector set and the target user preference tag word vector set to obtain the similarity matrix of the course to be updated and the preference label, and the similarity matrix is updated according to the similarity matrix of the course to be updated and the preference label library; obtain a course recommendation update request based on the data update request, use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain the to-be-recommended Each user ID of the set of user IDs corresponds to the target course recommendation result.

In the above-mentioned course recommendation method, firstly, when the data update request is a recommendation request after course update, the preset course batch processing parameters are used to segment each course name in the updated course name data set, and update the course name word segmentation database. When the data update request is a recommendation request after the user preference label is updated, the preset user batch processing parameters are used to update the user preference label database according to the user preference label data set to be updated, and then obtain from the course name word segmentation database based on the data update request. The result of the word segmentation of the course name is used to generate the word vector to obtain the word vector set of the target course name. Based on the data update request, the user preference label data is obtained from the user preference label database to generate the word vector to obtain the word vector set of the target course name. Name word segmentation word vector set and target user preference tag word vector set are used to generate similarity matrix to obtain the similarity matrix of courses to be updated and preference tags and update the similarity matrix library. Finally, by using preset recommendation rules, according to the similarity matrix library and courses The user identification set carried in the recommendation update request is used for course recommendation, and the target course recommendation result corresponding to each user identification of the user identification set to be recommended is obtained, thereby realizing dynamic, dynamic, Deeply excavating the learning courses that users are really interested in is conducive to improving the timeliness of course recommendation, improving the accuracy of course recommendation, and helping to mobilize the enthusiasm of users for learning.

The computer storage medium can be non-volatile or volatile.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other medium provided in this application and used in the embodiments may include non-volatile and/or volatile memory. Nonvolatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-rate SDRAM (SSRSDRAM), 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), etc.

It should be noted that, herein, the terms "comprising", "comprising" or any other variation thereof are intended to encompass non-exclusive inclusion, such that a process, device, article or method comprising a series of elements includes not only those elements, It also includes other elements not expressly listed or inherent to such a process, apparatus, article or method. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in the process, apparatus, article, or method that includes the element.

The above are only the preferred embodiments of the present application, and are not intended to limit the scope of the patent of the present application. Any equivalent structure or equivalent process transformation made by using the contents of the description and drawings of the present application, or directly or indirectly applied to other related The technical field is similarly included in the scope of patent protection of this application.

Claims (20)

1. A course recommendation method, wherein the method includes:

   Get data update request;

   When the data update request is a recommendation request after course update, obtain the course name data set to be updated, and use the preset course batch processing parameters to segment each course name in the to-be-updated course name data set, and obtain The word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated, according to the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated Update the course name thesaurus;

   When the data update request is a recommendation request after the user preference label is updated, the user preference label data set to be updated is obtained, and the user preference label database is updated according to the user preference label data set to be updated using preset user batch processing parameters ;

   Based on the data update request, the course name word segmentation result is obtained from the course name word segmentation database, and a target course name word segmentation result set is obtained, and each of the course name word segmentation results in the target course name word segmentation result set is performed generating a word vector, obtaining a target course name word segmentation word vector set corresponding to the target course name word segmentation result set;

   Obtain user preference tag data from the user preference tag library based on the data update request, obtain a target user preference tag data set, and perform word vector generation on each of the user preference tag data in the target user preference tag data set respectively. , obtain the target user preference label word vector set corresponding to the target user preference label data set;

   The similarity matrix is generated according to the target course name word segmentation word vector set and the target user preference label word vector set, and the similarity matrix between the courses to be updated and the preference labels is obtained. Matrix update similarity matrix library;

   Obtain a course recommendation update request based on the data update request, use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain the user to be recommended. Each user of the identification set identifies the corresponding target course recommendation result.
2. The course recommendation method according to claim 1, wherein the step of acquiring the data update request comprises:

   Get the data update notification sent by the Kafka message middleware;

   When the data update notice is a course name update notice, generating a recommendation request after the course update of the data update request;

   When the data update notification is a user preference tag update notification, a recommendation request after the user preference tag update of the data update request is generated.
3. The course recommendation method according to claim 1, wherein said using preset course batch processing parameters to segment each course name in the to-be-updated course-name data set to obtain the to-be-updated course name data set Corresponding course name word segmentation results corresponding to each of the course names, the step of updating the course name word segmentation database according to the course name word segmentation results corresponding to each of the course names corresponding to the to-be-updated course name data set, include:

   obtaining the preset course batch processing parameters;

   When the to-be-updated course title data set includes course title data that needs to be deleted, delete the course title thesaurus according to the course title data to be deleted;

   When the to-be-updated course name data set includes course name data that needs to be added, the preset course batch processing parameters are used to segment each of the course names in the to-be-added course name data, respectively, Obtain the word segmentation result of the course name corresponding to each of the course names in the course name data that needs to be added, and segment the course name corresponding to each of the course names in the course name data to be added. The results are stored in the course name lexicon.
4. The method for recommending courses according to claim 3, wherein said using said preset batch processing parameters of courses to segment each of said course names in said data of course names to be added, respectively, to obtain said Before the steps of the word segmentation results of the course names corresponding to each of the course names in the added course name data, the steps include:

   Obtain the batch processing duration monitoring results and the preset course batch processing parameters;

   When the batch processing duration monitoring result is greater than the batch interval data of the preset course batch processing parameters, determine the batch processing parameters of the courses to be updated according to the batch processing duration monitoring results;

   The preset course batch processing parameters are updated according to the to-be-updated course batch processing parameters.
5. The course recommendation method according to claim 4, wherein the step of determining the batch processing parameters of the courses to be updated according to the monitoring result of the batch processing duration comprises:

   Get the number of CPU cores, block interval data, and preset ratio;

   Multiplying the batch processing duration monitoring result and the preset ratio to obtain the batch processing interval data corresponding to the batch processing parameters of the courses to be updated;

   Perform sliding window size calculation and sliding interval data calculation according to the batch interval data corresponding to the course batch processing parameters to be updated, to obtain the sliding window size and the sliding window size corresponding to the to-be-updated course batch processing parameters interval data.
6. The course recommendation method according to claim 1, wherein the step of using preset user batch processing parameters to update the user preference tag database according to the to-be-updated user preference tag data set comprises:

   Using the preset user batch processing parameters to obtain a piece of the user preference label data from the to-be-updated user preference label data set, to obtain the to-be-processed user preference label data;

   According to the user identification corresponding to the user preference label data to be processed, the preference label that needs to be added in the to-be-processed user preference label data is added to the user preference label library;

   Deleting, according to the user identification corresponding to the to-be-processed user-favorite label data, a favorite tag that needs to be deleted in the to-be-processed user-favorite tag data from the user favorite tag library;

   Repeat the steps of obtaining a piece of user preference label data from the user preference label data set to be updated by using the preset user batch processing parameters, and obtaining the user preference label data to be processed, until the to-be-updated user preference label data is completed. The user preference label database is updated by all the user preference label data in the user preference label data set.
7. The method for recommending courses according to claim 1, wherein the recommending a course according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request by using a preset recommendation rule, and obtaining the to-be-recommended course recommendation method. The steps of the respective target course recommendation results corresponding to each user ID of the recommended user ID set include:

   Searching the set of user IDs to be recommended carried in the course recommendation update request in the similarity matrix library, and obtaining the respective courses and preference tags corresponding to each of the user IDs in the set of user IDs to be recommended similarity matrix;

   Get the preset similarity threshold;

   Obtain one of the user IDs from the set of user IDs to be recommended as the user ID to be recommended;

   Find out the similarity between courses and preference tags greater than the preset similarity threshold from the similarity matrix of the courses and preference tags corresponding to the user IDs to be recommended, and compare all the courses found to be similar to the preference tags All course names corresponding to the degree are used as the target course recommendation result corresponding to the user ID to be recommended;

   Repeatedly performing the step of obtaining one of the user IDs from the set of user IDs to be recommended as the user ID to be recommended, until it is determined that each of the user IDs in the set of user IDs to be recommended corresponds to the respective user IDs. Target course recommendation results.
8. A course recommendation device, wherein the device includes:

   The request acquisition module is used to acquire data update requests;

   The course name word segmentation database update module is used to obtain the data set of the course name to be updated when the data update request is a recommendation request after the course update, and use the preset course batch processing parameters to respectively update the course name to be updated. Perform word segmentation for each course name in the data set, and obtain the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated. The word segmentation result of the said course name corresponding to each name updates the course name word segmentation database;

   The user preference tag library update module is used to obtain the user preference tag data set to be updated when the data update request is a recommendation request after the user preference tag is updated, and use preset user batch processing parameters according to the user preference tag to be updated. The preference tag dataset updates the user preference tag library;

   The target course name word segmentation word vector set determination module is used to obtain the course name word segmentation result from the course name word segmentation database based on the data update request, and obtain the target course name word segmentation result set, respectively for the target course name. Perform word vector generation on each of the course name word segmentation results in the word segmentation result set, and obtain a target course name word segmentation word vector set corresponding to the target course name word segmentation result set;

   The target user preference tag word vector set determination module is configured to obtain user preference tag data from the user preference tag library based on the data update request, and obtain a target user preference tag data set, and respectively analyze the target user preference tag data set in the target user preference tag data set. Perform word vector generation on each of the user preference tag data to obtain a target user preference tag word vector set corresponding to the target user preference tag data set;

   The similarity matrix generation module is used to generate the similarity matrix according to the target course name word segmentation word vector set and the target user preference tag word vector set, and obtain the similarity matrix of the course to be updated and the preference tag, according to the to-be-updated course and preference tag similarity matrix. Update the similarity matrix of courses and preference tags; update the similarity matrix library;

   A course recommendation module, configured to obtain a course recommendation update request based on the data update request, and use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain Each user identifier of the user identifier set to be recommended corresponds to a target course recommendation result.
9. A computer device includes a memory and a processor, wherein the memory stores a computer program, wherein the processor implements the following method steps when executing the computer program:

   Get data update request;

   When the data update request is a recommendation request after course update, obtain the course name data set to be updated, and use the preset course batch processing parameters to segment each course name in the to-be-updated course name data set, and obtain The word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated, according to the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated Update the course name thesaurus;

   When the data update request is a recommendation request after the user preference label is updated, the user preference label data set to be updated is obtained, and the user preference label database is updated according to the user preference label data set to be updated using preset user batch processing parameters ;

   Based on the data update request, the course name word segmentation result is obtained from the course name word segmentation database, and a target course name word segmentation result set is obtained, and each of the course name word segmentation results in the target course name word segmentation result set is performed Word vector generation, to obtain the target course name word segmentation word vector set corresponding to the target course name word segmentation result set;

   Obtain user preference tag data from the user preference tag library based on the data update request, obtain a target user preference tag data set, and perform word vector generation on each of the user preference tag data in the target user preference tag data set respectively. , obtain the target user preference tag word vector set corresponding to the target user preference tag data set;

   The similarity matrix is generated according to the target course name word segmentation word vector set and the target user preference label word vector set, and the similarity matrix between the courses to be updated and the preference labels is obtained. Matrix update similarity matrix library;

   Obtain a course recommendation update request based on the data update request, use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain the user to be recommended. Each user of the identification set identifies the corresponding target course recommendation result.
10. The computer device according to claim 9, wherein, the step of acquiring the data update request comprises:

    Get the data update notification sent by the Kafka message middleware;

    When the data update notification is a course name update notification, generating a recommendation request after the course update of the data update request;

    When the data update notification is a user preference tag update notification, a recommendation request after the user preference tag update of the data update request is generated.
11. The computer device according to claim 9, wherein the preset course batch processing parameters are used to segment each course name in the to-be-updated course-name data set, and obtain the corresponding to-be-updated course name data set. The step of updating the course name thesaurus database according to the course name word segmentation results corresponding to each of the course names corresponding to the to-be-updated course name datasets, including :

    obtaining the batch processing parameters of the preset course;

    When the to-be-updated course title data set includes course title data that needs to be deleted, delete the course title thesaurus according to the course title data to be deleted;

    When the to-be-updated course name data set includes course name data that needs to be added, the preset course batch processing parameters are used to segment each of the course names in the to-be-added course name data, respectively, Obtain the word segmentation result of the course name corresponding to each of the course names in the course name data that needs to be added, and segment the course name corresponding to each of the course names in the course name data to be added. The results are stored in the course name lexicon.
12. The computer device according to claim 11, wherein the preset course batch processing parameters are used to segment each of the course names in the course name data that needs to be added, to obtain the need to add new courses. Before the step of the word segmentation result of the course name corresponding to each of the course names in the course name data, the steps include:

    Obtain batch processing duration monitoring results and the preset course batch processing parameters;

    When the batch processing duration monitoring result is greater than the batch interval data of the preset course batch processing parameters, determine the batch processing parameters of the courses to be updated according to the batch processing duration monitoring results;

    The preset course batch processing parameters are updated according to the to-be-updated course batch processing parameters.
13. The computer device according to claim 12, wherein the step of determining the batch processing parameters of the courses to be updated according to the batch processing duration monitoring result comprises:

    Get the number of CPU cores, block interval data, and preset ratio;

    Multiplying the batch processing duration monitoring result and the preset ratio to obtain the batch processing interval data corresponding to the batch processing parameters of the courses to be updated;

    Perform sliding window size calculation and sliding interval data calculation according to the batch interval data corresponding to the course batch processing parameters to be updated, to obtain the sliding window size and the sliding window size corresponding to the to-be-updated course batch processing parameters interval data.
14. The computer device according to claim 9, wherein the step of using preset user batch processing parameters to update the user preference tag database according to the user preference tag data set to be updated comprises:

    Using the preset user batch processing parameters to obtain a piece of the user preference label data from the user preference label data set to be updated to obtain the user preference label data to be processed;

    According to the user identification corresponding to the user preference label data to be processed, the preference label that needs to be added in the to-be-processed user preference label data is added to the user preference label library;

    Deleting, according to the user identifier corresponding to the to-be-processed user-favorite label data, a favorite tag that needs to be deleted in the to-be-processed user-favorite tag data from the user favorite tag library;

    Repeat the steps of obtaining a piece of user preference label data from the user preference label data set to be updated by using the preset user batch processing parameters, and obtaining the user preference label data to be processed, until the to-be-updated user preference label data is completed. The user preference label database is updated by all the user preference label data in the user preference label data set.
15. A computer-readable storage medium on which a computer program is stored, wherein when the computer program is executed by a processor, the following method steps are implemented:

    Get data update request;

    When the data update request is a recommendation request after course update, obtain the course name data set to be updated, and use the preset course batch processing parameters to segment each course name in the to-be-updated course name data set, and obtain The word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated, according to the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated Update the course name thesaurus;

    When the data update request is a recommendation request after the user preference label is updated, the user preference label data set to be updated is obtained, and the user preference label database is updated according to the user preference label data set to be updated using preset user batch processing parameters ;

    Based on the data update request, the course name word segmentation result is obtained from the course name word segmentation database, and a target course name word segmentation result set is obtained, and each of the course name word segmentation results in the target course name word segmentation result set is performed Word vector generation, to obtain the target course name word segmentation word vector set corresponding to the target course name word segmentation result set;

    Obtain user preference tag data from the user preference tag library based on the data update request, obtain a target user preference tag data set, and perform word vector generation on each of the user preference tag data in the target user preference tag data set respectively. , obtain the target user preference tag word vector set corresponding to the target user preference tag data set;

    The similarity matrix is generated according to the target course name word segmentation word vector set and the target user preference tag word vector set to obtain the similarity matrix of the course to be updated and the preference label. According to the similarity between the course to be updated and the preference label Matrix update similarity matrix library;

    Obtain a course recommendation update request based on the data update request, use preset recommendation rules to perform course recommendation according to the similarity matrix library and the set of user identifiers to be recommended carried in the course recommendation update request, and obtain the user to be recommended. Each user of the identification set identifies the corresponding target course recommendation result.
16. The computer-readable storage medium of claim 15, wherein the step of obtaining a data update request comprises:

    Get the data update notification sent by the Kafka message middleware;

    When the data update notification is a course name update notification, generating a recommendation request after the course update of the data update request;

    When the data update notification is a user preference tag update notification, a recommendation request after the user preference tag update of the data update request is generated.
17. The computer-readable storage medium according to claim 15, wherein, by using preset course batch processing parameters, word segmentation is performed on each course name in the to-be-updated course-name data set to obtain the to-be-updated course name The word segmentation result of the course name corresponding to each of the course names corresponding to the data set, according to the word segmentation result of the course name corresponding to each of the course names corresponding to the course name data set to be updated, update the course name word segmentation result. steps, including:

    obtaining the batch processing parameters of the preset course;

    When the to-be-updated course title data set includes course title data that needs to be deleted, delete the course title thesaurus according to the course title data to be deleted;

    When the to-be-updated course name data set includes course name data that needs to be added, the preset course batch processing parameters are used to segment each of the course names in the to-be-added course name data, respectively, Obtain the word segmentation result of the course name corresponding to each of the course names in the course name data that needs to be added, and segment the course name corresponding to each of the course names in the course name data to be added. The results are stored in the course name lexicon.
18. The computer-readable storage medium according to claim 17, wherein the pre-set course batch processing parameters are used to segment each of the course names in the course name data to be added, respectively, to obtain the Before the steps of the word segmentation results of the course names corresponding to each of the course names in the newly added course name data, the steps include:

    Obtain batch processing duration monitoring results and the preset course batch processing parameters;

    When the batch processing duration monitoring result is greater than the batch interval data of the preset course batch processing parameters, determine the batch processing parameters of the courses to be updated according to the batch processing duration monitoring results;

    The preset course batch processing parameters are updated according to the to-be-updated course batch processing parameters.
19. The computer-readable storage medium according to claim 18, wherein the step of determining the batch processing parameters of the courses to be updated according to the batch processing duration monitoring result comprises:

    Get the number of CPU cores, block interval data, and preset ratio;

    Multiplying the batch processing duration monitoring result and the preset ratio to obtain the batch processing interval data corresponding to the batch processing parameters of the courses to be updated;

    Perform sliding window size calculation and sliding interval data calculation according to the batch interval data corresponding to the course batch processing parameters to be updated, to obtain the sliding window size and the sliding window size corresponding to the to-be-updated course batch processing parameters interval data.
20. The computer-readable storage medium according to claim 15, wherein the step of updating the user preference tag library according to the user preference tag data set to be updated by using preset user batch processing parameters comprises:

    Using the preset user batch processing parameters to obtain a piece of the user preference label data from the user preference label data set to be updated to obtain the user preference label data to be processed;

    According to the user identification corresponding to the user preference label data to be processed, the preference label that needs to be added in the to-be-processed user preference label data is added to the user preference label library;

    Deleting, according to the user identifier corresponding to the to-be-processed user-favorite label data, a favorite tag that needs to be deleted in the to-be-processed user-favorite tag data from the user favorite tag library;

    Repeat the steps of obtaining a piece of the user preference label data from the user preference label data set to be updated by using the preset user batch processing parameters, and obtaining the user preference label data to be processed, until the to-be-updated user preference label data is completed. The user preference label database is updated by all the user preference label data in the user preference label data set.

Images