Personalized training resource recommendation method based on knowledge graph and ability evaluation

Abstract

The invention discloses a personalized training resource recommendation method based on a knowledge graph and capability evaluation, and relates to the field of deep learning. The method comprises the following steps: constructing a knowledge graph according to online learning platform training resources; generating knowledge graph nodes from the user school portrait, and matching online learning platform training resource nodes in a knowledge graph; and in combination with the matched nodes, selecting online learning platform training resource nodes higher than a preset threshold as recommendation objects. According to the method, intelligent distribution of the training resources is carried out based on the knowledge graph, classification and clustering system of the training resources, the hobbies and interests of the users, the portraits and intentions of the users and other information in combination with multiple application scenes, the resource utilization efficiency in each application scene can be greatly improved, and meanwhile, the user experience is improved.

Description

Technical Field

[0001] The present invention relates to the field of data mining, and more specifically to a personalized training resource recommendation method based on knowledge graph and ability assessment. Background Art

[0002] After 8 years of construction, the online learning platform has become the first truly all-staff application information system. It has been fully improved in the fields of functions, resources, applications and management systems, and has initially realized the informatization, process and mobility of training business, the systematization of courses and data standardization. In view of the new situation, new demands, new opportunities and new challenges faced by the digital transformation of education and training, it is necessary to further consolidate the foundation for the deep integration of talent business development, training management innovation, student service improvement and digital transformation.

[0003] With the development of science and technology in recent years, intelligent content distribution can be achieved in many industries by using a series of technical means with artificial intelligence technology as the core. At the same time, the development of hardware technology also provides an efficient computing platform for more complex artificial intelligence technology, which also effectively guarantees the practicality of this project.

[0004] The online learning platform has a large amount of training resources to serve all employees of the company. How to continuously make training resources better meet user needs has always been the goal pursued by online universities. Summary of the invention

[0005] In view of this, the present invention provides a personalized training resource recommendation method based on knowledge graph and capability assessment to achieve the goals in the background technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A personalized training resource recommendation method based on knowledge graph and ability assessment includes the following steps:

[0008] Build a knowledge graph based on the training resources of the online learning platform;

[0009] Generate knowledge graph nodes from user classroom portraits and match online learning platform training resource nodes in the knowledge graph;

[0010] Combined with the matched nodes, online learning platform training resource nodes above the preset threshold are selected as recommendation objects.

[0011] Optionally, constructing the knowledge graph includes classifying and clustering online learning platform training resources, clustering the classified online learning platform training resources, and constructing nodes of the knowledge graph.

[0012] Optionally, the user school portrait construction process is: combining the user's interests, hobbies, and intentions, sorting out various application scenarios of training resources, establishing a user data platform, analyzing user behavior, intelligently labeling users and stratifying populations, and establishing a user school.

[0013] Optionally, it also includes storing the knowledge graph nodes generated by the user classroom portrait, and clustering the user classroom portrait nodes with a similarity higher than 90%.

[0014] Optionally, the similarity is judged based on user behavior, user preference, user identity, user social relationship, content relationship, geographic location, and time dimension; it is used to increase the effective exposure nodes of content in various locations within the platform, attract user aggregation, and become a gathering point for various information.

[0015] Optionally, in combination with the matched nodes, online learning platform training resource nodes above a preset threshold are selected as recommendation objects, specifically including the following steps:

[0016] Based on the matched nodes, use the breadth-first algorithm to traverse the shortest path between any two nodes with similar matching degrees and store the shortest path;

[0017] After multiple traversals, the nodes with the most occurrences are screened, and the online learning platform training resource nodes corresponding to the nodes with the most occurrences are used as the recommended objects.

[0018] It can be seen from the above technical solution that compared with the prior art, the present invention discloses a personalized training resource recommendation method based on knowledge graph and capability assessment. Based on the knowledge graph, classification, clustering system of training resources and information such as user interests, user portraits and intentions, combined with multiple application scenarios, intelligent distribution of training resources is carried out, which can greatly improve the resource utilization efficiency in each application scenario and improve the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0020] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] The embodiment of the present invention discloses a personalized training resource recommendation method based on knowledge graph and ability assessment, such as Figure 1 As shown, the following steps are included:

[0023] Build a knowledge graph based on the training resources of the online learning platform;

[0024] Generate knowledge graph nodes from user classroom portraits and match online learning platform training resource nodes in the knowledge graph;

[0025] Combined with the matched nodes, select the online learning platform training resource nodes that are higher than the preset threshold as the recommended objects. Specifically, based on the matched nodes, use the breadth-first algorithm to traverse the shortest path between any two nodes with similar matching degrees, and store the shortest path;

[0026] After multiple traversals, the nodes with the most occurrences are screened, and the online learning platform training resource nodes corresponding to the nodes with the most occurrences are used as the recommended objects.

[0027] In this embodiment, it includes:

[0028] In the first phase, the distribution of content on the online learning platform will be enhanced from two levels: product design and algorithm-related distribution technology, to improve how content products can distribute content more efficiently.

[0029] By designing content nodes, we can improve the effective exposure nodes of content in various places on the platform, attract users to aggregate and become the gathering point of various information. Users gather together because of the nodes and can get more information through the nodes. Content distribution is based on user behavior, user preferences, user identity, user social relationships, content relationships, geographic location, and time dimensions.

[0030] Use recommendation-related machine learning algorithms to help users discover resources of interest from various learning resources.

[0031] The second stage is based on personalized recommendation algorithms, mainly based on collaborative filtering and content recommendation. The content recommendation method is to first characterize the news, and then use the user's positive feedback (such as clicks, reading time, sharing, collection, comments, etc.) and negative feedback (such as lack of interest, etc.) to establish a connection between users and news tags, so as to conduct statistical modeling.

[0032] The third stage is based on large-scale real-time machine learning algorithms. The architecture is divided into two layers: the retrieval layer, which has multiple retrieval branches and pulls out content candidates that users are interested in; the scoring layer, which uses real-time learning to model and score based on three categories of features: user features, content features, and environmental features. It is worth mentioning that the actual sorting is not completely based on the model score. Some specific business logic will be combined to make the final sorting and spit it out to the user.

[0033] Among them, users gather together because of nodes and can obtain more information through nodes. Content distribution is based on user behavior, user preferences, user identity, user social relationships, content relationships, geographic location, and time dimensions. This increases the effective exposure nodes of content in various places on the platform, attracting users to aggregate and become a gathering point for various information.

[0034] Establish user portraits for online learning platforms. Analyze the knowledge graph, classification, clustering system of online learning platform training resources, as well as information such as user interests, hobbies, and intentions, sort out various application scenarios of training resources, establish a user data platform, analyze user behavior, intelligently label users and group stratification, form effective customer insights, and establish user portraits.

[0035] Design personalized course recommendation distribution and sorting solutions. Course recommendations are divided into recommended course modules and all course modules. Course sorting will be personalized based on student user portraits.

[0036] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0037] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A personalized training resource recommendation method based on knowledge graph and ability assessment, characterized in that: The following steps are involved: Build a knowledge graph based on the training resources of the online learning platform; Generate knowledge graph nodes from user classroom portraits and match online learning platform training resource nodes in the knowledge graph; Combined with the matched nodes, online learning platform training resource nodes above the preset threshold are selected as recommendation objects.

2. According to claim 1, a personalized training resource recommendation method based on knowledge graph and ability assessment is characterized in that: Constructing the knowledge graph includes classifying and clustering online learning platform training resources. The classified online learning platform training resources are clustered to construct nodes of the knowledge graph.

3. According to claim 1, a personalized training resource recommendation method based on knowledge graph and ability assessment is characterized in that: The process of constructing the user school portrait is as follows: combining the user's interests, hobbies and intentions, sorting out various application scenarios of training resources, establishing a user data platform, analyzing the user's behavior, intelligently labeling users and stratifying the population, and establishing a user school.

4. According to claim 1, a personalized training resource recommendation method based on knowledge graph and ability assessment is characterized in that: It also includes storing the knowledge graph nodes generated by the user classroom portraits, and clustering the user classroom portrait nodes with a similarity higher than 90%.

5. According to claim 4, a personalized training resource recommendation method based on knowledge graph and ability assessment is characterized in that: The similarity is judged based on user behavior, user preference, user identity, user social relationship, content relationship, geographic location, and time dimension; It is used to increase the effective exposure nodes of content in various locations within the platform, attracting users to aggregate and becoming a gathering point for various information.

6. According to claim 1, a personalized training resource recommendation method based on knowledge graph and ability assessment is characterized in that: Combined with the matched nodes, online learning platform training resource nodes above the preset threshold are selected as recommendation objects, which specifically include the following steps: Based on the matched nodes, use the breadth-first algorithm to traverse the shortest path between any two nodes with similar matching degrees and store the shortest path; After multiple traversals, the nodes with the most occurrences are screened, and the online learning platform training resource nodes corresponding to the nodes with the most occurrences are used as the recommended objects.

Images