A talent supply chain platform information interaction method based on virtual reality

By building VR classification training classrooms, certification spaces and virtual recruitment spaces on the virtual reality technology platform, combining matching algorithms and machine learning, the problem that traditional platforms cannot meet the needs of enterprises in a timely manner is solved, and a closed loop of talent training and job matching is realized. Improved learning and recruitment experience.

CN118536956BActive Publication Date: 2024-12-03GUANGZHOU ZHIXIN CLOUD TECH CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202410466858.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-04-18
Publication Date
2024-12-03
Estimated Expiration
2044-04-18

AI Technical Summary

Technical Problem

Traditional schools, education and training institutions and talent service institutions are unable to meet the job needs of enterprises in a timely and accurate manner, resulting in low matching between talent training and jobs, poor user experience, limited offline trials, and untimely information exchange, which leads to difficulty in recruiting employees and students in enterprises. It's difficult to find a job.

Method used

Using virtual reality technology to build a talent supply chain platform, including VR classification training classrooms, certification spaces, virtual recruitment spaces and job display areas, match courses, positions and certificates through matching algorithms, provide real learning and recruitment scenarios, and use Unity to It interacts with technologies such as SteamVR, and combines machine learning algorithms for job screening and recommendation.

Benefits of technology

It effectively solved the problem of disconnection between talent training and job needs, opened up barriers between education and practice, optimized the connection between educational resources, improved learning effect, and cultivated compound talents with theoretical and practical abilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118536956B_ABST
    Figure CN118536956B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of information matching technology, constructs VR classification training classrooms based on virtual reality technology, constructs certification space and virtual recruitment space based on virtual reality technology, realizes VR job display area based on virtual reality technology, constructs matching algorithm, realizes job matching courses, job matching certificates and course matching certificates, matches and recommends courses, jobs and certificates according to matching algorithm; the present invention effectively solves the disconnection between talent training and job requirements, breaks through the barriers between education and practice, builds a complete talent training and service closed loop, ensures that teaching content is connected to job skills, and professional certificates are connected to employment ability. With the help of virtual reality technology, a realistic learning environment is provided for students. According to the present invention, by matching courses, certificates and jobs, and also feedback whether the matching is qualified based on the students' grades, this method is more conducive to cultivating compound talents with both theoretical and practical abilities.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of information matching, and in particular to an information interaction method for a talent supply chain platform based on virtual reality. Background Art

[0002] At present, traditional schools or educational institutions, talent service agencies, and certification agencies are all relatively independent platforms. "Courses, positions, and certificates" cannot be integrated in a timely manner. The courses offered by schools or educational institutions cannot be accurately matched with the job needs of enterprises. The knowledge, skills, and literacy requirements of enterprise positions obtained by talent service agencies cannot be timely matched with schools or educational institutions. Most certification agencies aim to obtain certificates for students, and cannot accurately match the employment needs of enterprises with the talent training programs of schools or educational institutions. In addition, traditional platforms are implemented through web2.0, which cannot truly simulate the scenarios of job recruitment and on-site learning, and the user experience is poor, resulting in most learning, recruitment, and certification businesses still being mainly offline. Offline interviews are in a relatively limited time and space range, and information is exchanged in a timely manner, but resources cannot be shared in a timely manner, resulting in a low degree of matching between people and positions. In addition, when people and positions do not match, employers and applicants can only choose to give up, and cannot form a solution in an effective time, which ultimately leads to difficulties for companies to recruit people and students to find jobs. Therefore, a talent supply chain platform information interaction method based on virtual reality is urgently needed to solve the above problems. Summary of the invention

[0003] In view of the limitations of the existing methods mentioned above, the purpose of the present invention is to propose a talent supply chain platform information interaction method based on virtual reality to solve one or more technical problems existing in the prior art and at least provide a beneficial choice or create conditions.

[0004] A talent supply chain platform information interaction method based on virtual reality, the method comprising the following steps:

[0005] S100: Building a VR classification training classroom based on virtual reality technology;

[0006] S200: Construct verification space and virtual recruitment space based on virtual reality technology;

[0007] S300: Realize VR job display area based on virtual reality technology;

[0008] S400: Build a matching algorithm and implement job matching courses, job matching certificates and course matching certificates;

[0009] S500: Match and recommend courses, positions and certificates based on matching algorithms.

[0010] Furthermore, in step S100, a 3D scene is created in Unity, and a SteamVR plug-in is imported, a virtual classroom is created in the scene, and a script is added to the virtual classroom to realize the interactive behavior of the control elements, wherein the interactive behavior includes: triggering corresponding actions through collision or induction detection to increase the authenticity of the virtual classroom;

[0011] By adding objects in the classroom or displaying relevant information on the projector, users can learn knowledge of different categories and realize the VR classification training function. The display of relevant information on the projector is represented by: accessing the back-end system, obtaining the content of classification training through network requests, and projecting the content of classification training; importing the teacher model, and realizing the interaction between students and teachers.

[0012] Furthermore, in the step S200, after training in the VR classification training classroom, a verification space and a virtual recruitment space are constructed to test the user's ability;

[0013] Constructing the examination space: Using the 3D model in Unity to create a virtual examination space in the scene, the virtual examination space includes: simulated examination scene, examination paper selection, examination paper answering; and can realize free roaming in the virtual examination scene, view the examination environment, examination equipment and examination-related information, control the perspective and action of the examinee through the keyboard or mouse, realize the interactive operation of the scene, use the extracted features and label data to train the classifier, adjust the model parameters, and improve the accuracy and generalization ability of the model;

[0014] Constructing a virtual recruitment space: creating a virtual recruitment space in the scene, the virtual recruitment space including: a job display area, an applicant registration area, and an interview room; by displaying job information and applicant information on the screen in the virtual recruitment space, allowing users to go through the recruitment process and access the back-end system, users obtain job and applicant information through network requests and display them in the virtual recruitment space; the virtual recruitment space can screen jobs and recommend candidates by applying machine learning algorithms.

[0015] Furthermore, in step S300, virtual reality technology creates a virtual job display area in the scene, sets a server interface, obtains and displays the job information in the background through network requests, and obtains and dynamically updates the job information by reading a local json file.

[0016] Furthermore, in step S400, VR classified training classrooms, certification spaces, virtual recruitment spaces and VR job display areas are constructed through virtual reality technology, and real-time user information, job information and certificate information of the four areas are obtained, background job information is obtained through the virtual job display area, job competency units are created based on the background job information, course knowledge points are obtained through the VR classified training classrooms, and certificate information is obtained through the certification space and virtual recruitment space;

[0017] Data preprocessing is performed on the job competency unit, course knowledge points, and certificate information. Word segmentation is performed according to keywords, topics, and categories, and the word segmentation is represented by word vectors. The word vectors are classified according to the corresponding data sources into job competency unit word vector sets, course knowledge point word vector sets, and certificate information word vector sets. A model is trained using Doc2Vec, where Doc2Vec is a model based on Word2Vec. The Doc2Vec model is used to calculate the cosine similarity A_cos corresponding to each element in the job competency unit word vector set and the course knowledge point word vector set, the cosine similarity B_cos corresponding to each element in the job competency unit word vector set and the certificate information word vector set, and the cosine similarity B_cos corresponding to each element in the course knowledge point word vector set and the certificate information word vector set. The cosine similarity C_cos corresponding to each element in the book information word vector set is used to construct sets [A_cos], [B_cos], and [C_cos] for A_cos, B_cos, and C_cos respectively. For the set [A_cos], the m course knowledge points with the highest similarity are selected. For each job capability unit, the m course knowledge points with the highest cosine similarity are selected, and the matching results are output. The job capability units and the corresponding course knowledge points are output as a dictionary; for the set [B_cos], the m skill certificates with the highest similarity are selected and output as a dictionary; for [C_cos], the m certificates with the highest similarity are selected and output as a dictionary, and the user's positions, courses, and certificates are matched in combination with the dictionary.

[0018] Furthermore, the user matches the courses by cosine similarity, and performs secondary matching on the job competency units, course knowledge points, and certificate information using the matching scores obtained in the certification space and the virtual recruitment space;

[0019] Read the sets [A_cos], [B_cos], [C_cos], and reorder all elements in the sets [A_cos], [B_cos], and [C_cos] according to the cosine similarity. The minimum value of the value in the set [B_cos] is A1, and the current maximum value of the value in the set [B_cos] is A2. B1=A1 / mean(B_cos), B2=A2 / mean(B_cos), mean(B_cos) represents the average value of all elements in the set [B_cos]. B1 and B2 represent the maximum and minimum ratios between certificates and positions.

[0020] Define Li=B1+(B2-B1)×(C_cosi-min(C_cos)) / (max(C_cos)-min(C_cos)), where C_cosi is the ith element in the set [C_cos], min(C_cos) is the value of the smallest element in the set [C_cos], and max(C_cos) is the value of the largest element in the set [C_cos]. The Li is expressed as judging the corresponding value of the course and the certificate by the ratio of the relationship between the position and the certificate, and constructing the array E1 with Li;

[0021] Define Ki=B1+(B2-B1)×(B_cosi-min(A_cos)) / (max(A_cos)-min(A_cos)), where A_cosi is the ith element in the set [A_cos], min(A_cos) is the value of the smallest element in the set [A_cos], and max(A_cos) is the value of the largest element in the set [A_cos]. The Ki is expressed as judging the corresponding value of the course and the position by the ratio of the relationship between the position and the certificate, and constructing the array E2 with Ki;

[0022] Construct feedback matrix E1 from arrays E1 and E2 E2, the method of performing secondary matching of the position competency units, course knowledge points, and certificate information by combining the feedback matrix with the matching scores is as follows:

[0023] S401: Obtain the grades of the matched certification space and virtual recruitment space, and define the certification space grade as E_grade, and the recruitment space grade as I_grade. Calculate the average values ​​of the user's overall certification space grade and recruitment space grade mean(E_grade) and mean(I_grade), respectively, and determine the overall pass rate of the user through calculation.

[0024] PE= + ;

[0025] PI= + ;

[0026] Wherein, λ is the passing ratio, which is 70%, max() is the maximum value function, min() is the minimum value function, M is the total number of users, and the formula is expressed as adding points to the passing line base according to the user's score ratio to determine whether the user's score reaches the passing line. If both of the user's scores reach the passing line, there is no need to perform secondary matching. If the user's score E_grade<PE, jump to S402; if the user's score I_grade<PI, jump to S403;

[0027] S402: cancel the cosine similarity matched this time from the m cosine similarities selected from the set [C_cos], re-output the remaining m-1 cosine similarities into the dictionary, and jump to S404;

[0028] S403: cancel the cosine similarity matched this time from the m cosine similarities selected from the set [A_cos], output the remaining m-1 cosine similarities to the dictionary again, and jump to S404;

[0029] S404: Create a set D and a set P, wherein the set D includes the updated m-1 sets [C_cos], and the set P includes the updated m-1 sets [A_cos], and input the set D and the set P into the feedback matrix E1 E2 performs secondary matching judgment and updates the matrix , the feedback matrix E1 in set D and set E Set E2 to infinity to start traversal, repeatedly traverse all the cosine similarities in set D and set P and delete one of the m, obtain the corresponding word vector, calculate the shortest distance between all word vectors in set D and the word vector corresponding to the deleted cosine similarity, traverse set P, and determine whether the source data is the same as that in set P. If the same, output the matrix , if different, update the matrix .

[0030] Furthermore, in step S500, the job competency units, course knowledge points, and certificate information are obtained through a matching algorithm and automatically matched according to the user's situation, and the matching algorithm is adjusted according to the user's performance feedback after use to obtain a more reasonable matching algorithm.

[0031] The beneficial effects of the present invention are: effectively solving the disconnect between talent training and job requirements, breaking through the barriers between education and practice, and building a complete closed loop of talent training and services. Optimize educational resources to ensure that teaching content is connected to job skills and professional certificates are connected to employment capabilities. With the help of virtual reality technology, a realistic learning environment is provided for students, and abstract knowledge is visualized, thereby greatly improving learning effects and igniting students' passion for learning. According to the present invention, by matching courses, certificates, and positions, and also feedback on whether the matching is qualified based on students' grades, this method is more conducive to cultivating compound talents with both theoretical and practical abilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The above and other features of the present invention will become more obvious by describing in detail the embodiments shown in the accompanying drawings. The same reference numerals in the accompanying drawings of the present invention represent the same or similar elements. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other accompanying drawings can be obtained based on these accompanying drawings without creative work. In the accompanying drawings:

[0033] Figure 1 Shown is a flow chart of a talent supply chain platform information interaction method based on virtual reality. DETAILED DESCRIPTION

[0034] The following will be combined with the embodiments and drawings to clearly and completely describe the concept, specific structure and technical effects of the present invention, so as to fully understand the purpose, scheme and effect of the present invention. It should be noted that the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0035] like Figure 1 As shown, 1. To realize the classification training classroom through virtual reality, it is necessary to use the Unity engine for development. The following are the implementation steps and methods:

[0036] 1.1 Create a scene: Create a 3D scene in Unity and import the SteamVR plug-in.

[0037] 1.2 Create a new scene and add the components required for the VR environment, such as VR Camera, VRController, etc.

[0038] 1.3 Create a virtual classroom in the scene, including elements such as a podium, tables and chairs, and a projector. You can use Unity's built-in 3D models or make your own.

[0039] 1.4 Add scripts to control the interactive behavior of elements. For example, when a user enters a classroom, collision detection can trigger corresponding actions, such as opening the door, adjusting the lights, etc.

[0040] 1.5 Implement the classification training function. By adding objects in the classroom or displaying relevant information on the projector, users can learn knowledge of different categories. This can be achieved using Unity's own UI system or third-party plug-ins.

[0041] 1.6 Access the backend system. Obtain the content of the classification training, such as pictures, text, audio, etc., through the network request, and then display it in the virtual classroom.

[0042] 1.7 Import the teacher model: Import the teacher model in Unity. You can use the model resources built into Unity or the model provided by a third-party plug-in. The model can be placed in the appropriate position in the scene so that students can see the teacher.

[0043] 1.8 Implement the teacher model's movements and animations: To make the teacher model more realistic, its movements and animations need to be implemented. You can use Unity's built-in animation system or third-party plug-ins to create animations. For example, when the teacher speaks, you can use Unity's audio system to play the teacher's voice and synchronize the teacher model's mouth movements with the voice.

[0044] 1.9 Implementing interaction between students and teachers: In order to enable students to interact with teachers, some interactive functions need to be implemented. For example, students can click on the teacher model in the scene to ask or answer questions to the teacher. You can use Unity's input system to implement interactive functions.

[0045] 1.10 Implementing a virtual blackboard and projector: To facilitate teachers' teaching and students' listening, you can create a virtual blackboard and projector in the scene. You can use Unity's UI system to create these elements and add the required text, images, and video content to them.

[0046] 1.11 Add background music and sound effects: To enhance the atmosphere of the scene, you can add appropriate background music and sound effects. You can use Unity's audio system to play music and sound effects.

[0047] 1.12 Implementing network functions: In order to allow multiple students to participate in the virtual classroom at different locations, network functions need to be implemented. You can use Unity's network system to implement multi-person online interaction functions.

[0048] 2. To realize the verification space through virtual reality, you need to use the Unity engine for development. The following are the steps and methods:

[0049] 2.1 Create a Unity project and import the SteamVR plugin.

[0050] 2.2 Create a new scene and add the components required for the VR environment, such as VR Camera, VR Controller, etc.

[0051] 2.3 Use Unity's built-in 3D model to create a virtual examination space in the scene, including simulated examination scenes, test paper selection, answering and other elements.

[0052] 2.4 Use Unity's built-in 3D model to roam freely in the virtual examination scene, view the examination environment, examination equipment and examination-related information. Use the keyboard or mouse to control the examinee's perspective and actions to achieve interactive operation of the scene.

[0053] 2.5 Use Unity's built-in UI system to interact with objects in the virtual examination scene, such as opening or closing examination equipment, flipping through examination papers, etc. Use devices such as a mouse or a controller to perform operations, and implement interactive operations on objects based on feedback from the results of the operations.

[0054] 2.6 Use Unity's built-in UI system to answer questions in the virtual exam scene, including multiple-choice questions, fill-in-the-blank questions, true-or-false questions, etc. Use the mouse or keyboard to select or enter answers, and the system will automatically judge the answers and give corresponding feedback information.

[0055] 2.7 Use Unity's built-in UI system to interact with the interface in the virtual exam scene, such as viewing exam rules, checking scores, etc. Use devices such as a mouse or touch screen to interact with the interface.

[0056] 2.8 Use Unity's built-in UI system to hear relevant sound prompts in the virtual exam scene, such as the voice prompts for the start and end of the exam, and voice feedback for correct or incorrect answers. The system will automatically play the corresponding sound prompts as needed to achieve interactive sound operations.

[0057] 2.9 Collect a large amount of question and answer data, including correct answers and incorrect answers. The data needs to be preprocessed and cleaned to ensure accuracy and consistency.

[0058] 2.10 Extract useful features from the question and answer data, including text features, image features, numerical features, etc.

[0059] 2.11 Select appropriate machine learning algorithms as classifiers based on the type of problem and data characteristics, including decision trees, support vector machines, naive Bayes, neural networks, etc.

[0060] 2.12 Use the extracted features and labeled data to train the classifier and adjust the model parameters to improve the accuracy and generalization ability of the model.

[0061] 2.13 Use the test data set to evaluate the performance of the model. Common evaluation indicators include accuracy, precision, recall, F1 value, etc.

[0062] 2.14 Apply the trained model to actual scenarios, classify and judge the answers to new questions, judge the candidates' answers, and give scores.

[0063] 2.15 Testing and optimization. Use VR headsets and controllers for testing, and promptly fix and optimize any problems found.

[0064] 3. To realize the virtual recruitment space through virtual reality, you need to use the Unity engine for development. The following are the steps and methods:

[0065] 3.1 Create a Unity project and import the SteamVR plugin.

[0066] 3.2 Create a new scene and add the components required for the VR environment, such as VR Camera, VR Controller, etc.

[0067] 3.3 Create a virtual recruitment space in the scene, including elements such as job display area, applicant registration area, interview room, etc. You can use Unity's built-in 3D models or make your own.

[0068] 3.4 Add scripts to control the interactive behavior of elements. For example, when a user enters a virtual recruitment space, collision detection can trigger corresponding actions, such as opening doors, adjusting lights, etc.

[0069] 3.5 Implement the recruitment function. Allow users to go through the recruitment process by adding objects to the virtual recruitment space or displaying job information and applicant information on the screen. This can be achieved using Unity's built-in UI system or third-party plug-ins.

[0070] 3.6 Access the back-end system. Obtain information about positions and applicants through the Internet, and then display them in the virtual recruitment space.

[0071] 3.7 Apply machine learning algorithms to screen and recommend candidates.

[0072] 3.8 Testing and optimization. Use VR headsets and controllers for testing, and promptly fix and optimize any problems found.

[0073] 4. To realize a VR space for job display area with updateable content, you need to use Unity engine for development. The following are the steps and methods to achieve this:

[0074] 4.1 Create a Unity project and import the SteamVR plugin.

[0075] 4.2 Create a new scene and add the components required for the VR environment, such as VR Camera, VRController, etc.

[0076] 4.3 Use Unity's built-in 3D models to create a virtual job display area in the scene.

[0077] 4.4 Add scripts to control the interactive behavior of elements. For example, when a user enters a job display area, collision detection can trigger corresponding actions, such as opening doors, adjusting lights, etc.

[0078] 4.5 Update job information. This can be done using Unity's own UI system or third-party plug-ins, or by remotely requesting it through the network.

[0079] 4.6 Testing and optimization. Use VR headsets and controllers for testing, and promptly fix and optimize any problems found.

[0080] 5. To realize a VR space that can publish job positions of multiple companies in the background, you need to use the Unity engine for development and combine the background development technology to realize dynamic data update. The following are the steps and methods to achieve this:

[0081] 5.1 Create a Unity project and import the SteamVR plugin.

[0082] 5.2 Create a new scene and add the components required for the VR environment, such as VR Camera, VRController, etc.

[0083] 5.3 You can use Unity's built-in 3D models to create a virtual job display area in the scene.

[0084] 5.4 Add scripts to control the interactive behavior of elements. For example, when a user enters a job display area, collision detection can trigger corresponding actions, such as opening doors, adjusting lights, etc.

[0085] 5.5 Implement job information update. You can design a server interface to obtain job information from the background through network requests and display it. You can also directly read local json files in the code to obtain job information and refresh it.

[0086] 5.6 Backend development technology realizes dynamic data update. Use Java backend programming language and database technology to realize the addition, deletion, modification, query and dynamic update of job information.

[0087] 5.7 Use the NavMesh system in Unity to implement mobile control and provide users with an immersive scene interaction experience.

[0088] 5.8 Add spatial UI elements, such as buttons and input boxes, in the job display area to facilitate user operations and information input.

[0089] 5.9 Get backend job information through network request. You can use Unity's own WWW class or third-party plug-ins such as UnityWebRequest, etc., and use the GET or POST method to get data from the server. Then parse the data and display it on the VR space UI element.

[0090] 5.10 Write interfaces in the background to ensure the security and integrity of information. Add functions such as data upload (add or update job information), data deletion, and data query in the job display area to achieve dynamic update of job information.

[0091] 5.11 Testing and optimization. Use VR headsets and controllers for testing, and promptly fix and optimize any problems found.

[0092] Use virtual reality technology to build VR classified training classrooms, certification spaces, virtual recruitment spaces and VR job display areas, and obtain real-time user information, job information and certificate information in the four areas, obtain background job information through the virtual job display area, create job competency units based on the background job information, obtain course knowledge points through VR classified training classrooms, and obtain certificate information through certification spaces and virtual recruitment spaces;

[0093] Data preprocessing is performed on the job competency unit, course knowledge points, and certificate information. Word segmentation is performed according to keywords, topics, and categories, and the word segmentation is represented by word vectors. The word vectors are classified according to the corresponding data sources into job competency unit word vector sets, course knowledge point word vector sets, and certificate information word vector sets. A model is trained using Doc2Vec, where Doc2Vec is a model based on Word2Vec. The Doc2Vec model is used to calculate the cosine similarity A_cos corresponding to each element in the job competency unit word vector set and the course knowledge point word vector set, the cosine similarity B_cos corresponding to each element in the job competency unit word vector set and the certificate information word vector set, and the cosine similarity B_cos corresponding to each element in the course knowledge point word vector set and the certificate information word vector set. The cosine similarity C_cos corresponding to each element in the book information word vector set is used to construct sets [A_cos], [B_cos], and [C_cos] for A_cos, B_cos, and C_cos respectively. For the set [A_cos], the m course knowledge points with the highest similarity are selected. For each job capability unit, the m course knowledge points with the highest cosine similarity are selected, and the matching results are output. The job capability units and the corresponding course knowledge points are output as a dictionary; for the set [B_cos], the m skill certificates with the highest similarity are selected and output as a dictionary; for [C_cos], the m certificates with the highest similarity are selected and output as a dictionary, and the user's positions, courses, and certificates are matched in combination with the dictionary.

[0094] Users match courses by cosine similarity, and conduct secondary matching of the job competency units, course knowledge points, and certificate information in the certification space and virtual recruitment space with the matching scores;

[0095] Read the sets [A_cos], [B_cos], [C_cos], and reorder all elements in the sets [A_cos], [B_cos], and [C_cos] according to the cosine similarity. The minimum value of the value in the set [B_cos] is A1, and the current maximum value of the value in the set [B_cos] is A2. B1=A1 / mean(B_cos), B2=A2 / mean(B_cos), mean(B_cos) represents the average value of all elements in the set [B_cos]. B1 and B2 represent the maximum and minimum ratios between certificates and positions.

[0096] Define Li=B1+(B2-B1)×(C_cosi-min(C_cos)) / (max(C_cos)-min(C_cos)), where C_cosi is the ith element in the set [C_cos], min(C_cos) is the value of the smallest element in the set [C_cos], and max(C_cos) is the value of the largest element in the set [C_cos]. The Li is expressed as judging the corresponding value of the course and the certificate by the ratio of the relationship between the position and the certificate, and constructing the array E1 with Li;

[0097] Define Ki=B1+(B2-B1)×(B_cosi-min(A_cos)) / (max(A_cos)-min(A_cos)), where A_cosi is the ith element in the set [A_cos], min(A_cos) is the value of the smallest element in the set [A_cos], and max(A_cos) is the value of the largest element in the set [A_cos]. The Ki is expressed as judging the corresponding value of the course and the position by the ratio of the relationship between the position and the certificate, and constructing the array E2 with Ki;

[0098] Construct feedback matrix E1 from arrays E1 and E2 E2, the method of performing secondary matching of the position competency units, course knowledge points, and certificate information by combining the feedback matrix with the matching scores is as follows:

[0099] (1) Obtain the scores of the matched certification space and virtual recruitment space, and define the certification space score as E_grade and the recruitment space score as I_grade. Calculate the average values ​​of the user's overall certification space score and recruitment space score, mean(E_grade) and mean(I_grade), respectively. Determine the overall pass rate of the user through calculation.

[0100] PE= + ;

[0101] PI= + ;

[0102] Wherein, λ is the passing ratio, which is 70%, max() is the maximum value function, min() is the minimum value function, and M is the total number of users. The formula is expressed as adding points to the passing line based on the user's score ratio to determine whether the user's score reaches the passing line. If both of the user's scores reach the passing line, there is no need to perform secondary matching. If the user's score E_grade<PE, jump to step (2); if the user's score I_grade<PI, jump to step (3);

[0103] (2) Cancel the cosine similarity matched this time from the m cosine similarities selected from the set [C_cos], output the remaining m-1 cosine similarities to the dictionary again, and jump to step (4);

[0104] (3): Cancel the cosine similarity matched this time from the m cosine similarities selected from the set [A_cos], re-output the remaining m-1 cosine similarities into the dictionary, and jump to step (4);

[0105] (4): Create a set D and a set P, where the set D includes the updated m-1 sets [C_cos], and the set P includes the updated m-1 sets [A_cos], and input the set D and the set P into the feedback matrix E1 E2 performs secondary matching judgment and updates the matrix , the feedback matrix E1 in set D and set E Set E2 to infinity to start traversal, repeatedly traverse all the cosine similarities in set D and set P and delete one of the m, obtain the corresponding word vector, calculate the shortest distance between all word vectors in set D and the word vector corresponding to the deleted cosine similarity, traverse set P, and determine whether the source data is the same as that in set P. If the same, output the matrix , if different, update the matrix .

[0106] Through the matching algorithm, the job competency units, course knowledge points, and certificate information are automatically matched according to the user's situation. After the user uses it, the matching algorithm is adjusted based on the user's performance feedback to obtain a more reasonable matching algorithm.

[0107] Although the description of the present invention has been quite detailed and has been described in particular with respect to several described embodiments, it is not intended to be limited to any of these details or embodiments or any particular embodiment, so as to effectively cover the intended scope of the present invention. In addition, the present invention is described above with the embodiments foreseeable by the inventors, and its purpose is to provide a useful description, and those non-substantial changes to the present invention that are not currently foreseen may still represent equivalent changes of the present invention.

Claims

1. A talent supply chain platform information interaction method based on virtual reality, characterized in that: The method comprises the following steps: S100: Building a VR classification training classroom based on virtual reality technology; S200: Construct verification space and virtual recruitment space based on virtual reality technology; S300: Realize VR job display area based on virtual reality technology; S400: Build a matching algorithm and implement job matching courses, job matching certificates and course matching certificates; S500: Match and recommend courses, positions and certificates based on matching algorithms; In the step S200, after training in the VR classification training classroom, a verification space and a virtual recruitment space are constructed to test the user's ability; Constructing the examination space: Using the 3D model in Unity to create a virtual examination space in the scene, the virtual examination space includes: simulated examination scene, examination paper selection, examination paper answering; and can realize free roaming in the virtual examination scene, view the examination environment, examination equipment and examination-related information, control the perspective and action of the examinee through the keyboard or mouse, realize the interactive operation of the scene, use the extracted features and label data to train the classifier, adjust the model parameters, and improve the accuracy and generalization ability of the model; Constructing a virtual recruitment space: creating a virtual recruitment space in the scene, the virtual recruitment space includes: a job display area, an applicant registration area, and an interview room; by displaying job information and applicant information on the screen in the virtual recruitment space, users can go through the recruitment process and access the back-end system. Users can obtain job and applicant information through network requests and display them in the virtual recruitment space; the virtual recruitment space screens and recommends candidates for positions by applying machine learning algorithms; In step S300, virtual reality technology creates a virtual job display area in the scene, sets a server interface, obtains and displays the job information in the background through network requests, obtains and dynamically updates the job information by reading a local json file; In step S400, VR classified training classrooms, certification spaces, virtual recruitment spaces, and VR job display areas are constructed through virtual reality technology, and real-time user information, job information, and certificate information of the four areas are obtained. Backstage job information is obtained through the virtual job display area, and job competency units are created based on the backstage job information. Course knowledge points are obtained through the VR classified training classrooms, and certificate information is obtained through the certification space and virtual recruitment space. Data preprocessing is performed on the job competency unit, course knowledge points, and certificate information. Word segmentation is performed according to keywords, topics, and categories, and the word segmentation is represented by word vectors. The word vectors are classified according to the corresponding data sources into job competency unit word vector sets, course knowledge point word vector sets, and certificate information word vector sets. A model is trained using Doc2Vec, where Doc2Vec is a model based on Word2Vec. The Doc2Vec model is used to calculate the cosine similarity A_cos corresponding to each element in the job competency unit word vector set and the course knowledge point word vector set, the cosine similarity B_cos corresponding to each element in the job competency unit word vector set and the certificate information word vector set, and the cosine similarity B_cos corresponding to each element in the course knowledge point word vector set and the certificate information word vector set. The cosine similarity C_cos corresponding to each element in the book information word vector set is used to construct sets [A_cos], [B_cos], and [C_cos] for A_cos, B_cos, and C_cos respectively. For set [A_cos], the m course knowledge points with the highest similarity are selected. For each job competency unit, the m course knowledge points with the highest cosine similarity are selected, and the matching results are output. The job competency units and the corresponding course knowledge points are output as a dictionary; for set [B_cos], the m skill certificates with the highest similarity are selected and output as a dictionary; for [C_cos], the m certificates with the highest similarity are selected and output as a dictionary, and the user's positions, courses, and certificates are matched in combination with the dictionary; Users match courses by cosine similarity, and conduct secondary matching of the job competency units, course knowledge points, and certificate information in the certification space and virtual recruitment space with the matching scores; Read the sets [A_cos], [B_cos], [C_cos], and reorder all elements in the sets [A_cos], [B_cos], and [C_cos] according to the cosine similarity. The minimum value of the value in the set [B_cos] is A1, and the current maximum value of the value in the set [B_cos] is A2. B1=A1 / mean(B_cos), B2=A2 / mean(B_cos), mean(B_cos) represents the average value of all elements in the set [B_cos]. B1 and B2 represent the maximum and minimum ratios between certificates and positions. Define Li=B1+(B2-B1)×(C_cosi-min(C_cos)) / (max(C_cos)-min(C_cos)), where C_cosi is the ith element in the set [C_cos], min(C_cos) is the value of the smallest element in the set [C_cos], and max(C_cos) is the value of the largest element in the set [C_cos]. The Li is expressed as judging the corresponding value of the course and the certificate by the ratio of the relationship between the position and the certificate, and constructing the array E1 with Li; Define Ki=B1+(B2-B1)×(B_cosi-min(A_cos)) / (max(A_cos)-min(A_cos)), where A_cosi is the ith element in the set [A_cos], min(A_cos) is the value of the smallest element in the set [A_cos], and max(A_cos) is the value of the largest element in the set [A_cos]. The Ki is expressed as judging the corresponding value of the course and the position by the ratio of the relationship between the position and the certificate, and constructing the array E2 with Ki; Construct feedback matrix E1 from arrays E1 and E2 E2, the method of performing secondary matching of the position competency units, course knowledge points, and certificate information by combining the feedback matrix with the matching scores is as follows: S401: Obtain the grades of the matched certification space and virtual recruitment space, and define the certification space grade as E_grade, and the recruitment space grade as I_grade. Calculate the average values ​​of the user's overall certification space grade and recruitment space grade mean(E_grade) and mean(I_grade), respectively, and determine the overall pass rate of the user through calculation. PE= + ; PI= + ; Among them, PE represents the passing rate of the user in the certification space, PI represents the passing rate of the user in the recruitment space, λ is the passing ratio, which is 70%, max() is the maximum value function, min() is the minimum value function, M is the total number of users, and the formula is expressed as adding points to the base of the passing line according to the user's score ratio to determine whether the user's score reaches the passing line. If both of the user's scores reach the passing line, there is no need to perform secondary matching. If the user's score E_grade<PE, jump to S402; if the user's score I_grade<PI, jump to S403; S402: cancel the cosine similarity matched this time from the m cosine similarities selected from the set [C_cos], re-output the remaining m-1 cosine similarities into the dictionary, and jump to S404; S403: cancel the cosine similarity matched this time from the m cosine similarities selected from the set [A_cos], output the remaining m-1 cosine similarities to the dictionary again, and jump to S404; S404: Create a set D and a set P, wherein the set D includes the updated m-1 sets [C_cos], and the set P includes the updated m-1 sets [A_cos], and input the set D and the set P into the feedback matrix E1 E2 performs secondary matching judgment and updates the matrix , the feedback matrix E1 in set D and set E Set E2 to infinity to start traversal, repeatedly traverse all the cosine similarities in set D and set P and delete one of the m, obtain the corresponding word vector, calculate the shortest distance between all word vectors in set D and the word vector corresponding to the deleted cosine similarity, traverse set P, and determine whether the source data is the same as that in set P. If the same, output the matrix , if different, update the matrix .

2. According to the method of information interaction of a talent supply chain platform based on virtual reality in claim 1, it is characterized in that: In step S100, a 3D scene is created in Unity, and a SteamVR plug-in is imported, a virtual classroom is created in the scene, and a script is added to the virtual classroom to realize the interactive behavior of the control elements. The interactive behavior includes: triggering corresponding actions through collision or induction detection to increase the authenticity of the virtual classroom; By adding objects in the classroom or displaying relevant information on the projector, users can learn knowledge of different categories and realize the VR classification training function. The display of relevant information on the projector is represented by: accessing the back-end system, obtaining the content of classification training through network requests, and projecting the content of classification training; importing the teacher model, and realizing the interaction between students and teachers.

3. The method for information interaction of a talent supply chain platform based on virtual reality according to claim 1, characterized in that: In step S500, the job competency units, course knowledge points, and certificate information are obtained through a matching algorithm and automatically matched according to the user's situation. After the user uses it, the matching algorithm is adjusted through user performance feedback to obtain a more reasonable matching algorithm.

Citation Information

Patent Citations

  • Human resource-oriented multi-source heterogeneous big data fusion basic information system

    CN110689321A

  • Network online education system

    CN112288314A

  • Employment competitiveness analysis system based on online education

    CN112597365A

  • Training system based on Internet virtual reality

    CN115756157A