Educational Evaluation Method for Dynamic Vocational Competence and Post Matching Based on Knowledge Graph

Abstract

The present invention discloses an educational evaluation method for dynamic vocational ability and job matching based on a knowledge graph. S1. Construct a vocational ability knowledge graph; S2. Construct a job requirement knowledge graph; S3. Design and train a double-layer generative adversarial network model; S4. Match the vocational ability enhancement plan and job requirement changes generated by the double-layer generative adversarial network model with the existing vocational ability knowledge graph and job requirement knowledge graph; S5. Based on the dynamic changes in market demand and the development of individual vocational abilities, update the vocational ability knowledge graph and job requirement knowledge graph in real time, and use the double-layer generative adversarial network model to generate new matching relationships to continuously optimize the matching results; S6. Generate an educational evaluation report and career planning suggestions according to the optimized matching results. The present invention significantly improves the dynamic adaptability and matching accuracy of vocational ability and job matching, and realizes the precise docking between job seekers and job requirements.

Description

Technical Field

[0001] The present invention relates to the technical field of educational assessment, and particularly to an educational assessment method for dynamic vocational ability and job matching based on a knowledge graph. Background Art

[0002] In the prior art, the matching of vocational ability and job mainly relies on traditional assessment methods and static matching algorithms. Traditional methods usually determine an individual's vocational ability by analyzing static data such as resumes, skill test results, and interview feedback, and make job recommendations based on preset matching rules or historical data. However, this method gradually shows many deficiencies when dealing with the rapidly changing market demands and the dynamic improvement of individual vocational abilities.

[0003] First of all, the vocational ability assessment in the prior art often based on fixed criteria and rules, which cannot timely reflect the dynamic changes of individual vocational abilities, resulting in the fact that the actual abilities and potentials of individuals are difficult to be fully reflected in the career planning process, thus affecting the accuracy and personalization of the matching. At the same time, traditional matching algorithms are usually based on static rules and are difficult to adjust the matching strategy in real time according to the changes in market demands. The static matching method often leads to the lag of job recommendations behind market demands and cannot meet the rapidly changing talent needs of enterprises.

[0004] Secondly, there is no effective linkage mechanism between vocational education assessment and job requirements in the prior art. The training content of vocational education usually cannot reflect the latest market demands, resulting in the disconnection between educational achievements and the employment market. In addition, existing vocational assessment methods and job matching systems are usually independent of each other, and the educational assessment results are not effectively applied to the job matching process. There is a lack of a closed-loop feedback system to guide the individual's career development path, reducing the pertinence and actual effect of education and training.

[0005] In addition, the existing vocational ability and job matching technologies have obvious deficiencies in dealing with dynamic data and real-time optimization. With the development of the market environment and technology, the requirements for job skills and market trends in enterprises are constantly changing, while the existing matching technologies often cannot update the matching rules and algorithms in time, resulting in the lag of job matching results, thus affecting the employment efficiency of enterprises and the career development of individuals.

[0006] In summary, the existing technologies mainly have the following disadvantages: First, the assessment of vocational abilities and job matching rely on static data and cannot dynamically reflect the changes and development of individual vocational abilities. Second, there is a lack of effective linkage between vocational education and job requirements, and the results of educational assessment cannot be directly used for job matching. Finally, the existing technologies have deficiencies in processing dynamic data and real-time updating of matching strategies, and it is difficult to cope with the rapidly changing market demands and job requirements. These problems significantly limit the accuracy and timeliness of the matching between vocational abilities and jobs and cannot meet the needs of modern enterprises and individuals for efficient matching. Summary of the Invention

[0007] An object of the present invention is to provide an educational assessment method for dynamic vocational ability and job matching based on a knowledge graph. The present invention significantly improves the dynamic adaptability and matching accuracy of vocational ability and job matching, and realizes the precise docking between job seekers and job requirements.

[0008] An educational assessment method for dynamic vocational ability and job matching based on a knowledge graph according to an embodiment of the present invention includes the following steps:

[0009] S1. Collect and organize data on an individual's educational background, work experience, skill certificates, and project experience, and structure them into a vocational ability knowledge graph to represent the vocational ability characteristics and association relationships of the individual;

[0010] S2. Collect data on the skill requirements, industry standards, and market trends of enterprise positions, and structure them into a job requirement knowledge graph to represent the job characteristics and requirements, and dynamically adjust the job requirement knowledge graph according to the update cycle of market data;

[0011] S3. Design and train a two-layer generative adversarial network model. The first-layer generative adversarial network is used to generate possible vocational ability enhancement schemes, and the second-layer generative adversarial network is used to simulate the impact of changes in market demand on job requirements;

[0012] S4. By matching the vocational ability enhancement schemes and job requirement changes generated by the two-layer generative adversarial network model with the existing vocational ability knowledge graph and job requirement knowledge graph, and using the potential matching relationships generated by the two-layer generative adversarial network model, optimize the matching algorithm;

[0013] S5. Based on the dynamic changes in market demand and the development of individual vocational abilities, real-time update the vocational ability knowledge graph and job requirement knowledge graph, and use the two-layer generative adversarial network model to generate new matching relationships to continuously optimize the matching results;

[0014] S6. Generate an education assessment report and career planning suggestions based on the optimized matching results, and feedback them to educational institutions and individuals to guide adjustments between the improvement of vocational capabilities and job requirements, ultimately forming a closed-loop vocational education and job matching system.

[0015] Optionally, the S1 includes the following sub-steps:

[0016] S11. Obtain the educational background data of an individual, where the educational background data includes academic qualifications, majors studied, graduation institutions, and educational achievement information, and convert the educational background data into an educational background vector E i ;

[0017] S12. Obtain the work experience data of an individual, where the work experience data includes work history, job responsibilities, and industry experience information, and convert the work experience data into a work experience vector W i ;

[0018] S13. Obtain the skill certificate data of an individual, where the skill certificate data includes technical qualification certificates, professional certifications, and language proficiency certificates information, and convert the skill certificate data into a skill certificate vector C i ;

[0019] S14. Obtain the project experience data of an individual, where the project experience data includes project names, project roles, and project outcome information, and convert the project experience data into a project experience vector P i ;

[0020] S15. Combine the educational background vector E i , work experience vector W i , skill certificate vector C i and project experience vector P i to generate the comprehensive vocational ability vector A i :

[0021] A i = α·E i + β·W i + γ·C i + δ·P i ;

[0022] where α, β, γ, δ are the weight coefficients of the corresponding vectors, representing the influence weights of different vocational ability characteristics on the comprehensive vocational ability;

[0023] S16. Based on the comprehensive vocational ability vector A iAs a result, by using knowledge graph technology, the educational background, work experience, skill certificates, and project experience information of an individual are structurally represented to generate nodes of the individual's professional ability knowledge graph, and the association relationships between the nodes are established to represent the individual's professional ability characteristics and the information of each associated dimension.

[0024] Optionally, S16 includes the following sub-steps:

[0025] S161. Decompose each data feature in the comprehensive professional ability vector A of the individual i into independent knowledge graph nodes to generate a node set N i ;

[0026] S162. Assign a unique identifier N i，j to each node, where N represents a knowledge graph node, i represents an individual identifier, j represents the feature dimension of the node in the comprehensive professional ability vector A i , and the relationship between the nodes is described as follows:

[0027]

[0028] Among them, φ is a non-linear activation function used to project the comprehensive professional ability vector A i into the node space through the feature weight matrix W j , λ m is the cross-correlation coefficient between features, and η i，j，m represents the interaction effect between different features and is used to adjust the uniqueness and discrimination of the nodes;

[0029] S163. Construct an edge set R i，j，k according to the association relationship between the nodes, where R i，j，k represents the association relationship between node N i，j and node N i，k :

[0030]

[0031] Among them, ψ is an association mapping function, sim(N i，j , N i，k ) represents the similarity measurement between nodes, ζ j，k，p is a parameter for adjusting the similarity measurement, Δ j，k，p represents the difference measurement of a specific feature dimension, ω t is a weight parameter, and v i，j，k，t is a multi-dimensional interaction term for correcting the association relationship;

[0032] S164. Analyze the potential association relationships between the nodes and adjust the weight value r of the edge set R i，j，k according to the inference resulti，j，k :

[0033]

[0034] Among them, ξ is a function for updating the weight value, and κ v represents the influence factor of the relationship between different nodes on the correlation weight during the inference process, and μ y is the edge adjustment parameter, and X i，j，k，y is the multi-dimensional potential correlation relationship discovered during the inference process;

[0035] S165. Generate the occupational ability knowledge graph G of the individual i =(N i , R i ), where N i represents the set of all nodes of the individual's occupational ability knowledge graph, and R i represents the set of all association relationships between nodes. The generation of the occupational ability knowledge graph is expressed as:

[0036]

[0037] Among them, θ a，c is the adjustment parameter in the generation of the occupational ability knowledge graph, and Ω i，a，c is the high-dimensional relationship item used to enhance the structure of the occupational ability knowledge graph.

[0038] Optionally, the S2 includes the following sub-steps:

[0039] S21. Obtain the skill requirement data of the enterprise positions. The skill requirement data includes the technical capabilities, professional knowledge, and soft skills required for the positions, and convert the skill requirement data into a skill demand vector Q j ;

[0040] S22. Obtain the industry standard data. The industry standard data includes industry certification standards, industry norms, and best practices, and convert the industry standard data into an industry standard vector S j ;

[0041] S23. Obtain the market trend data. The market trend data includes market demand, technological development trends, and economic environment changes, and convert the market trend data into a market trend vector T j ;

[0042] S24. Merge the skill demand vector Q j , the industry standard vector S j and the market trend vector T j to generate a position demand vector D j :

[0043] D j = ρ·Qj +σ·S j +τ·T j ;

[0044] Among them, ρ, σ, and τ are the weight coefficients of skill requirements, industry standards, and market trends respectively, used to adjust the influence degree of each vector on the job requirements;

[0045] S25. Use the knowledge graph technology to structurally represent the job requirement vector D j as the job requirement knowledge graph node, generating the job requirement node set M j , where each node represents a specific job feature, and generating the job requirement knowledge graph G j :

[0046] G j =(M j , R j );

[0047] Among them, M j represents the set of job requirement nodes, and R j represents the association relationship between nodes. The association relationship is realized by constructing the edge set R j，p，q，r :

[0048]

[0049] Among them, sim(M j，p , M j，q , M j，r ) represents the similarity measure ζ p，q，r，u is the parameter for adjusting the similarity measure, Δ p，q，r，u represents the difference measure of a specific feature dimension, ω w is the weight parameter, and v j，p，q，r，w is the multi-dimensional interaction term for correcting the association relationship;

[0050] S26. Dynamically adjust the job requirement knowledge graph G j according to the update period of the market data, updating the weight values r j in the node set M j and the edge set R j，p，q，r .

[0051] Optionally, the S3 includes the following sub-steps:

[0052] S31. Design and initialize the first-layer generative adversarial network for generating the professional ability enhancement plan. The first-layer generative adversarial network consists of the first-layer generator G 1 (z 1 ) and the first-layer discriminator D 1 (A i ), where z1 is the input noise vector, and the first-layer generator G 1 (z 1 ) generates a vector representation of the enhanced vocational ability plan, denoted as The first-layer discriminator D 1 (A i ) is used to distinguish the generated enhanced vocational ability plan from the original vocational ability vector A i ;

[0053] S32. Design and initialize the second-layer generative adversarial network for simulating the impact of market demand changes on job requirements. The second-layer generative adversarial network consists of the second-layer generator G 2 (z 2 ) and the second-layer discriminator D 2 (D j ) where z 2 is the input noise vector, and the vector of job demand changes generated by the second-layer generator G 2 (z 2 ) is denoted as The second-layer discriminator D 2 (D j ) is used to distinguish the generated job demand changes from the original job demand vector D j ;

[0054] S33. Jointly train the first-layer and second-layer generative adversarial networks, adopt the adversarial training mechanism, and optimize the first-layer generator G 1 and the second-layer generator G 2 . The objective function of the joint training is expressed as:

[0055]

[0056] S34. Match the enhanced vocational ability plan generated by the trained first-layer generative adversarial network with the vocational ability knowledge graph G i to analyze and optimize the individual vocational ability improvement strategy;

[0057] S35. Match the job demand changes generated by the trained second-layer generative adversarial network with the job demand knowledge graph G j to evaluate the impact of market demand changes on job requirements and optimize the job matching strategy;

[0058] S36. Based on the matching results of the first-layer and second-layer generative adversarial networks, optimize the matching algorithm for vocational ability and job demand, and adjust the matching parameters to make the matching degree function reach the optimal:

[0059]

[0060] Among them, is a matching degree function, which is used to evaluate the matching degree between the professional ability enhancement plan and the changes in job requirements.

[0061] Optionally, the step S6 includes the following sub-steps:

[0062] S61. Based on the matching result between the professional ability and the job requirements, calculate the matching degree score between the individual's professional ability and the target job requirements. The scoring range is from 0 to 1, where 1 represents a perfect match and 0 represents no match;

[0063] S62. Generate an education evaluation report according to the matching degree score. The content of the education evaluation report includes the advantages and disadvantages of the individual's current professional ability, the specific requirements of the target job, and the professional skill fields that need to be improved;

[0064] S63. Compare the analysis results in the education evaluation report with the professional ability knowledge graph G i to identify the professional ability characteristics that need to be further improved, and generate the individual's career planning advice P i . The career planning advice includes the courses recommended to study, the skill certificates required to obtain, and the practical projects recommended to participate in:

[0065]

[0066] Among them, is the target professional ability characteristic value, is the current professional ability characteristic value, and w k is the weight coefficient of each professional ability characteristic in the career planning;

[0067] S64. Feed back the career planning advice P i to the educational institution and the individual to guide them to make corresponding adjustments in the process of professional ability improvement. The educational institution adjusts the curriculum according to the advice, and the individual selects the learning path and practical activities according to the advice;

[0068] S65. Dynamically adjust the professional ability knowledge graph G i and the job requirement knowledge graph G j according to the feedback data of the educational institution and the individual, and recalculate the matching degree score;

[0069] S66. Through multiple feedbacks and optimization iterations, finally form a closed-loop vocational education and job matching system. The system provides the optimal career planning advice and job matching plan in real time according to the changes in market demand and the development of individual professional ability.

[0070] The beneficial effects of the present invention are:

[0071] (1) The present invention constructs a vocational ability knowledge graph and a job requirement knowledge graph, achieving an accurate characterization of the correlation between job seekers' abilities and job requirements. Different from traditional static matching methods, the system can dynamically adjust the nodes and edge weights in the knowledge graph according to the changes in market demand and the development of individual vocational abilities. Through the application of a double-layer generative adversarial network model, the system can generate vocational ability enhancement plans and simulate the changes in market demand, thereby achieving adaptive optimization. This not only improves the real-time performance and accuracy of matching but also effectively responds to the rapidly changing market demand, making the matching results more personalized and forward-looking.

[0072] (2) The present invention, through a double-layer generative adversarial network model, especially through the market demand change simulation function of the second-layer generative adversarial network, predicts and captures the potential change trends of job requirements, and combines with the vocational ability enhancement plan generated by the first-layer generative adversarial network to discover potential matching relationships. The system can predict the rise of certain emerging technologies and thus proactively suggest that job seekers acquire relevant skills to enhance their market competitiveness. Through the discovery of potential matching relationships, the system can not only make effective matches in the current market environment but also prepare for future job requirements, improving the forward-looking and accuracy of talent allocation.

[0073] (3) The present invention combines educational evaluation with the job matching process to form a closed-loop feedback mechanism. Through the comprehensive application of the knowledge graph and the generative adversarial network model, the system can not only generate matching results but also generate detailed educational evaluation reports and career planning suggestions based on the matching degree analysis and changes in market demand. Description of the Drawings

[0074] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0075] Figure 1 is a flowchart of an educational evaluation method for dynamic vocational ability and job matching based on a knowledge graph proposed by the present invention. Detailed Embodiments

[0076] Now, the present invention will be further described in detail with reference to the drawings. These drawings are all simplified schematic diagrams, only showing the basic structure of the present invention in a schematic way, so they only show the components related to the present invention.

[0077] Reference Figure 1 , an educational evaluation method for dynamic vocational ability and job matching based on a knowledge graph, includes the following steps:

[0078] S1. Collect and organize the individual's educational background, work experience, skill certificates, and project experience data, and structure them into a career ability knowledge graph to represent the individual's career ability characteristics and association relationships;

[0079] S2. Collect the skill requirements, industry standards, and market trend data of enterprise positions, and structure them into a position demand knowledge graph to represent the position characteristics and requirements, and dynamically adjust the position demand knowledge graph according to the update cycle of market data;

[0080] S3. Design and train a two-layer generative adversarial network model. The first-layer generative adversarial network is used to generate possible career ability enhancement plans, and the second-layer generative adversarial network is used to simulate the impact of changes in market demand on position requirements;

[0081] S4. By matching the career ability enhancement plans and position demand changes generated by the two-layer generative adversarial network model with the existing career ability knowledge graph and position demand knowledge graph, and using the potential matching relationships generated by the two-layer generative adversarial network model, optimize the matching algorithm;

[0082] S5. Based on the dynamic changes in market demand and the development of individual career abilities, update the career ability knowledge graph and position demand knowledge graph in real time, and use the two-layer generative adversarial network model to generate new matching relationships to continuously optimize the matching results;

[0083] S6. According to the optimized matching results, generate an education assessment report and career planning suggestions, and feedback them to educational institutions and individuals to guide adjustments between career ability improvement and position requirements, and finally form a closed-loop vocational education and position matching system.

[0084] In this embodiment, S1 includes the following sub-steps:

[0085] S11. Obtain the individual's educational background data, which includes academic qualifications, majors studied, graduation institutions, and educational achievement information, and convert the educational background data into an educational background vector E i ;

[0086] S12. Obtain the individual's work experience data, which includes work experience, job responsibilities, and industry experience information, and convert the work experience data into a work experience vector W i ;

[0087] S13. Obtain the individual's skill certificate data, which includes technical qualification certificates, professional certifications, and language ability certificate information, and convert the skill certificate data into a skill certificate vector C i ;

[0088] S14. Obtain the project experience data of an individual, where the project experience data includes the project name, project role, and project result information, and convert the project experience data into a project experience vector P i ;

[0089] S15. Combine the educational background vector E i , professional experience vector W i , skill certificate vector C i and project experience vector P i to generate the comprehensive professional ability vector A of the individual i :

[0090] A i = α·E i + β·W i + γ·C i + δ·P i ;

[0091] Among them, α, β, γ, and δ are the weight coefficients of the corresponding vectors, representing the influence weights of different professional ability characteristics on the comprehensive professional ability;

[0092] S16. Based on the result of the comprehensive professional ability vector A i , use knowledge graph technology to structurally represent the educational background, professional experience, skill certificates, and project experience information of the individual, generate the knowledge graph nodes of the individual's professional ability, and establish the association relationships between the nodes to represent the individual's professional ability characteristics and the information of each associated dimension.

[0093] In this embodiment, S16 includes the following sub-steps:

[0094] S161. Decompose the data features in the comprehensive professional ability vector A of the individual i into independent knowledge graph nodes to generate a node set N i ;

[0095] S162. Assign a unique identifier N to each node i，j , where N represents the knowledge graph node, i represents the individual identifier, j represents the feature dimension of the node in the comprehensive professional ability vector A i , and the relationship between the nodes is described as follows:

[0096]

[0097] Among them, φ is a non-linear activation function used to project the comprehensive professional ability vector A i into the node space through the feature weight matrix W j , λ m is the cross-correlation coefficient between the features, and η i，j，mRepresents the interaction effect between different features, used to adjust the uniqueness and distinctiveness of nodes;

[0098] S163. Construct an edge set R according to the association relationship between nodes i，j，k , where R i，j，k represents the association relationship between node N i，j and node N i，k :

[0099]

[0100] where ψ is an association mapping function, sim(N i，j , N i，k ) represents the similarity measure between nodes, ζ j，k，p is a parameter for adjusting the similarity measure, Δ j，k，p represents the difference measure of a specific feature dimension, ω t is a weight parameter, and V i，j，k，t is a multi-dimensional interaction term for correcting the association relationship;

[0101] S164. Analyze the potential association relationships between nodes and adjust the weight value r i，j，k of the edge set R i，j，k :

[0102]

[0103] where ξ is a function for updating the weight value, Kv represents the influence factor of the relationship between different nodes on the association weight during the reasoning process, μ y is an edge adjustment parameter, and χ i，j，k，y is the multi-dimensional potential association relationship discovered during the reasoning process;

[0104] S165. Generate the occupational ability knowledge graph G i =(N i , R i ), where N i represents the set of all nodes of the individual's occupational ability knowledge graph, and R i represents the set of all association relationships between nodes. The generation of the occupational ability knowledge graph is expressed as:

[0105]

[0106] where θ a，c is an adjustment parameter in the generation of the occupational ability knowledge graph, and Ω i，a，c is a high-dimensional relationship term for enhancing the structure of the occupational ability knowledge graph.

[0107] In this embodiment, S2 includes the following sub-steps:

[0108] S21. Obtain the skill requirement data for enterprise positions. The skill requirement data includes the technical capabilities, professional knowledge, and soft skills required for the positions, and convert the skill requirement data into a skill demand vector Q j ;

[0109] S22. Obtain the industry standard data. The industry standard data includes industry certification standards, industry norms, and best practices, and convert the industry standard data into an industry standard vector S j ;

[0110] S23. Obtain the market trend data. The market trend data includes market demand, technological development trends, and changes in the economic environment, and convert the market trend data into a market trend vector T j ;

[0111] S24. Merge the skill demand vector Q j , the industry standard vector S j , and the market trend vector T j to generate a position demand vector D j :

[0112] D j = ρ·Q j + σ·S j + τ·T j ;

[0113] Where ρ, σ, and τ are the weight coefficients of skill requirements, industry standards, and market trends respectively, used to adjust the influence degree of each vector on the position demand;

[0114] S25. Use knowledge graph technology to structurally represent the position demand vector D j as a position demand knowledge graph node to generate a position demand node set M j , where each node represents a specific position feature, and generate a position demand knowledge graph G j :

[0115] G j = (M j , R j );

[0116] Where M j represents the set of position demand nodes, and R j represents the association relationship between nodes. The association relationship is realized by constructing an edge set R j，p，q，r :

[0117]

[0118] Where sim(M j，p , M j，q , Mj，r ) represents the similarity measure ζ between nodes p，q，r，u is the parameter for adjusting the similarity measure, Δ p，q，r，u represents the difference measure ω of a specific feature dimension w is the weight parameter, v j，p，q，r，w is the multi-dimensional interaction term for correcting the association relationship;

[0119] S26. Dynamically adjust the job demand knowledge graph G according to the update period of market data j , update the node set M j and the weight value r in the edge set R j j，p，q，r .

[0120] In this embodiment, S3 includes the following sub-steps:

[0121] S31. Design and initialize the first-layer generative adversarial network for generating the professional ability enhancement plan. The first-layer generative adversarial network consists of the first-layer generator G 1 (z 1 ) and the first-layer discriminator D 1 (A i ), where z 1 is the input noise vector, and the professional ability enhancement plan vector generated by the first-layer generator G 1 (z 1 ) is represented as The first-layer discriminator D 1 (A i ) is used to distinguish the generated professional ability enhancement plan from the original professional ability vector A i ;

[0122] S32. Design and initialize the second-layer generative adversarial network for simulating the impact of market demand changes on job requirements. The second-layer generative adversarial network consists of the second-layer generator G 2 (z 2 ) and the second-layer discriminator D 2 (D j ), where z 2 is the input noise vector, and the job demand change vector generated by the second-layer generator G 2 z 2 is represented as The second-layer discriminator D 2 D j is used to distinguish the generated job demand change from the original job demand vector D j ;

[0123] S33. Jointly train the first-layer generative adversarial network and the second-layer generative adversarial network, and adopt the adversarial training mechanism to optimize the first-layer generator G​1 and the second - layer generator G 2 , the objective function for joint training is expressed as:

[0124]

[0125] S34. Match the enhanced vocational ability plan generated by the trained first - layer generative adversarial network with the vocational ability knowledge graph G i to analyze and optimize the individual vocational ability improvement strategy;

[0126] S35. Match the job demand changes generated by the trained second - layer generative adversarial network with the job demand knowledge graph G j to evaluate the impact of market demand changes on job requirements and optimize the job matching strategy;

[0127] S36. Based on the matching results of the first - layer and second - layer generative adversarial networks, optimize the matching algorithm for vocational ability and job demand, and adjust the matching parameters to make the matching degree function reach the optimal:

[0128]

[0129] where, is the matching degree function, which is used to evaluate the matching degree between the enhanced vocational ability plan and the job demand changes.

[0130] In this embodiment, S6 includes the following sub - steps:

[0131] S61. Based on the matching results of vocational ability and job demand, calculate the matching degree score between the individual's vocational ability and the target job demand. The score range is from 0 to 1, where 1 represents a perfect match and 0 represents no match;

[0132] S62. Generate an education evaluation report according to the matching degree score. The content of the education evaluation report includes the advantages and disadvantages of the individual's current vocational ability, the specific requirements of the target job, and the vocational skill areas that need to be improved;

[0133] S63. Compare the analysis results in the education evaluation report with the vocational ability knowledge graph G i to identify the vocational ability characteristics that need to be further improved, and generate the individual's career planning advice P i , and the career planning advice includes the courses recommended to study, the skill certificates required to obtain, and the practical projects recommended to participate in:

[0134]

[0135] Among them, is the target vocational ability eigenvalue, is the current vocational ability eigenvalue, w k is the weight coefficient of each vocational ability characteristic in career planning;

[0136] S64. Provide the career planning advice P i to the educational institution and the individual, guiding them to make corresponding adjustments in the process of vocational ability improvement. The educational institution adjusts the curriculum according to the advice, and the individual selects the learning path and practical activities according to the advice;

[0137] S65. Dynamically adjust the vocational ability knowledge graph G i and the job requirement knowledge graph G j according to the feedback data of the educational institution and the individual, and recalculate the matching degree score;

[0138] S66. Through multiple feedbacks and optimization iterations, finally form a closed-loop vocational education and job matching system. The system provides the optimal career planning advice and job matching plan in real time according to the changes in market demand and the development of individual vocational ability.

[0139] Example 1:

[0140] In June 2024, the HR department of a technology enterprise in Beijing was recruiting a senior AI algorithm engineer. In order to cope with the rapid changes in the market and the complexity of job requirements, the enterprise decided to conduct an experiment using the present invention.

[0141] The HR department first extracted the detailed information of 10 job seekers from its talent database. The backgrounds of the job seekers covered the fields of computer science and artificial intelligence. The educational background, work experience, skill certificates and project experiences of each job seeker were collected by the system and converted into a vocational ability knowledge graph.

[0142] In the example, job seeker Zhang San (ID number A001), male, 31 years old, graduated from the computer science major of University A in 2015, with a master's degree. His work experience includes working as a data scientist in a well-known Internet company, participating in the development of multiple AI projects, and having a TensorFlow advanced development certification. After structuring the information, the system generated Zhang San's vocational ability knowledge graph, which includes multiple nodes, such as the educational background node (master of computer science major at University A), the work experience node (5 years of data scientist experience), the skill certificate node (TensorFlow advanced development certification), and the project experience node (participating in multiple AI projects).

[0143] Meanwhile, the requirements for the enterprise AI algorithm engineer position have also been structured into a job requirements knowledge graph. This position requires candidates to have professional knowledge in machine learning and deep learning, be able to handle large-scale data, and have more than 3 years of relevant work experience. In addition, due to the rapidly growing market demand for edge computing technology, familiarity with edge computing technology is particularly emphasized in the job requirements graph.

[0144] After systematically analyzing Zhang San's career ability knowledge graph and the job requirements graph for AI algorithm engineers, it was found that Zhang San's career ability score reached 85 points (out of 100). Subsequently, the system used a two-layer generative adversarial network (GAN) model for further optimization. First, the first layer of the generative adversarial network generated a career ability enhancement plan, suggesting that Zhang San could participate in a two-week intensive training on edge computing technology to improve his matching degree. Then, the second layer of the generative adversarial network simulated the changes in market demand and predicted that the demand for edge computing would continue to increase within the next six months. Therefore, it was recommended that the HR department give priority to candidates with edge computing skills.

[0145] After the optimized matching algorithm, Zhang San's career ability score was increased to 92 points. The system generated a detailed matching report, recommending that the HR department invite Zhang San to participate in the next round of interviews and providing Zhang San with specific career development suggestions. The report listed Zhang San's current advantages, including his rich experience in the field of AI algorithms and solid academic background, while also pointing out his room for improvement in edge computing.

[0146] To verify the effectiveness of the system, the HR department decided to compare the traditional recruitment method with the new system at the same time. During the two-month recruitment cycle, the HR department interviewed 20 candidates using the traditional method, among which 5 entered the final interview round, and only 1 passed the probation period in the end. The overall recruitment process took up to 50 days.

[0147] Through the knowledge graph-based system, the HR department screened 10 high-potential candidates within the same time. Among them, 7 entered the final interview round, and finally 4 passed the probation period. The whole process only took 25 days. The matching accuracy rate provided by the system reached 90%, which was 20% higher than the traditional method. At the same time, through the career development suggestions generated by the system, the 4 employees who passed the probation period performed excellently at work. Among them, 2 successfully completed edge computing-related projects during the probation period, creating significant economic benefits for the company.

[0148] Table 1 Comparison data of the method of the present invention and the traditional method in terms of recruitment cycle and job matching

[0149]

[0150] As can be seen from Example 1, the educational evaluation method for dynamic vocational ability and job matching based on the knowledge graph not only significantly shortens the recruitment cycle, improves the accuracy of job matching, but also effectively increases the passing rate of the probation period and job adaptability of employees. The HR department finally decides to fully adopt this system and plans to further promote the application of this technology in future recruitment and career development to better meet the market changes and enterprise development needs.

[0151] By constructing a vocational ability knowledge graph and a job requirement knowledge graph, the present invention realizes the accurate characterization of the correlation between the abilities of job seekers and job requirements. Different from the traditional static matching method, the system can dynamically adjust the nodes and edge weights in the knowledge graph according to the changes in market demand and the development of individual vocational abilities. Through the application of the double-layer generative adversarial network model, the system can generate vocational ability enhancement plans and simulate the changes in market demand, thereby realizing adaptive optimization. This not only improves the real-time performance and accuracy of matching, but also effectively responds to the rapidly changing market demand, making the matching results more personalized and forward-looking.

[0152] Through the double-layer generative adversarial network model, especially through the market demand change simulation function of the second-layer generative adversarial network, the present invention predicts and captures the potential change trends of job requirements, and combines with the vocational ability enhancement plan generated by the first-layer generative adversarial network to discover potential matching relationships. The system can predict the rise of some emerging technologies and thus proactively suggest that job seekers acquire relevant skills to enhance their market competitiveness. Through the discovery of potential matching relationships, the system can not only make effective matches in the current market environment, but also prepare for future job requirements, improving the forward-looking and accuracy of talent allocation.

[0153] The present invention combines the educational evaluation with the job matching process to form a closed-loop feedback mechanism. Through the comprehensive application of the knowledge graph and the generative adversarial network model, the system can not only generate matching results, but also generate detailed educational evaluation reports and career planning suggestions according to the matching degree analysis and market demand changes.

[0154] The above is only the preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and all should be covered by the protection scope of the present invention.

Claims

1. An education assessment method for dynamic professional ability and job matching based on knowledge graph, characterized in that: The steps include: S1. Collect and organize individual education background, professional experience, skill certificates and project experience data, and structure them into a professional competency knowledge graph; S2. Collect the skill requirements, industry standards and market trend data of enterprise positions and structure them into a job demand knowledge map; S3. Design and train a two-layer generative adversarial network model. The first layer of the generative adversarial network is used to generate professional ability enhancement plans, and the second layer of the generative adversarial network is used to simulate the impact of changes in market demand on job requirements; S4. By matching the occupational capability enhancement scheme and job requirement changes generated by the two-layer generative adversarial network model with the existing occupational capability knowledge graph and job requirement knowledge graph, the matching algorithm is optimized using the potential matching relationship generated by the two-layer generative adversarial network model. S5. Based on the dynamic changes in market demand and the development of individual professional capabilities, the professional capability knowledge graph and job requirement knowledge graph are updated in real time, and a two-layer generative adversarial network model is used to generate new matching relationships and continuously optimize matching results; S6. Generate education assessment reports and career planning suggestions based on the optimized matching results, and provide feedback to educational institutions and individuals to guide them in making adjustments between career improvement and job requirements, ultimately forming a closed-loop vocational education and job matching system; The S3 comprises the following sub-steps: S31. Design and initialize the first-layer generative adversarial network to generate a professional ability enhancement plan. The first-layer generative adversarial network consists of the first-layer generator G1(z1) and the first-layer discriminator D1(A i ), where z1 is the input noise vector, and the occupational ability enhancement solution vector generated by the first layer generator G1(z1) is expressed as : The first layer discriminator D1 (A i ) is used to distinguish the generated occupational capability enhancement scheme from the original occupational capability vector A i ; S32, design and initialize the second-layer generative adversarial network to simulate the impact of changes in market demand on job requirements. The second-layer generative adversarial network consists of the second-layer generator G2(z2) and the second-layer discriminator D2(D j ), where z2 is the input noise vector, and the job demand change vector generated by the second-layer generator G2(z2) is expressed as : The second layer discriminator D2 (D j ) is used to distinguish the generated job demand changes from the original job demand vector D j ; S33, jointly train the first-layer generative adversarial network and the second-layer generative adversarial network, adopt an adversarial training mechanism, optimize the first-layer generator G1 and the second-layer generator G2, and the objective function of the joint training is expressed as: S34. Enhance the professional ability of the first layer of the trained generative adversarial network and professional competence knowledge graph G i Match, analyze and optimize strategies to improve individual professional capabilities; S35. Change the job requirements generated by the second layer of the trained generative adversarial network and job requirements knowledge graph G j Conduct matching, evaluate the impact of changes in market demand on job requirements, and optimize job matching strategies; S36. Based on the matching results of the first-layer generative adversarial network and the second-layer generative adversarial network, optimize the matching algorithm between professional ability and job requirements and adjust the matching parameters Make the matching function To achieve the best: in, It is a matching function, which is used to evaluate the matching degree between the occupational capability enhancement plan and the changes in job requirements.

2. According to claim 1, a method for dynamic professional ability and job matching education assessment based on knowledge graph is characterized in that: The S1 comprises the following sub-steps: S11. Obtain the individual's educational background data, which includes academic qualifications, majors studied, graduate schools, and educational achievement information, and convert the educational background data into an educational background vector E i ; S12. Obtain individual professional experience data, which includes work experience, job responsibilities and industry experience information, and convert the professional experience data into a professional experience vector W i ; S13. Obtain individual skill certificate data, which includes technical qualification certificates, professional certifications, and language proficiency certificate information, and convert the skill certificate data into a skill certificate vector C i ; S14. Obtain individual project experience data, which includes project name, project role and project result information, and convert the project experience data into a project experience vector P. i ; S15, the educational background vector E i , professional experience vector W i , Skill Certificate Vector C i and the project experience vector P i Merge to generate the individual's comprehensive professional ability vector A i : A i =α·E i +β·W i +γ·C i +δ·P i ; Among them, α, β, γ, and δ are the weight coefficients of the corresponding vectors, representing the weights of the influence of different professional ability characteristics on comprehensive professional ability; S16, based on comprehensive professional ability vector A i Based on the results, the knowledge graph technology is used to structure the individual's educational background, professional experience, skill certificates and project experience information, generate the individual's professional ability knowledge graph nodes, and establish the association relationship between the nodes to represent the individual's professional ability characteristics and the various dimensions of information associated with them.

3. According to claim 2, a method for dynamic professional ability and job matching education assessment based on knowledge graph is characterized in that: The S16 comprises the following sub-steps: S161, the individual's comprehensive professional ability vector A i Each data feature in is decomposed into independent knowledge graph nodes, generating a node set N i ; S1 62. Assign a unique identifier N to each node i,j , where N represents the knowledge graph node, i represents the individual identifier, and j represents the node in the comprehensive professional ability vector A i The relationship between nodes in the feature dimension is described as follows: Among them, φ is a nonlinear activation function used to transform the comprehensive professional ability vector A i Through the feature weight matrix W j Projected into the node space, λ m is the correlation coefficient between the features, η i,j,m Represents the interaction effect between different features and is used to adjust the uniqueness and discrimination of nodes; S163. Construct edge set R based on the association relationship between nodes i,j,k , where R i,j,k Represents node N i,j With node N i,k The relationship between: Among them, ψ is an association mapping function, sim(N i,j ,N i,k ) represents the similarity measure between nodes, ζ j,k,p is a parameter to adjust the similarity measure, Δ j,k,p Represents the difference measure of a specific feature dimension, ω t is the weight parameter, ν i,j,k,t is a multidimensional interaction term used to correct the association relationship; S164, analyze the potential association relationship between each node, and adjust the edge set R according to the reasoning result i,j,k The weight value r i,j,k : Among them, ξ is the function used to update the weight value, κ v Indicates the influence factor of the relationship between different nodes on the association weight during the reasoning process, μ y is the edge adjustment parameter, χ i,j,k,y It is the multi-dimensional potential correlation relationship discovered in the reasoning process; S165. Generate individual professional competence knowledge graph G i =(N i ,R i ), where N i Represents the set of all individual professional competence knowledge graph nodes, R i Represents the set of all associations between nodes. The generation of the professional competence knowledge graph is expressed as: Among them, θ a,c is the adjustment parameter in the generation of professional competence knowledge graph, Ω i,a,c It is a high-dimensional relational item used to enhance the structure of the professional competence knowledge graph.

4. According to claim 1, a method for dynamic professional ability and job matching education assessment based on knowledge graph is characterized in that: The S2 comprises the following sub-steps: S21. Obtain the skill requirement data of the enterprise positions. The skill requirement data includes the technical capabilities, professional knowledge and soft skills required for the positions, and convert the skill requirement data into a skill demand vector Q j ; S22. Obtain industry standard data, which includes industry certification standards, industry specifications and best practices, and convert the industry standard data into industry standard vector S j ; S23, obtaining market trend data, the market trend data including market demand, technology development trend and economic environment change, and converting the market trend data into a market trend vector T j ; S24. Transform the skill demand vector Q j , industry standard vector S j and the market trend vector T j Merge to generate job demand vector D j : D j =ρ·Q j +σ·S j +τ·T j ; Among them, ρ, σ and τ are the weight coefficients of skill requirements, industry standards and market trends, which are used to adjust the impact of each vector on job demand; S25. Use knowledge graph technology to transform the job demand vector D j The structured representation is a job requirement knowledge graph node, generating a job requirement node set M j , where each node represents a specific job feature and generates a job requirement knowledge graph G j : G j =(M j ,R j ); Among them, M j represents the set of job demand nodes, R j Represents the association relationship between nodes, which is constructed by constructing the edge set R j,p,q,r accomplish: Among them, sim(M j,p ,M j,q ,M j,r ) represents the similarity measure ζ between nodes p,q,r,u To adjust the parameters of the similarity measure, Δ p,q,r,u Represents the difference measure of a specific feature dimension, ω w is the weight parameter, ν j,p,q,r,w To correct the multidimensional interaction terms of the association relationship; S26. Dynamically adjust the job demand knowledge graph G according to the update cycle of market data j , update the node set M j and the edge set R j The weight value r in j,p,q,r .

5. According to claim 1, a method for dynamic professional ability and job matching education assessment based on knowledge graph is characterized in that: The S6 comprises the following sub-steps: S61. Based on the matching results between professional ability and job requirements, calculate the matching score between individual professional ability and target job requirements, with the score ranging from 0 to 1, where 1 indicates a perfect match and 0 indicates a mismatch; S62. Generate an education assessment report based on the matching score. The content of the education assessment report includes the strengths and weaknesses of the individual's current professional ability, the specific requirements of the target position, and the professional skills areas that need to be improved; S63. Combine the analysis results in the education evaluation report with the professional competence knowledge map G i Compare and identify the professional capabilities that need to be further improved, and generate individual career planning suggestions i , career planning suggestions include recommended courses to study, required skill certificates to obtain, and recommended practical projects to participate in: in, is the target professional ability characteristic value, is the characteristic value of the current professional ability, w k is the weight coefficient of each professional ability characteristic in career planning; S64. Provide career planning advice i Feedback is provided to educational institutions and individuals to guide them to make corresponding adjustments in the process of improving their professional capabilities. Educational institutions adjust their curriculum according to the suggestions, and individuals choose their learning paths and practical activities based on the suggestions. S65. Dynamically adjust the professional competence knowledge graph G based on feedback from educational institutions and individuals i and job requirements knowledge graph G j , and recalculate the matching score; S66. Through multiple feedback and optimization iterations, a closed-loop vocational education and job matching system will eventually be formed. The system will provide the best career planning suggestions and job matching solutions in real time based on changes in market demand and the development of individual professional capabilities.

Images