A talent information processing method and system based on professional quality and big data analysis

Abstract

The application discloses a talent information processing method and system based on professional quality and big data analysis, and relates to the technical field of talent information processing. The method comprises the following steps: integrating talent information from multiple channels through a multi-source data fusion algorithm to obtain a talent information set; extracting features from the information in the talent information set to identify unstructured features related to the professional quality of talents; determining a multi-level professional quality evaluation and prediction model based on the unstructured features related to the professional quality; determining a talent matching algorithm based on graph theory according to the output results of the multi-level professional quality evaluation and prediction model; wherein the talent matching algorithm is used to output a matching graph between talents and positions; and determining a talent allocation plan for a target object according to the matching graph between talents and positions output by the talent matching algorithm.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of talent information processing, in particular to a talent information processing method and system based on professional competence and big data analysis. BACKGROUND

[0002] With the advent of economic globalization and the information age, enterprises are facing increasing competition and challenges, and the importance of talent information processing is increasingly prominent. Talent information processing not only relates to whether an enterprise can efficiently attract and retain key talents, but also directly affects the operational efficiency and strategic decision-making of the enterprise.

[0003] Effective talent information processing helps to identify and prevent potential human resource risks such as employee turnover and skill gaps, thereby reducing the negative impact on business. And talent information processing can reveal new work patterns and talent trends, providing data support for enterprise innovation and business development. Existing talent information processing methods often rely on manual screening and matching, which is inefficient and easily affected by subjective factors.

[0004] In view of the above problems, no effective solutions have been proposed so far. SUMMARY

[0005] The embodiments of the present application provide a talent information processing method and system based on professional competence and big data analysis to solve the above technical problems.

[0006] The present application provides a talent information processing method based on professional competence and big data analysis, comprising: integrating talent information from multiple channels through a multi-source data fusion algorithm to obtain a talent information set;

[0007] Feature extraction is performed on the information in the talent information set to identify unstructured features related to the professional competence of talents;

[0008] Based on the unstructured features related to the professional competence, a multi-level professional competence evaluation and prediction model is determined;

[0009] According to the output result of the multi-level professional competence evaluation and prediction model, a talent matching algorithm based on graph theory is determined; wherein the talent matching algorithm is used to output a matching graph between talents and positions;

[0010] According to the matching graph between talents and positions output by the talent matching algorithm, a talent allocation plan is determined for a target object.

[0011] Further, the multi-level professional competence evaluation and prediction model based on the unstructured features related to the professional competence is configured to:

[0012] determine the individual comprehensive ability evaluation model, the occupation adaptability and development potential sub-model and the social network and influence evaluation sub-model based on the unstructured features related to the occupation literacy;

[0013] fuse the individual comprehensive ability evaluation model, the occupation adaptability and development potential sub-model and the social network and influence evaluation sub-model to obtain the multi-level occupation literacy evaluation and prediction model.

[0014] Further, the plurality of channels includes social media, online resumes and enterprise data.

[0015] The target object is a college, an enterprise or a unit.

[0016] The present application provides a talent information processing system based on occupation literacy and big data analysis, comprising: a talent information integration module for integrating talent information from multiple channels through a multi-source data fusion algorithm to obtain a talent information set;

[0017] An occupation literacy information extraction module is configured to extract features from the information in the talent information set to identify unstructured features related to the occupation literacy of the talent.

[0018] An occupation literacy evaluation and prediction module is configured to determine a multi-level occupation literacy evaluation and prediction model based on the unstructured features related to the occupation literacy.

[0019] A talent matching algorithm construction module is configured to determine a talent matching algorithm based on graph theory according to the output results of the multi-level occupation literacy evaluation and prediction model, wherein the talent matching algorithm is configured to output a matching graph between talents and positions.

[0020] A talent allocation plan determination unit is configured to determine a talent allocation plan for a target object according to the matching graph between talents and positions output by the talent matching algorithm.

[0021] Further, the occupation literacy evaluation and prediction model includes three levels of sub-models, namely an individual comprehensive ability evaluation model, an occupation adaptability and development potential sub-model and a social network and influence evaluation sub-model. The occupation literacy evaluation and prediction module is configured to determine a multi-level occupation literacy evaluation and prediction model based on unstructured features related to the occupation literacy.

[0022] determine the individual comprehensive ability evaluation model, the occupation adaptability and development potential sub-model and the social network and influence evaluation sub-model based on the unstructured features related to the occupation literacy;

[0023] The individual comprehensive ability evaluation model, the occupation adaptability and development potential sub-model, and the social network and influence evaluation sub-model are fused to obtain the multi-level occupation accomplishment evaluation and prediction model.

[0024] Based on the embodiments provided in the present application, talent information from multiple channels is integrated by a multi-source data fusion algorithm to obtain a talent information set; information in the talent information set is subjected to feature extraction to identify unstructured features related to the occupation accomplishment of the talent; based on the unstructured features related to the occupation accomplishment, a multi-level occupation accomplishment evaluation and prediction model is determined; based on the output result of the multi-level occupation accomplishment evaluation and prediction model, a talent matching algorithm based on graph theory is determined; wherein the talent matching algorithm is used to output a matching graph between talents and positions; based on the matching graph between talents and positions output by the talent matching algorithm, a talent allocation plan is determined for a target object. Thus, the processing of talent information based on occupation accomplishment and big data analysis is automatically realized, and the efficiency and accuracy of processing talent information are improved. Specifically, the following beneficial effects are achieved: through the multi-source data fusion algorithm, talent information from different channels can be integrated to form a comprehensive talent information set. This method helps to capture the all-around features of talents, and improves the richness and accuracy of talent information; through feature extraction on the talent information set, unstructured features related to occupation accomplishment are identified, which enables the evaluation model to more deeply understand the potential ability and quality of talents, thereby improving the accuracy and reliability of evaluation; based on the unstructured features, a multi-level occupation accomplishment evaluation and prediction model is established, which not only can evaluate the current occupation accomplishment of an individual, but also can predict the future development potential, providing a more long-term perspective for talent management; using the talent matching algorithm based on graph theory, the best match between talents and positions can be determined according to the output result of the multi-level evaluation and prediction model; this method helps to realize the precise matching of talents and positions, and improve the recruitment efficiency and employee satisfaction; the matching graph between talents and positions output by the talent matching algorithm provides a visual way for decision makers to understand and analyze talent allocation schemes, making the talent allocation process more transparent and easy to manage; based on the matching graph, a talent allocation plan can be determined for a target object, which helps enterprises to more effectively utilize talent resources, improve organizational efficiency, and provide positions more suitable for the career development path of talents; through multi-level evaluation, it helps enterprises to better understand the long-term development potential of employees, thereby designing more effective career development paths and retention strategies, enhancing the loyalty and satisfaction of employees. BRIEF DESCRIPTION OF DRAWINGS

[0025] The accompanying drawings, which are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and serve to explain the principles of the application. In the drawings:

[0026] Figure 1 An optional flow chart of a talent information processing method based on professional accomplishment and big data analysis according to an embodiment of the present application;

[0027] Figure 2 An optional structure diagram of a talent information processing system based on professional accomplishment and big data analysis according to an embodiment of the present application.

[0028] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely with reference to the accompanying drawings.

[0030] Optionally, as shown in the figure, Figure 1 The present application provides a talent information processing method based on professional accomplishment and big data analysis, comprising:

[0031] S101, integrating talent information from multiple channels through a multi-source data fusion algorithm to obtain a talent information set;

[0032] S102, extracting features from the information in the talent information set to identify unstructured features related to the professional accomplishment of the talent;

[0033] S103, determining a multi-level professional accomplishment evaluation and prediction model based on the unstructured features related to the professional accomplishment;

[0034] S104, determining a talent matching algorithm based on graph theory according to the output results of the multi-level professional accomplishment evaluation and prediction model; wherein the talent matching algorithm is used to output a matching graph between talents and positions;

[0035] S105, determining a talent allocation plan for a target object according to the matching graph between talents and positions output by the talent matching algorithm.

[0036] Based on the embodiments provided in the present application, the talent information from multiple channels is integrated through a multi-source data fusion algorithm to obtain a talent information set; the information in the talent information set is subjected to feature extraction to identify unstructured features related to the professional competence of the talent; based on the unstructured features related to the professional competence, a multi-level professional competence evaluation and prediction model is determined; based on the output result of the multi-level professional competence evaluation and prediction model, a talent matching algorithm based on graph theory is determined; wherein the talent matching algorithm is used to output a matching graph between the talent and the position; based on the matching graph between the talent and the position output by the talent matching algorithm, a talent allocation plan is determined for the target object. Thus, the processing of talent information based on professional competence and big data analysis is automatically realized, and the efficiency and accuracy of processing talent information are improved. Specifically, the following beneficial effects are achieved: through the multi-source data fusion algorithm, talent information from different channels can be integrated to form a comprehensive talent information set. This method helps to capture the all-around features of talents, improving the richness and accuracy of talent information; by extracting features from the talent information set, unstructured features related to professional competence are identified, which enables the evaluation model to better understand the potential ability and quality of the talent, thereby improving the accuracy and reliability of the evaluation; based on the unstructured features, a multi-level professional competence evaluation and prediction model is established, which not only evaluates the current professional competence of the individual, but also predicts the future development potential, providing a more long-term perspective for talent management; using the talent matching algorithm based on graph theory, the best match between the talent and the position can be determined based on the output result of the multi-level evaluation and prediction model; this method helps to achieve precise matching of talents and positions, improving recruitment efficiency and employee satisfaction; the matching graph between the talent and the position output by the talent matching algorithm provides a visual way for decision-makers to understand and analyze talent allocation schemes, making the talent allocation process more transparent and easy to manage; based on the matching graph, a talent allocation plan can be determined for the target object, which helps enterprises to more effectively utilize talent resources, improve organizational efficiency, and provide positions that are more suitable for the career development path of talents; through multi-level evaluation, it helps enterprises to better understand the long-term development potential of employees, thereby designing more effective career development paths and retention strategies to enhance employee loyalty and satisfaction.

[0037] Further, based on the unstructured features related to the professional competence, a multi-level professional competence evaluation and prediction model is determined, which is configured to:

[0038] Based on the unstructured features related to the professional competence, an individual comprehensive ability evaluation model, a professional adaptability and development potential sub-model, and a social network and influence evaluation sub-model are determined;

[0039] The individual comprehensive ability evaluation model, the occupation adaptability and development potential sub-model, and the social network and influence evaluation sub-model are fused to obtain a multi-level occupation accomplishment evaluation and prediction model.

[0040] Further, the multiple channels include social media, online resumes, and enterprise data.

[0041] The target object is a college, an enterprise, or a unit.

[0042] Optionally, as shown in the application, a talent information processing system based on occupation accomplishment and big data analysis is provided, which includes: Figure 2

[0043] The talent information integration module 201 is configured to integrate talent information from multiple channels through a multi-source data fusion algorithm to obtain a talent information set.

[0044] The occupation accomplishment information extraction module 202 is configured to extract features from the information in the talent information set to identify unstructured features related to the occupation accomplishment of the talent.

[0045] The occupation accomplishment evaluation and prediction module 203 is configured to determine a multi-level occupation accomplishment evaluation and prediction model based on the unstructured features related to the occupation accomplishment.

[0046] The talent matching algorithm construction module 204 is configured to determine a talent matching algorithm based on graph theory according to the output result of the multi-level occupation accomplishment evaluation and prediction model, wherein the talent matching algorithm is used to output a matching graph between talents and positions.

[0047] The talent allocation plan determination unit 205 is configured to determine a talent allocation plan for the target object according to the matching graph between talents and positions output by the talent matching algorithm.

[0048] Further, the occupation accomplishment evaluation and prediction model includes three levels of sub-models, namely, an individual comprehensive ability evaluation model, an occupation adaptability and development potential sub-model, and a social network and influence evaluation sub-model.

[0049] Based on the unstructured features related to the occupation accomplishment, the individual comprehensive ability evaluation model, the occupation adaptability and development potential sub-model, and the social network and influence evaluation sub-model are determined.

[0050] The individual comprehensive ability evaluation model, the occupation adaptability and development potential sub-model, and the social network and influence evaluation sub-model are fused to obtain a multi-level occupation accomplishment evaluation and prediction model.

[0051] ​Further, based on the unstructured features related to professional competence, an individual comprehensive ability assessment model, a professional adaptability and development potential sub-model, and a social network and influence assessment sub-model are determined, which are configured to:

[0052] based on the unstructured features related to professional competence train the individual comprehensive ability assessment model ; wherein, ; is a function combining machine learning and psychometrics; the output of the individual comprehensive ability assessment model is a score of the individual's ability in multiple dimensions; the multiple dimensions of ability include cognitive ability, emotional intelligence, and leadership;

[0053] based on the unstructured features related to professional competence and the output of the individual comprehensive ability assessment model, use time series analysis and reinforcement learning algorithms to simulate the individual's performance in a virtual professional path to assess the individual's professional adaptability and development potential in different professional environments; train the professional adaptability and development potential sub-model according to the individual's adaptability and development potential in different professional environments ;

[0054] wherein, ; is a function combining time series analysis and reinforcement learning algorithms; the professional adaptability and development potential sub-model is used to predict the individual's future growth trajectory and professional development potential;

[0055] based on the unstructured features related to professional competence and the individual's future growth trajectory and professional development potential predicted by the professional adaptability and development potential sub-model, combine social network analysis algorithms and natural language processing algorithms to analyze the individual's interaction patterns in social media and professional forums to determine the professional adaptability and development potential sub-model ;

[0056] wherein, ; is a function combining social network analysis algorithms and natural language processing algorithms; the professional adaptability and development potential sub-model is used to assess the individual's influence and social ability in the professional network to quantify the individual's social influence and network centrality;

[0057] based on the individual's influence and social ability in the professional network assessed by the professional adaptability and development potential sub-model, update the professional adaptability and development potential sub-model to obtain an updated professional adaptability and development potential sub-model ;

[0058]

[0059] wherein, is an update function for updating the career adaptability and development potential sub-model according to the individual's influence and social ability in the professional network.

[0060] Further, according to the output results of the multi-level career competence evaluation and prediction model, a talent matching algorithm based on graph theory is determined, which is configured to:

[0061] construct an initial graph model; wherein the initial graph model includes talent nodes and position nodes; the node features of each talent node include the output of the individual comprehensive ability evaluation model, the career adaptability and development potential sub-model, and the social network and influence evaluation sub-model; the node features of each position node include the position description, the required skills, and the social influence;

[0062] combine the use of term frequency-inverse document frequency algorithm and word embedding algorithm to calculate the first similarity between talent nodes and position nodes;

[0063] by analyzing the relationship between talent nodes and the correlation between position nodes, the second similarity of the edges between talent nodes and the third similarity of the edges between position nodes are analyzed;

[0064] according to the first similarity, the second similarity and the third similarity, the correlation matrix in graph matching is calculated;

[0065] combine the use of decomposition graph matching algorithm and probabilistic graph matching algorithm to construct talent matching algorithm according to the correlation matrix in graph matching;

[0066] the output results of the multi-level career competence evaluation and prediction model are taken as weighted factors in the talent matching algorithm.

[0067] Further, according to the first similarity, the second similarity and the third similarity, the correlation matrix in graph matching is calculated based on the following formula:

[0068]

[0069] wherein, is an element of the correlation matrix in graph matching, representing the comprehensive similarity between talent nodes and position nodes ; is the first similarity between talent nodes and position nodes ; is the second similarity between talent nodes and talent nodes ; is the third similarity between position nodes and position nodes a third similarity between the talent node , , are respectively the output of the individual comprehensive ability evaluation model, the occupation adaptability and development potential sub-model and the social network and influence evaluation sub-model of the talent node ; , , are respectively the job description, the required skill and the social influence feature vector of the job node ; , , are all weight coefficients.

[0070] Further, a first similarity between the talent node and the job node is calculated based on the following formula:

[0071]

[0072] wherein, is the similarity calculated based on the term frequency-inverse document frequency algorithm; is the similarity calculated based on the word embedding algorithm; , are respectively the output of the individual comprehensive ability evaluation model and the occupation adaptability and development potential sub-model of the talent node ; , are respectively the job description and the required skill feature vector of the job node ; , , are all weight coefficients.

[0073] Further, the multiple channels include social media, online resumes and enterprise data; and the target object is a college, an enterprise or a unit.

[0074] It should be noted that, in the present application, the embodiments implemented by the talent information processing system based on occupation literacy and big data analysis can be mutually referenced with the embodiments implemented by the talent information processing method based on occupation literacy and big data analysis, and the present application will not be repeated here.

[0075] The above are only preferred embodiments of the present application, and do not limit the patent scope of the present application, and any equivalent structure or equivalent process transformation based on the content of the present application specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A talent information processing method based on professional qualities and big data analysis, characterized in that, include: Talent information from multiple sources is integrated using a multi-source data fusion algorithm to obtain a talent information set; Feature extraction is performed on the information in the talent information set to identify unstructured features related to the professional qualities of the talent; Based on the unstructured characteristics related to professional competence, a multi-level professional competence assessment and prediction model is determined. It is configured as follows: An individual comprehensive ability assessment model Y is obtained by training on unstructured features X related to professional competence; where Y = f1(X); f1 is a function combining machine learning and psychometrics; a career adaptability and development potential sub-model Z is obtained by training on the individual's adaptability and development potential in different occupational environments; where Z = f2(X,Y); f2 is a function combining time series analysis and reinforcement learning algorithms; based on the unstructured features X related to professional competence and the individual's future growth trajectory and career development potential predicted by the career adaptability and development potential sub-model, the interaction patterns of the individual in social media and professional forums are analyzed using social network analysis algorithms and natural language processing algorithms to determine the career adaptability and development potential sub-model W; where W = f3(X,Z); f3 is a function combining social network analysis algorithms and natural language processing algorithms; based on the individual's influence and social skills in professional networks assessed by the career adaptability and development potential sub-model, the career adaptability and development potential sub-model is updated to obtain the updated career adaptability and development potential sub-model Y. ' ;Y ' =f4(W,Y) Wherein, f4 is an update function used to update the career adaptability and development potential sub-model based on an individual's influence and social skills in the professional network; based on the output of the multi-level career competence assessment and prediction model, a graph theory-based talent matching algorithm is determined; wherein, the talent matching algorithm is used to output a matching graph between talent and position; By combining decomposition graph matching and probabilistic graph matching algorithms, a talent matching algorithm is constructed based on the association matrix in graph matching. The association matrix in graph matching is calculated based on the following formula, using the first similarity, second similarity, and third similarity: K i,j =S1(i,j)+α×[S 2_talent (i,k)+S 2_position (j,l)]+β×Y ' i ×D i +γ×Z i ×Q i +δ×W i ×IMP i Among them, K i,j S1(i,j) is an element of the association matrix in graph matching, representing the comprehensive similarity between talent node i and job node j; S1(i,j) is the first similarity between talent node i and job node j; S 2_talent (i,k) is the second similarity between talent node i and talent node k; S 2_position (j,l) is the third similarity between job node j and job node l; Y ' i Z i W i These are the outputs of the individual comprehensive ability assessment model for talent node i, the career adaptability and development potential sub-model, and the social network and influence assessment sub-model; D i Q i IMP i These are the feature vectors of job description, required skills, and social influence for job node j; α, β, γ, and δ are all weight coefficients; based on the matching graph between talent and job output by the talent matching algorithm, a talent allocation plan is determined for the target object.

2. The talent information processing method based on professional qualities and big data analysis according to claim 1, characterized in that, The multiple channels include social media, online resumes, and corporate data; The target groups are universities, enterprises, or organizations.

3. A talent information processing system based on professional competence and big data analysis, wherein the system implements the talent information processing method based on professional competence and big data analysis as described in claim 1, characterized in that, include: The talent information integration module is used to integrate talent information from multiple channels through a multi-source data fusion algorithm to obtain a talent information set; The professional competence information extraction module is used to extract features from the information in the talent information set in order to identify unstructured features related to the professional competence of talents. The professional competence assessment and prediction module is used to determine multi-level professional competence assessment and prediction models based on unstructured characteristics related to professional competence. The talent matching algorithm construction module is used to determine a graph-based talent matching algorithm based on the output results of a multi-level professional competence assessment and prediction model; wherein, the talent matching algorithm is used to output a matching graph between talents and positions; The talent allocation plan determination unit is used to determine a talent allocation plan for the target object based on the talent-position matching graph output by the talent matching algorithm.

4. The talent information processing system based on professional qualities and big data analysis according to claim 3, characterized in that, Based on the output results of the multi-level professional competence assessment and prediction model, the graph theory-based talent matching algorithm is configured as follows: Construct an initial graph model; wherein the initial graph model includes talent nodes and job nodes; the node features of each talent node include the outputs of the individual comprehensive ability assessment model, the career adaptability and development potential sub-model, and the social network and influence assessment sub-model; the node features of each job node include job description, required skills, and social influence; The first similarity between the talent node and the job node is calculated by combining the term frequency-inverse document frequency algorithm with the word embedding algorithm.

Images