Entrepreneurial team establishment recommendation method based on skill map matching

Abstract

The invention relates to the technical field of skill atlas matching, and discloses a skill atlas matching-based entrepreneurial team establishment recommendation method, which comprises the following steps of S1, constructing a dynamic skill atlas, extracting skill labels from a public data source and user input, defining skill nodes, relationships among the nodes and association weights, and monitoring technical trend data in real time to obtain a dynamic skill atlas; dynamically updating nodes and weights of the skill atlas; s2, analyzing entrepreneurship project demands, and analyzing project description texts; s3, evaluating skills of candidate members; s4, generating a team multi-recommendation scheme; and S5, scheme comparative analysis and selection. Through a multi-objective optimization algorithm, at least three team establishment recommendation schemes are generated, diversified entrepreneurship demands are covered, scheme selection in different directions is realized, and meanwhile, an entrepreneurship person can select the most suitable team establishment mode from a plurality of high-quality schemes according to the actual condition of a project, so that the flexibility and decision space are increased.

Description

Technical Field

[0001] The present invention relates to the technical field of skill graph matching, and specifically to a recommendation method for startup team formation based on skill graph matching. Background Art

[0002] In the current startup wave, the market environment is complex and changeable, and innovative projects are emerging continuously. The formation of startup teams is crucial. As a tool that can effectively integrate and present skill information, a skill graph can clearly display various skills and their relationships. Forming a startup team based on skill graph matching, docking project requirements with members' skills, injecting suitable talents into the startup team, enhancing the team's competitiveness, and helping the startup project set sail successfully.

[0003] However, in the current technology, traditional startup team formation methods are mostly based on limited experience and information, determining a team formation model based on past cases or intuition, which is difficult to form multiple recommendation schemes in terms of multi-dimensional focuses, reducing the flexibility of the formation scheme and the decision-making space. Summary of the Invention

[0004] Aiming at the deficiencies of the existing technology, the present invention provides a recommendation method for startup team formation based on skill graph matching, which solves the problem of difficult to form multiple recommendation schemes in terms of multi-dimensional focuses, reducing the flexibility of the formation scheme and the decision-making space.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: In the first aspect, a recommendation method for startup team formation based on skill graph matching includes the following steps: S1. Construct a dynamic skill graph, extract skill tags, define skill nodes, relationships between nodes and association weights from public data sources and user inputs, and dynamically update the nodes and weights of the skill graph by real-time monitoring of technology trend data; S2. Analyze the requirements of startup projects, parse the project description text, extract core skills and auxiliary skills, and assign priority weights; S3. Evaluate the skills of candidate members, and quantify the scores of explicit skills and implicit skills based on resume data and collaboration behavior logs; S4. Generate multiple team recommendation schemes, search for subgraphs in the skill graph that match the project requirements based on graph neural networks, and generate at least three recommendation suggestions through multi-objective optimization algorithms; S5. Compare and analyze the schemes and make a selection, display the comparison of the quantitative indicators of each scheme through a visual interface, and generate a summary of the advantages and disadvantages of each scheme.

[0006] Preferably, the construction of the dynamic skill graph specifically includes the following steps: S101. New skill tags through user input and public data such as online recruitment websites; S102. Use large language models to automatically identify the association between new skills and existing graph nodes; S103. Trigger graph weight adjustment according to the growth rate of skill search volume.

[0007] Preferably, the specific steps of the S2 analysis of startup project requirements are as follows: S201. Identify and extract skill keywords in the project description text through natural language processing. The project description text includes project plans, business proposals, and requirement documents; S202. Extract skill modification relationships through dependency syntax analysis, disambiguate the fuzzy expressions of skills, and then match the preset skill templates based on industry classification; S203. Automatically fill in the weights of core skills and auxiliary skills through industry benchmark data, and dynamically adjust the threshold according to the cost scale.

[0008] Preferably, the evaluation of candidate members' skills includes the following steps: S301. Score based on the candidate's resume, position, and skill field by years of work experience; S302. Collect the interaction data of the candidate in historical project collaborations, calculate the response speed and task completion rate indicators, and map the execution and leadership labels and incorporate them into the skill graph.

[0009] Preferably, in S4, a configurable set of optimization goals is defined, including: Minimum cost, minimizing the total team salary; Highest efficiency, minimizing the average delivery cycle of historical similar projects; Highest quality, maximizing the average skill score and project experience value of members; Generate a Pareto front solution set through a multi-objective optimization algorithm, and select the extreme solutions in each objective direction as recommended suggestions.

[0010] Preferably, in S4, the weights of each objective can be customized and adjusted to generate a hybrid optimization plan. Currently, the weights include cost, time, and quality.

[0011] Preferably, the specific steps of the program comparison analysis and selection are as follows: S501. Establish a visual tree diagram to compare the costs, times, and qualities of each program; S502. Based on the tree diagram, output the summary of the advantages and disadvantages of each program in the quantitative indicators, including project cost control, project completion time, and project quality; S503. Through the selected program, distribute the matched candidate components to form a startup team.

[0012] In a second aspect, a startup team formation recommendation system based on skill graph matching, the system includes: A dynamic skill graph construction module, which is used to extract skill tags, define skill nodes and association weights from public data sources and user inputs, identify associations through recruitment websites using large language models, and dynamically update the graph based on the search volume growth rate; A startup project requirement analysis module, which is used to parse the project description text, extract skill keywords, eliminate ambiguity through dependency syntax analysis and match preset templates, and fill in skill weights and adjust thresholds according to industry data; A candidate member skill evaluation module, which is used to evaluate the skills of candidates and quantify explicit and implicit skills; A team recommendation plan generation module, which is used to search for matching subgraphs through a graph neural network, define a set of cost, efficiency, and quality goals, obtain a Pareto front solution set through a multi-objective optimization algorithm, select extreme solutions for recommendation, and generate a hybrid plan by customizing objective weights; A plan comparison analysis and selection module, which is used to establish a visual tree diagram to compare the quantitative indicators of each plan, output a summary of advantages and disadvantages, and issue candidates to form a team through the user's selected plan.

[0013] In a third aspect, the invention provides the following technical solution, a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the computer program, it implements the above-mentioned startup team formation recommendation method based on skill graph matching.

[0014] In a fourth aspect, the invention provides the following technical solution, a readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the above-mentioned startup team formation recommendation method based on skill graph matching.

[0015] The invention provides a startup team formation recommendation method based on skill graph matching. It has the following beneficial effects: 1. Through a multi-objective optimization algorithm, the invention generates at least three team formation recommendation plans, covering diverse startup needs, realizing plan selection in different directions. At the same time, entrepreneurs can select the most suitable team formation method from multiple high-quality plans according to the actual situation of the project, thus increasing flexibility and decision-making space.

[0016] 2. By collecting skill tags from public data sources, combining user inputs to improve information, defining skill nodes, associations, and weights, and tracking technology trends in real time to dynamically update the skill graph, making it conform to industry changes, the invention realizes dynamic update of the skill framework and improves the accuracy and adaptability of the recommendation plan in different industry scenarios.

[0017] 3. The present invention comprehensively combines resumes and collaboration behavior logs to quantitatively evaluate the skills of candidate members. Explicit skills are extracted from resumes, and implicit skills are mined from logs for multi-modal scoring, providing data support for generating team recommendation plans, ensuring that the recommended members meet the project skill requirements, improving the overall skill level and collaboration potential of the team, and facilitating the formation of a high-quality team suitable for the project. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flowchart of the method for recommending the formation of a startup team based on skill graph matching according to the present invention; Figure 2 is an architecture diagram of the system for recommending the formation of a startup team based on skill graph matching according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] Embodiment 1: Referring to Figure 1 , in the first embodiment of the present invention, the present invention provides a method for recommending the formation of a startup team based on skill graph matching, including the following steps: S1. Construct a dynamic skill graph, extract skill tags, define skill nodes, relationships between nodes and association weights from public data sources and user inputs, and dynamically update the nodes and weights of the skill graph by real-time monitoring of technology trend data; S2. Analyze the requirements of the startup project, parse the project description text, extract core skills and auxiliary skills, and assign priority weights; S3. Evaluate the skills of candidate members, and quantitatively score explicit skills and implicit skills based on resume data and collaboration behavior logs; S4. Generate multiple team recommendation plans, search for subgraphs in the skill graph that match the project requirements based on a graph neural network, and generate at least three recommendation suggestions through a multi-objective optimization algorithm; S5. Compare and analyze the plans and make a selection, display the comparison of quantitative indicators of each plan through a visualization interface, and generate a summary of the advantages and disadvantages of each plan.

[0021] Specifically, through S1, skill tags are widely collected from public data sources, supplemented with characteristic information by integrating user inputs, the skill nodes and their associations are defined, and weights are assigned to measure the importance. By tracking the technical trend data in real time, it ensures that the skill map always fits the changing industry environment, providing a dynamic and rich skill framework for the subsequent steps, thereby improving the accuracy and adaptability of the recommendation scheme; Through S2, interpret the description text of the startup project, extract the core skills on which the smooth progress of the project depends and the skills that play an auxiliary supporting role, and assign priority weights according to the importance of the skills to the project, clarify the focus and level of the project's skill requirements, and transform the vague project requirements into a specific and measurable skill requirement system, providing a clear and highly targeted reference standard for the subsequent screening and matching of candidate members; Through S3, comprehensively utilize resume data and collaboration behavior logs to conduct a comprehensive and in-depth quantitative assessment of the skills of candidate members, extract the explicit skills presented in the resume and the implicit skills reflected in the collaboration behavior logs, and achieve multi-modal scoring analysis, providing a data basis of technical personnel for the subsequent generation of team recommendation schemes, ensuring that the recommended members have the ability to meet the project requirements, and enhancing the overall skill strength and collaboration potential of the team; Through S4, with the help of a graph neural network, find the subgraph that matches the project requirements in the skill map, and then use a multi-objective optimization algorithm to generate at least three team formation recommendation schemes, providing diverse choices for the startup team formation, achieving different degrees of balance in different schemes, and entrepreneurs can select the most suitable team formation method from multiple high-quality schemes according to the actual situation of the project, increasing flexibility and decision-making space; Through S5, the visualization interface intuitively shows the differences between the schemes in quantitative indicators, helps decision-makers clearly understand the characteristics of each scheme at a glance, and generates a summary of the advantages and disadvantages of each scheme, providing a reference basis for the final decision-making, thereby simplifying the process of scheme comparison and selection, accurately selecting the team formation scheme that best meets the needs and expected goals of the startup project, and improving the decision-making efficiency and quality.

[0022] The construction of the dynamic skill map specifically includes the following steps: S101. Obtain new skill tags through user inputs and public data such as online recruitment websites; S102. Use a large language model to automatically identify the association relationships between new skills and existing map nodes; S103. Trigger the adjustment of the map weights according to the growth rate of skill search volume.

[0023] Specifically, by reflecting the current enterprise employment skill requirements through the recruitment website in S101 and capturing new skill tags, the cutting-edge dynamics in the technology field and the actual market demand can be captured in a timely manner, new skill information can be actively obtained, the data pool of the skill map can be expanded, ensuring that the content of the skill map keeps up with the industry development rhythm and does not miss the skill changes brought by emerging technologies, thereby broadening the coverage of the skill map; Through S102, the large language model (LLM) sorts out the newly captured skill tags, thereby connecting the new skills with the map nodes. For example, whether the new skills are expansions, fusions, or completely new derivatives based on existing skills, so as to reasonably embed the new skills into the existing skill knowledge network, improve the structural system of the skill map, and make the newly added skills and the original skills form an organic whole; Through S103, the growth rate of skill search volume is used to adjust the map weight to reflect the dynamic measurement of the importance and popularity of skills, thereby giving higher weights to skills and reducing them conversely, ensuring that the skill map is focused and in line with the current technology and market development trends, and improving the accuracy and efficiency of matching based on the skill map.

[0024] The analysis of startup project requirements in S2 specifically includes the following steps: S201: Identify and extract skill keywords in the project description text through natural language processing. The project description text includes project plans, business proposals, and requirement documents; S202: Extract skill modification relationships through dependency syntax analysis, disambiguate the ambiguous expressions of skills, and then match the preset skill templates based on industry classification; S203: Automatically fill in the weights of core skills and auxiliary skills through industry benchmark data, and dynamically adjust the threshold according to the cost scale.

[0025] Specifically, through natural language processing in S201, quickly locate and extract skill keywords from complex document contents such as project plans, transform complex project descriptions into clear skill elements, and thus sort out a detailed list of skill keywords required for the project, providing a reliable data starting point for the entire project requirement analysis process; By mining the modification relationships between skills in S202, clarify the specific action methods and interrelationships of skills in the project context. For ambiguous skill expressions that may exist in the text, through disambiguation processing, such as mapping "familiar with cloud services" to "basic operation and maintenance of AWS / Azure / GCP", eliminate the uncertainty in understanding, and make the skill meaning clear and definite; The industry benchmark data in S203 is used to reflect the general importance of various skills in different industries in a project, and these data are used to automatically fill in the weights of core skills and auxiliary skills, providing an initial standard based on industry commonalities for the quantitative assessment of project skill requirements, making the weight setting have an objective basis. At the same time, considering the huge differences in cost scales among different startup projects, dynamically adjusting the threshold according to the cost scale can flexibly incorporate the personalized cost factors of the project into the weight allocation process, balancing skill importance and project resource constraints.

[0026] Evaluating the skills of candidate members includes the following steps: S301. Score based on the candidate's resume, position, and skill field by working years; S302. Collect the interaction data of the candidate in historical project collaborations, calculate the response speed and task completion rate indicators, and map the execution and leadership labels and incorporate them into the skill map.

[0027] Specifically, the candidate's resume in S301 can present their past experiences, the position reflects their scope of responsibilities, the skill field clarifies the professional direction, and the working years can measure the accumulation and proficiency of skills to a certain extent, giving a preliminary quantitative score to the candidate's ability in a specific skill field, being able to quickly distinguish the candidate's explicit skill level and laying a foundation for subsequent in-depth evaluation; The historical project collaboration interaction data in S302 contains the candidate's actual work behavior performance. The response speed reflects the candidate's immediate feedback ability to task arrangements and team communication, and the task completion rate reflects the effectiveness of their task execution. Mapping these indicators to execution and leadership labels and incorporating them into the skill map enriches the evaluation dimensions of the candidate's implicit skills and makes up for the deficiency of evaluating solely based on static information such as resumes.

[0028] In S4, a set of configurable optimization goals is defined, including: Minimum cost, minimizing the total team salary; Highest efficiency, minimizing the average delivery cycle of historical similar projects; Highest quality, maximizing the average skill score and project experience value of members; Generate a Pareto front solution set through a multi-objective optimization algorithm, and select the extreme solutions in each objective direction as recommended suggestions.

[0029] Specifically, defining a set of configurable optimization goals provides a clear and adjustable optimization direction for the formation of a startup team. The lowest cost focuses on capital investment. By minimizing the total team salary, it helps the startup project form a team with a limited budget. The highest efficiency takes the delivery cycle of past similar projects as a reference, aiming to shorten the project execution time so that the project can be launched to the market faster and gain the first-mover advantage. The highest quality emphasizes the professional qualities of team members, maximizing the average skill score and project experience value of members to ensure the quality and level of project completion.

[0030] In S4, the weights of each goal can be customized and adjusted to generate a hybrid optimization plan. Currently, the weights include cost, time, and quality.

[0031] Specifically, customizing and adjusting the goal weights such as cost, time, and quality to generate a hybrid optimization plan allows the startup team to flexibly change the importance of each goal in the consideration of team formation according to the unique needs and strategic priorities of the project, thereby launching a personalized team formation plan that suits the actual situation of the project, avoiding ignoring project characteristics due to a unified standard, helping the team find the most suitable balance among different goals, and enhancing the adaptability of team formation to the project.

[0032] The comparison analysis and selection of the plan specifically include the following steps: S501. Establish a visual tree diagram to compare the costs, time, and quality of each plan; S502. Based on the tree diagram, output a summary of the advantages and disadvantages of each plan in the quantitative indicators, including project cost control, project completion time, and project quality; S503. Through the selected plan, allocate the matched candidate components to form a startup team.

[0033] Specifically, through the visual tree diagram in S501, the cost, time, and quality data are presented in an intuitive and well-organized graph to show the performance of each plan in different indicators, helping decision-makers quickly capture the differences between each plan and providing a visual basis for subsequent in-depth analysis; Through S502, summarize the information in the tree diagram, analyze and output the advantages and disadvantages of each plan in terms of project cost control, completion time, and quality in text form, convert the visual data into clear text descriptions, highlight the key features of each plan, and help decision-makers quickly understand the core value and potential problems of each plan; Through S503, determine the most suitable team formation plan for the startup project, implement the plan, transfer the candidate information that matches the plan to relevant personnel, start the startup team formation process, and complete the closed-loop from requirements analysis to team building, ensuring that the project can be quickly promoted according to the selected plan.

[0034] Example 2: Refer toFigure 2 , in the second embodiment of the present invention, the present invention provides a startup team formation recommendation system based on skill graph matching. The system includes: A dynamic skill graph construction module, which is used to extract skill tags, define skill nodes and association weights from public data sources and user inputs, identify associations using a large language model through recruitment websites, and dynamically update the graph based on the search volume growth rate; A startup project requirement analysis module, which is used to parse the project description text, extract skill keywords, eliminate ambiguity through dependency syntax analysis and match preset templates, and fill in skill weights and adjust thresholds according to industry data; A candidate member skill evaluation module, which is used to evaluate the skills of candidates and quantify explicit and implicit skills; A team recommendation plan generation module, which is used to search for and match subgraphs through a graph neural network, define a set of cost, efficiency, and quality objectives, and obtain a Pareto front solution set through a multi-objective optimization algorithm, select extreme solutions for recommendation, and generate a hybrid plan by customizing objective weights; A plan comparison analysis and selection module, which is used to establish a visual tree diagram to compare the quantitative indicators of each plan, output a summary of advantages and disadvantages, and issue candidates to form a team based on the user's selected plan.

[0035] Specifically, the dynamic skill graph construction module widely accepts public and user input information, uses a large language model to sort out skill associations, dynamically updates with search popularity, and constructs a real-time and comprehensive skill knowledge network, laying a solid foundation for subsequent processes. Then, the startup project requirement analysis module deeply digs into the project description text, accurately extracts skill keywords, eliminates ambiguity and matches industry templates, and scientifically sets skill weights according to industry data, concretizing project requirements into a clear skill requirement system. The candidate member skill evaluation module integrates resumes and collaboration logs, comprehensively quantifies the explicit and implicit skills of candidates, provides three-dimensional member skill data, and uses the team recommendation plan generation module to match subgraphs through a graph neural network, generating a Pareto front solution set for multi-objective optimization around cost, efficiency, and quality, giving diverse recommendations and a hybrid plan with customized weights. Finally, the plan comparison analysis and selection module intuitively presents the quantitative indicators of each plan with a visual tree diagram, outputs a summary of advantages and disadvantages to assist in decision-making, and accurately issues candidates to form a team after selecting a plan, thereby improving the scientificity and efficiency of startup team formation.

[0036] Embodiment Three: In the third embodiment of the present invention, based on the same inventive concept, the present invention proposes a computer-readable storage medium storing a computer program, which when executed by a processor implements the steps of the startup team formation recommendation method based on skill graph matching in the above embodiment.

[0037] Embodiment Four: The fourth embodiment of the present invention. Based on the same inventive concept, a computer proposed by the present invention includes: a processor and a memory; the processor and the memory communicate with each other; the memory is used to store instructions; the processor is used to execute the instructions in the memory to execute the method for recommending the formation of an entrepreneurial team based on skill map matching in the above embodiments.

[0038] It should be understood that each part of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in the memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following technologies well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logic functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0039] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A startup team formation recommendation method based on skill graph matching, characterized in that, It includes the following steps: S1. Construct a dynamic skill graph, extract skill tags, define skill nodes, relationships between nodes and association weights from public data sources and user inputs, and dynamically update the nodes and weights of the skill graph by real-time monitoring of technology trend data; S2. Analyze the requirements of startup projects, parse the project description text, extract core skills and auxiliary skills, and assign priority weights; S3. Evaluate the skills of candidate members, and quantify the scores of explicit skills and implicit skills based on resume data and collaboration behavior logs; S4. Generate multiple team recommendation plans, search for subgraphs in the skill graph that match the project requirements based on graph neural networks, and generate at least three recommendation suggestions through multi-objective optimization algorithms; S5. Compare and analyze the plans and make a selection, display the comparison of quantitative indicators of each plan through a visual interface, and generate a summary of the advantages and disadvantages of each plan.

2. The startup team formation recommendation method based on skill graph matching according to claim 1, wherein The construction of the dynamic skill graph specifically includes the following steps: S101. Obtain new skill tags from user inputs and public data such as online recruitment websites; S102. Use large language models to automatically identify the association relationships between new skills and existing graph nodes; S103. Trigger graph weight adjustment according to the growth rate of skill search volume.

3. The method for recommending the formation of an entrepreneurial team based on skill graph matching according to claim 1, wherein The S2 analysis of startup project requirements specifically includes the following steps: S201. Identify and extract skill keywords in the project description text through natural language processing. The project description text includes project plans, business proposals, and requirement documents; S202. Extract skill modification relationships through dependency syntactic analysis, disambiguate ambiguous expressions of skills, and then match preset skill templates based on industry classification; S203. Automatically fill in the weights of core skills and auxiliary skills through industry benchmark data, and dynamically adjust the thresholds according to the cost scale.

4. The startup team formation recommendation method based on skill graph matching according to claim 1, characterized in that The evaluation of candidate members' skills includes the following steps: S301. Score based on the candidate's resume, position, and skill field by working years; S302. Collect the interaction data of candidates in historical project collaborations, calculate response speed and task completion rate indicators, and map execution and leadership labels and incorporate them into the skill graph.

5. The method for recommending the formation of an entrepreneurial team based on skill graph matching according to claim 1, wherein In S4, a configurable set of optimization objectives is defined, including: Lowest cost, minimizing the total team salary; Highest efficiency, minimizing the average delivery cycle of historical similar projects; Highest quality, maximizing the average skill score and project experience value of members; Generate a Pareto front solution set through the multi-objective optimization algorithm, and select the extreme solutions in each objective direction as recommendation suggestions.

6. The startup team formation recommendation method based on skill graph matching according to claim 1, wherein In S4, the weights of each objective can be custom-adjusted to generate a hybrid optimization plan. Currently, the weights include cost, time, and quality.

7. The method for recommending the formation of an entrepreneurial team based on skill graph matching according to claim 1, wherein The comparison, analysis, and selection of the plans specifically include the following steps: S501. Establish a visual tree diagram to compare the cost, time, and quality of each plan; S502. Based on the tree diagram, output a summary of the advantages and disadvantages of each plan in quantitative indicators, including project cost control, project completion time, and project quality; S503. Through the selected plan, assign the matched candidates to form a startup team.

8. A startup team formation recommendation system based on skill graph matching, characterized in that, For the startup team formation recommendation method based on skill graph matching according to any one of claims 1-7, the system includes: A dynamic skill graph construction module, which is used to extract skill tags, define skill nodes and association weights from public data sources and user inputs, identify associations through recruitment websites using large language models, and dynamically update the graph based on the search volume growth rate; An entrepreneurial project requirement analysis module, which is used to parse the project description text, extract skill keywords, eliminate ambiguity through dependency syntactic analysis and match preset templates, and fill in skill weights and adjust thresholds according to industry data; A candidate member skill evaluation module, which is used to evaluate the skills of candidates and quantify explicit and implicit skills; A team recommendation solution generation module, which is used to search and match subgraphs through a graph neural network, define a set of cost, efficiency, and quality objectives, obtain a Pareto front solution set through a multi-objective optimization algorithm, select extreme solutions for recommendation, and generate a hybrid solution by customizing objective weights; A solution comparison analysis and selection module, which is used to establish a visual tree diagram to compare the quantitative indicators of each solution, output a summary of advantages and disadvantages, and issue candidates to form a team according to the user's selected solution.

9. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the entrepreneurial team formation recommendation method based on skill graph matching according to any one of claims 1 to 7.

10. A readable storage medium, characterized in that, A computer program is stored on the readable storage medium, and when the computer program is executed by the processor, it implements the entrepreneurial team formation recommendation method based on skill graph matching according to any one of claims 1 to 7.

Images