A method for constructing an organizational team model

Abstract

This invention provides a method for constructing an organizational team model, comprising: collecting and preprocessing organizational employee data; inputting the preprocessed data into a pre-trained large language model for parsing and extracting core value dimensions and constraints, quantifying them, and constructing a constraint field; generating a large number of initial role sets from a skill base, evaluating the survivability of each initial role set in the constraint field, and performing role evolution to form a deterministic role set and corresponding role vectors; constructing a personal capability graph centered on employees and with various skills and cognitive patterns as nodes, and inputting role vectors into the personal capability graph for role matching catalysis to form a categorized role set; using all employees and categorized roles as network nodes, calculating the connection strength between them through collaboration strength and synergy potential, and outputting the optimal organizational team structure. This technical solution effectively quantifies the long-term impact of team configuration on organizational order and efficiency, providing interpretable long-term value predictions for organizational decision-making.

Description

Technical Field

[0001] This invention relates to the field of algorithm model technology, and specifically to a method for constructing an organizational team model. Technical Background

[0002] In modern enterprise management, how to efficiently and scientifically build teams and allocate talent has always been a core challenge for organizational development and human resource management. As enterprises expand and business complexity increases, traditional talent allocation methods relying on managerial experience or simple rule-based matching are no longer sufficient to meet dynamically changing market demands and organizational strategic goals. Existing talent management systems or team-building tools are mostly based on static skill tags and historical performance for matching, lacking comprehensive consideration of organizational strategy, dynamic role evolution, team synergy potential, and long-term organizational effectiveness. Furthermore, because they are generally based on pre-set, fixed job roles, they cannot dynamically generate or optimize role definitions and skills according to changes in organizational strategy and adjustments to business goals, making it difficult to adapt to rapidly changing business environments. In addition, existing talent structure assessment methods often focus on single or a few dimensions such as skill data, failing to comprehensively consider multi-dimensional information such as employees' cognitive patterns, potential, and behavioral characteristics, and also lacking a systematic assessment of the overall team synergy.

[0003] Technical content

[0004] The present invention aims to solve at least one of the above-mentioned technical problems.

[0005] To address the above problems, this invention provides a method for constructing an organizational team model, comprising:

[0006] Collect and preprocess organizational employee data, including organizational evolution data, skills data, cognitive data, behavioral data, and outcome data.

[0007] The preprocessed data is input into the pre-trained large language model for parsing, and the core value dimensions and constraints are extracted, quantified, and then a constraint field is constructed.

[0008] A large number of initial character sets are generated from the skill library. The survivability of each initial character set in the constraint field is evaluated and the characters are evolved to form a deterministic character set and the corresponding character vector.

[0009] Construct an individual competency graph centered on employees and with various skills and cognitive patterns as nodes. Input the role vectors into the individual competency graph for role matching catalysis to form a set of idiosyncratic roles. Using all employees and the idiosyncratic roles as network nodes, calculate the connection strength between them through collaboration intensity and synergy potential, and output the optimal organizational team structure.

[0010] Preferably, the construction method further includes predictive evaluation of the output organizational team architecture:

[0011] The output organizational team architecture is deployed to the organizational network, and the current organizational entropy is calculated; the future organizational entropy is calculated by evolving the future organizational network over time; and the entropy reduction effect is calculated.

[0012] Based on the micro-matching degree, a comprehensive feature vector is constructed and the long-term value is predicted, and a comprehensive evaluation report is output.

[0013] Preferably, the step of inputting the preprocessed data into the pre-trained large language model for parsing, extracting the core value dimensions and constraints, and then quantifying them to construct a constraint field includes:

[0014] Each dimension defines a coordinate axis position P(R), and each constraint condition C i Define the boundary function B in the constrained field. i (P(R)), the closer the function value is to 0, the more the constraint is satisfied;

[0015] Construct a time feature vector of urgency.

[0016] Preferably, the step of generating a large number of initial character sets from the skill library and evaluating the survivability of each initial character set in the constrained field includes:

[0017] A character is represented by a skill probability amplitude vector Ψ: Ψ=[ψ1,ψ2,...,ψ M ] T During initialization, the probability amplitude of all skills is set to be equal.

[0018] Given a character Ψ that collapses into a classic character set R, the probability that skill i is included in the character set R is |ψ|. i | 2 ;

[0019] A quantum role set Ψ is generated by connecting the constraint field and the role set R through a fitness function F(R). R .

[0020] Preferably, the fitness function is: F(R) = α*Innovation(R) + β*Feasibility(R,Orgs) + γ*Complementarity(R,Orgs) + δ*Cohesion(R), where:

[0021] Innovation(R): The novelty score of role R, which is calculated by determining the Jaccard dissimilarity of its skill set to all positions in the organization's history.

[0022] Shills R and Shills H It is a One-Hot vector representing the skill set of character R and historical character H;

[0023] Feasibility(R,Orgs), a feasibility score, is calculated based on the organization's existing skill set to determine the coverage ratio of skills required to achieve role R.

[0024] Complementarity(R,Orgs): Complementarity score, which calculates the inverse cosine similarity between the skill set of R and the existing skill set of the organization;

[0025] Cohesion(R): Calculates the set of skills R within a character in the skill compatibility graph G. Sill Average connection strength in;

[0026] α, β, γ, δ are adjustable weight parameters determined by strategic value and organizational evolution data.

[0027] Preferably, the character evolution improves the desired fitness and selects through a gradient algorithm, undergoing multiple rounds of iterative evolution, including:

[0028] Select two parent quantum states ψ p1 and ψ p2 offspring quantum state ψ offspring It is generated by the linear superposition of their probability amplitudes;

[0029] With a small probability P mutation Randomly select several skill components ψ from the character's quantum state Ψ. i Perform a phase rotation θ on it to change the probability amplitude of the skill, and calculate the fitness gradient of the current amplitude;

[0030] Using gradient information, the probability amplitude of Ψ is fine-tuned; after each generation of evolution, N with the highest expected fitness is retained. elte Each character directly enters the next generation, and after N generations of evolution, a set of characters {ψ1,ψ2,...,ψ3} is output. K};

[0031] Extract | ψ i | 2 The most relevant skills.

[0032] Preferably, the selection of two parent quantum states ψ p1 and ψ p2 Before that, it also includes:

[0033] Calculate the role Ψ R Average fitness;

[0034] Based on the aforementioned role ΨR Expected fitness Make a choice, with the probability of selection being:

[0035] Preferably, the tissue entropy is defined as the entropy of disorder, and the calculation formula is: S = -∑ i (p i ·logp i ), where p i It is the normalized proportion of the information flow on the i-th edge in the network.

[0036] Preferably, the calculation of the connection strength between the two through cooperation strength and cooperation potential includes:

[0037] The edges are defined as having a cooperative relationship, and their weights are W. ij Based on cooperation strength CI ij and synergistic potential CP ij The initial calculation formula is: W ij =ω hist CI ij +ω potential ·CP ij ;

[0038] The cooperation strength is through Calculation, Comm ij To improve information flow efficiency, Co-work ij To assess the degree of completion of the work;

[0039] The synergistic potential CP ij =Complementarity(P i ,P j Synergy(S) i ,S j ),in:

[0040] Complementarity (P) i ,P j ) = 1 - Cosine Simiarity (P i ,P j ),

[0041] ||S i ∪S j || is the model of the union of employee skills, ||S i ||and||S j || is the modulus of the individual skill vector.

[0042] Preferably, the role vector is input into the personal ability graph for role matching catalysis, including:

[0043] The role vector V role Propagate within the graph, activating relevant skill nodes, with node i having an activation value A. i The calculation is as follows: Among them W ij σ is the weight, and σ is the activation function;

[0044] The activated cluster of nodes forms a response vector V. response Feedback is given to the set of roles V role The system calculates V response With V role The differences are considered, and the skill requirement weights of the character set are adjusted accordingly. The adjustment formula is as follows: Where η is the learning rate, and the output is stable after multiple rounds of iteration.

[0045] The technical solution of this invention systematically evaluates the compatibility between employees and between employees and roles through role dynamic evolution, team synergy potential adaptability analysis, and long-term organizational effectiveness and organizational entropy reduction analysis and evaluation. It effectively ensures that the combination of role capabilities is always highly aligned with the organization's development direction, enhances the overall team synergy effect, quantifies the long-term impact of team configuration on organizational order and efficiency, and provides interpretable long-term value predictions for organizational decision-making.

[0046] Instruction manual illustrations

[0047] Figure 1 This is a flowchart illustrating the method for constructing an organizational team model according to an embodiment of the present invention.

[0048] Figure 2 This is a schematic diagram of the team collaboration adaptability prediction process in an embodiment of the present invention.

[0049] Figure 3 This is a schematic diagram illustrating the process of simulating and predicting organizational team structure in an embodiment of the present invention. Detailed Implementation

[0050] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0051] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0052] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0053] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0054] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0055] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for constructing an organizational team model, comprising:

[0056] Step S1: Collect and preprocess employee data.

[0057] Specifically, this data is based on the company's overall strategy and development goals, as well as the input for building the organizational team's business objectives and strategic value statements. It also includes structured and unstructured data on the organization and its employees obtained from internal systems, including but not limited to organizational evolution data (such as core evolution data, dynamics and impact data, and environmental context data); skills data (such as hard and soft skills like technical, business, and language skills); behavioral data (such as collaboration data, contribution data, and working hours data); cognitive data (such as potential cognitive data like the abstractness, divergence, and logic of thinking); and outcome data (such as results, indicators, influence, and market feedback). The collected data is cleaned to remove noise and outliers, fill in missing values, and perform feature extraction and dimensionality reduction through multi-layer perceptron analysis to facilitate subsequent analysis and modeling.

[0058] Step S2: Input the preprocessed data into the pre-trained large language model, parse it, extract the core value dimensions and constraints, quantify them, and then construct the constraint field.

[0059] Specifically, core value dimensions include user growth, brand influence, and technological innovation; constraints can include time span, Southeast Asian market, and budget. By quantifying dimensions, constructing a constraint field tensor, and quantifying value dimensions and constraints, the dimensions and conditions are quantified into a multi-dimensional constraint field. This includes defining a coordinate axis position P(R) for each dimension and a constraint condition C. i Define the boundary function B in the constrained field. i (P(R)), the closer the function value is to 0, the more the constraint is satisfied; construct the time feature vector of urgency.

[0060] Step S3: Generate a large number of initial character sets from the skill library, evaluate the survivability of each initial character set in the constraint field, and perform character evolution to form a deterministic character set and corresponding character vectors.

[0061] Specifically, for a skill library with M skills, each character is no longer a fixed set, but is represented by a skill probability amplitude vector Ψ: Ψ = [ψ1, ψ2, ..., ψ M ] T Among them, ψ i It is a complex number, |ψ i | 2 This indicates the probability that skill i appears in this character.

[0062] During initialization, the probability amplitude of all skills is set to be equal, that is... Let Ψ represent an initial character in a maximal superposition state. When calculating fitness, observe a character Ψ and collapse it into a classic character set R, with the probability that skill i is included in set R being P(i) = |ψ i | 2 .

[0063] Then, the constraint field and the set of roles are connected through the fitness function F(R) to evaluate the survivability of a set of roles R in the constraint field. The calculation formula is as follows:

[0064] F(R)=α*Innovation(R)+β*Feasibility(R,Orgs)+γ*Complementarity(R,Orgs)+δ*Cohesion(R)

[0065] in:

[0066] Innovation(R): The novelty score of role R, calculated by determining the dissimilarity of its skill set to the Jaccard average of all positions in the organization's history.

[0067] Among them, Shills R and Shills H It is a One-Hot vector representing the skill set of character R and historical character H.

[0068] Feasibility(R,Orgs): Feasibility score, calculated based on the organization's existing skill set, representing the percentage of skills required to achieve R. Where I(·) is the indicator function, Skills Orgs It is the union of the organization's existing skill sets.

[0069] Complementarity(R,Orgs): Complementarity score, calculated as the inverse of the cosine similarity (1 - similarity) between the skill set of R and the organization's existing skill set. Complementarity(R,Orgs) = 1 - Cosine_Similarity(V R V Org ), where V R and V Orgs These are vector representations of character R and the overall skill distribution of the organization, respectively.

[0070] Cohesion(R): Calculates the set of skills R within a character in the skill compatibility graph G. Sill The average connection strength in.

[0071]

[0072] α, β, γ, δ are adjustable weight parameters determined by strategic value and organizational evolution data.

[0073] An initial set of quantum roles is generated, and in this embodiment, the top 1000 roles are preferred based on fitness ranking.

[0074] Then based on the character Ψ R Expected fitness Make a selection. Simultaneously, you can refer to the organization's historical evolution data and drive the evolution of the quantum role through multiple instantiations.

[0075] Calculate the role Ψ R The average fitness of . The selection probability is:

[0076]

[0077] Select two parent quantum states ψ p1 and ψp2 Sub-quantum state ψ offspring It is generated by the linear superposition of their probability amplitudes: The probability distribution of the advantageous skills of the two parents is combined. With a small probability P... mutation Randomly select several skill components ψ from the character's quantum state Ψ. i Perform a phase rotation θ on it: ψ'i = ψ i ·e iθ ,θ~U(0,2π), which changes the probability amplitude of the skill, and actively optimizes it for each offspring quantum character Ψ.

[0078] Calculate the fitness gradient of the current amplitude:

[0079] fitness_gradient=self._calculate_fitness_gradient(role,constraint_field)

[0080] Using gradient information, the probability amplitude of Ψ is fine-tuned to improve its expected fitness. After each generation of evolution, N with the highest expected fitness is retained. elte Each character directly enters the next generation, and after N generations of evolution, the system outputs a set of characters {ψ1,ψ2,...,ψ...}. K}

[0081] Finally, these |ψ i | 2 Extract the most relevant skills to form a deterministic set of characters R. k and their corresponding character vectors

[0082] Step S4, Team Collaboration Adaptability Prediction: Construct an individual capability graph centered on employees and with various skills and cognitive patterns as nodes, and input role vectors into the individual capability graph for role matching catalysis to form a set of idiosyncratic roles; using all employees and idiosyncratic roles as network nodes, calculate the connection strength between them through collaboration intensity and collaboration potential, and output the optimal organizational team structure.

[0083] Specifically, employees are used as the center of the graph, with their various skills and cognitive patterns as nodes and connected together, and historical collaborations adding edge relationships. The weights of the boundaries between nodes are dynamically calculated using a reinforcement learning model. This model is trained using an employee's success history as a positive reward and failure as a negative reward to form an individual competency graph.

[0084] Then the corresponding character vector V role Input into the individual ability graph for role matching and catalyst. Vrole It propagates within the graph, activating relevant skill nodes. The activation value A of node i. i The calculation is as follows:

[0085] Among them W ij σ is the weight, and σ is the activation function.

[0086] The activated cluster of nodes forms a response vector V. response Feedback is then sent to the set of roles. The system calculates V. response With V role The differences are considered, and the skill requirement weights of the character set are adjusted accordingly. The adjustment formula is:

[0087]

[0088] Where η is the learning rate, and the output is stable after multiple rounds of iteration.

[0089] Then, the organization is modeled as a heterogeneous information network, with nodes representing employees and roles. Each node has attributes, including: skill vectors, personal ability graphs, and potential vectors. Edges are set as collaborative relationships, with edge weights W. ij Based on cooperation strength CI ij and synergistic potential CP ij The initial calculation formula is as follows:

[0090] W ij =ω hist CI ij +ω potential ·CP ij ,

[0091] in:

[0092] Cooperation strength through Calculation, Comm ij To improve information flow efficiency, Co-work ij To assess the degree of completion of the work;

[0093] Synergistic potential is calculated using complementary potential and synergistic skills, and the formula is as follows:

[0094] CP ij =Complementarity(P i ,P j Synergy(S) i ,S j )

[0095] in:

[0096] Potential complementarity calculation matches employees based on their potential dimensions, aiming to achieve a 1+1>2 effect where the potential of the matched employees complement each other. The lower the cosine similarity, the higher the complementarity.

[0097] Complementarity (P) i ,P j ) = 1 - Cosine Simiarity (P i ,P j )

[0098] Skill synergy calculations determine whether matching employees' skill sets can cover broader or deeper competency areas.

[0099]

[0100] ||S i ∪S j || is the model of the union of employee skills, ||S i ||and||S j || is the modulus of the individual skill vector. The closer this value is to 1, the lower the skill overlap and the wider the scope of collaboration.

[0101] After multiple rounds of iteration, both parties reached a stable state and output the optimal organizational team structure after specific reshaping. The completed organizational team structure can support evaluation and analysis across multiple dimensions, including internal synergy strength, skill coverage completeness, cognitive diversity, and role matching, ensuring that the talent organization is a comprehensive and highly efficient objective match with the enterprise.

[0102] Of course, after outputting the optimal organizational team structure, the method of this embodiment may also include: step S5, simulating and predicting the output organizational team structure.

[0103] Specifically, the organizational entropy S is defined as a measure of the disorder of the system, and the calculation formula is: S = -∑ i (p i ·logp i ), where p i It is the normalized proportion of the information flow on the i-th edge in the network. High entropy indicates that information is blocked on a few paths and has poor flow; low entropy indicates that information flows evenly and smoothly.

[0104] The output organizational team structure is then placed back into its larger organizational context, i.e., the macro-organizational network, to assess its systemic impact. The macro-organizational network is a large-scale, heterogeneous information network model that abstracts the structure and dynamics of the entire organization. It includes nodes for employees / departments / roles, edges for collaboration / communication, existing tight group structures, key personnel nodes, shortest paths for information or resource dissemination, and robust organizational structures. It transcends the boundaries of individual teams, aiming to capture the complex, systemic interconnections and dependencies between all nodes within the organization, for the analysis and prediction of overall organizational behavior and effectiveness.

[0105] Calculate the current network entropy S respectively now And the S of the future network future The current network entropy calculation is based on role distribution, communication patterns, collaboration efficiency, and skill distribution to obtain a comprehensive organizational entropy. When simulating future network entropy calculation, the current network is replicated as a starting point, and its evolution is simulated monthly based on improvements in personnel collaboration efficiency, communication status, and skill enhancements; the entropy reduction effect ΔS is: ΔS = S now -S future ΔS>0 indicates that the organizational team structure combination can bring about improvements in order and efficiency to the organization. The absolute reduction, relative reduction rate, and overall entropy reduction effect are also calculated.

[0106] By combining micro-level matching degree and macro-level network effects, a feature vector is constructed:

[0107] X = △S,P historical Semantic Similarity ,...]

[0108] Long-term value V long Calculated by a regression model. The final tissue assessment score is then given through an interpretable machine learning model.

[0109] The system outputs a comprehensive assessment report to organizational decision-makers, including suggested organizational structure and roles, corresponding personnel and personnel assessment conclusions, team collaboration and fit conclusions, and organizational assessment conclusions.

[0110] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc. In this application, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention according to actual needs. Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units can be implemented in hardware or as software functional units.

[0111] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method for constructing an organizational team model, characterized in that... include: Collect and preprocess organizational employee data, including organizational evolution data, skills data, cognitive data, behavioral data, and outcome data. The preprocessed data is input into the pre-trained large language model for parsing, and the core value dimensions and constraints are extracted, quantified, and then a constraint field is constructed. A large number of initial character sets are generated from the skill library. The survivability of each initial character set in the constraint field is evaluated and the characters are evolved to form a deterministic character set and the corresponding character vector. Construct an individual competency graph centered on employees and with various skills and cognitive patterns as nodes. Input the role vectors into the individual competency graph for role matching catalysis to form a set of idiosyncratic roles. Using all employees and the idiosyncratic roles as network nodes, calculate the connection strength between them through collaboration intensity and synergy potential, and output the optimal organizational team structure.

2. The method for constructing an organizational team model as described in claim 1, characterized in that... The construction method also includes predictive evaluation of the output organizational team architecture: The output organizational team architecture is deployed to the organizational network, and the current organizational entropy is calculated; and the future organizational entropy is calculated by evolving the future organizational network over time. Calculate the entropy reduction effect; Based on the micro-matching degree, a comprehensive feature vector is constructed and the long-term value is predicted, and a comprehensive evaluation report is output.

3. The method for constructing an organizational team model as described in claim 1, characterized in that... The process of inputting the preprocessed data into a pre-trained large language model for parsing, extracting core value dimensions and constraints, and then quantifying them to construct a constraint field includes: Each dimension defines a coordinate axis position P(R), and each constraint condition C i Define the boundary function B in the constrained field. i (P(R)), the closer the function value is to 0, the more the constraint is satisfied; Construct a time feature vector of urgency.

4. A method for constructing an organizational team model as described in claim 1 or 3, characterized in that... The process of generating a large number of initial character sets from the skill library and evaluating the survivability of each initial character set in the constrained field includes: A character is represented by a skill probability amplitude vector Ψ: Ψ=[ψ1,ψ2,...,ψ M ] T During initialization, the probability amplitude of all skills is set to be equal. Given a character Ψ that collapses into a classic character set R, the probability that skill i is included in the character set R is |ψ|. i | 2 ; A quantum role set Ψ is generated by connecting the constraint field and the role set R through a fitness function F(R). R .

5. The method for constructing an organizational team model as described in claim 4, characterized in that... The fitness function is: F(R)=α*Innovation(R)+β*Feasibility(R,Orgs)+γ * Complementarity(R,Orgs)+δ*Cohesion(R), among which: Innovation(R): The novelty score of role R, which is calculated by determining the Jaccard dissimilarity of its skill set to all positions in the organization's history. Shills R and Shills H It is a One-Hot vector representing the skill set of character R and historical character H; Feasibility(R,Orgs), a feasibility score, is calculated based on the organization's existing skill set to determine the coverage ratio of skills required to achieve role R. Complementarity(R,Orgs): Complementarity score, which calculates the inverse cosine similarity between the skill set of R and the existing skill set of the organization; Cohesion(R): Calculates the set of skills R within a character in the skill compatibility graph G. Sill Average connection strength in; α, β, γ, δ are adjustable weight parameters determined by strategic value and organizational evolution data.

6. The method for constructing an organizational team model as described in claim 4, characterized in that... The character evolution process involves improving the desired fitness through a gradient algorithm and selecting characters, undergoing multiple rounds of iterative evolution, including: Select two parent quantum states ψ p1 and ψ p2 offspring quantum state ψ offspring It is generated by the linear superposition of their probability amplitudes; With a small probability P mutation Randomly select several skill components ψ from the character's quantum state Ψ. i Perform a phase rotation θ on it to change the probability amplitude of the skill, and calculate the fitness gradient of the current amplitude; Using gradient information, the probability amplitude of Ψ is fine-tuned; after each generation of evolution, N with the highest expected fitness is retained. elte Each character directly enters the next generation, and after N generations of evolution, a set of characters {ψ1,ψ2,...,ψ3} is output. K }; Extract | ψ i | 2 The most relevant skills.

7. The method for constructing an organizational team model as described in claim 6, characterized in that... The selection of two parent quantum states ψ p1 and ψ p2 Before that, it also includes: Calculate the role Ψ R Average fitness; Based on the aforementioned role Ψ R Expected fitness Make a choice, with the probability of selection being:

8. The method for constructing an organizational team model as described in claim 2, characterized in that... The organizational entropy is defined as the entropy of disorder, and is calculated using the formula: S = -∑ i (p i ·log p i ), where p i It is the normalized proportion of the information flow on the i-th edge in the network.

9. The method for constructing an organizational team model as described in claim 1, characterized in that... The calculation of the connection strength between the two through cooperation strength and cooperation potential includes: The edges are defined as having a cooperative relationship, and their weights are W. ij Based on cooperation strength CI ij and synergistic potential CP ij The initial calculation formula is: W ij =ω hist CI ij +ω potential ·CP ij ; The cooperation strength is through Calculation, Comm ij To improve information flow efficiency, Co-work ij To assess the degree of completion of the work; The synergistic potential CP ij =Complementarity(P i ,P j Synergy(S) i ,S j ),in: Complementarity(P i ,P j )=1-Cosine Simiarity (P i ,P j ), ||S i ∪S j || is the model of the union of employee skills, ||S i ||and||S j || is the modulus of the individual skill vector.

10. The method for constructing an organizational team model as described in claim 1, characterized in that... The character vector is input into the personal ability graph for character matching catalysis, including: The role vector V role Propagate within the graph, activating relevant skill nodes, with node i having an activation value A. i The calculation is as follows: Among them W ij σ is the weight, and σ is the activation function; The activated cluster of nodes forms a response vector V. response Feedback is given to the set of roles V role The system calculates V response With V role The differences are considered, and the skill requirement weights of the character set are adjusted accordingly. The adjustment formula is as follows: Where η is the learning rate, and the output is stable after multiple rounds of iteration.

Images