A job growth accessibility calculation method and system based on capability gap coverage and reinforcement action set selection

Abstract

The application provides a post growth accessibility calculation method and system based on capability gap coverage and reinforcement action set selection, comprising the following steps: step 1: obtaining a student's current capability vector, a post capability requirement vector and a capability dimension reinforcement difficulty; step 2: calculating a capability gap; step 3: obtaining a reinforcement action set; step 4: calculating the coverage of the candidate reinforcement action subset to the capability gap; step 5: calculating the reinforcement cost of the candidate reinforcement action subset; step 6: calculating the post growth accessibility score; step 7: selecting the optimal reinforcement action subset from the reinforcement action set according to the post growth accessibility score, and outputting at least one of the post growth accessibility score, the optimal reinforcement action subset, the candidate post entry marker or the post sorting weight; the technical solution quantitatively models the capability gap and further calculates the coverage of the reinforced capability to the post requirement, thereby realizing the judgment of the post.

Description

Technical Field

[0001] This invention relates to the field of data processing and job recommendation auxiliary calculation technology, and in particular to a method and system for calculating job growth attainability based on competency gap coverage and the selection of a set of reinforcement actions. Background Technology

[0002] In the fields of job recommendation, career recommendation, and employment support decision-making, existing technologies typically involve the following categories of foundational technologies.

[0003] The first category is job matching technology based on feature extraction and similarity calculation. This type of method typically extracts keywords, tags, or vector representations from resume text and job descriptions, and calculates the degree of matching between users and jobs through cosine similarity, Euclidean distance, or machine learning models, thereby outputting job recommendation results.

[0004] The second category is recommendation techniques based on ability tags or ability models. This type of method represents user abilities as a set of ability tags or ability vectors, and job requirements as corresponding ability requirements, and then matches users with job ability requirements based on the differences or similarities between them.

[0005] The third category is job recommendation technology based on knowledge graphs. This type of method constructs a graph of relationships between entities such as jobs, skills, industries, and users, and uses graph queries or graph reasoning to obtain candidate jobs, thereby improving the structured nature of the recommendations.

[0006] The fourth category is ranking and filtering techniques based on recommender systems. These methods typically include candidate recall, ranking models, and re-ranking mechanisms. They rank candidate positions using various features (such as user features, job features, and interaction features) to obtain the final recommendation results.

[0007] The aforementioned technologies provide the basic data processing, feature representation, and recommendation computation capabilities for this application, but they mainly focus on matching or relevance computation in the current state.

[0008] Existing job and career recommendation methods are mostly based on current matching, that is, determining the recommendation result based on the similarity, matching degree, or relevance between the student's current resume, current ability tags, current profile, or current assessment results and the job requirements. This type of method is suitable for determining whether a user currently meets the job requirements, but it still has the following technical shortcomings in the context of student career planning.

[0009] First, existing methods are prone to overlooking growth-oriented positions. For college students or users lacking complete work experience, there is often a gap between their current skills and the requirements of their target positions. Existing static matching methods typically lower the ranking of such positions or even filter them out entirely. However, some of these positions may only require a small amount of coursework, project experience, certification, or internships to meet the basic requirements. Therefore, relying solely on current matching scores for recall or ranking can easily lead to the omission of positions that are "not currently a perfect match but can be achieved with reasonable improvements."

[0010] Second, existing methods struggle to translate skill gaps into a calculable set of reinforcement actions. Even when existing career recommendation systems can alert users to missing skills, certifications, or practical experience, they typically offer only textual suggestions or labeled shortcomings. For example, while the system might suggest a lack of database skills, project experience, or industry certifications, it cannot further calculate the coverage effect of different reinforcement actions on the skill gaps, nor can it compare the reinforcement costs of different combinations of actions.

[0011] Third, existing methods struggle to incorporate accessibility enhancement into candidate job evaluation or ranking weight calculations. Current job recommendation rankings typically rely primarily on current matching degree, text similarity, tag overlap, or graph correlation. Even when career path or competency analysis modules exist, they often fail to convert the "remaining gap after the competency gap is covered by enhancement actions" and the "cost of enhancement actions" into job growth accessibility scores, which are then used as entry criteria or ranking weights for candidate jobs.

[0012] Therefore, the technical problem to be solved by this application is: how to calculate the coverage of the capability gap of the set of reinforcement actions by combining the capability gain vector and reinforcement cost of the reinforcement actions during the auxiliary calculation process of job recommendation, based on the capability gap between the student's current ability and the capability requirements of the target job, and generate the job growth attainable score and the set of reinforcement actions accordingly, so as to determine whether the target job belongs to the candidate job that can be attained after reasonable reinforcement. Summary of the Invention

[0013] In view of this, the purpose of this invention is to provide a method and system for calculating job growth attainability based on capability gap coverage and selection of reinforcement action set. By quantitatively modeling capability gaps and further calculating the coverage of the capabilities after reinforcement to the job requirements, the method can determine whether the job is "attainable after reinforcement".

[0014] To achieve the above objectives, the present invention adopts the following technical solution: a method for calculating job growth attainability based on capability gap coverage and the selection of a set of reinforcement actions, comprising the following steps:

[0015] Step 1: Obtain the student's current ability vector, job ability requirement vector, and difficulty of improving ability dimensions;

[0016] Step 2: Calculate the capability gap based on the student's current capability vector and the job capability requirement vector;

[0017] Step 3: Obtain the set of reinforcement actions, wherein each reinforcement action in the set of reinforcement actions includes a capability gain vector and a reinforcement cost;

[0018] Step 4: For the candidate reinforcement action subset, calculate the coverage of the candidate reinforcement action subset on the capability gap based on the capability gap, the capability dimension reinforcement difficulty, and the capability gain vector of each reinforcement action in the candidate reinforcement action subset;

[0019] Step 5: Calculate the reinforcement cost of the candidate reinforcement action subset based on the reinforcement cost of each reinforcement action in the candidate reinforcement action subset;

[0020] Step 6: Calculate the job growth achievable score based on the coverage and reinforcement costs;

[0021] Step 7: Based on the job growth attainable score, select the optimal subset of reinforcement actions from the set of reinforcement actions, and output at least one of the job growth attainable score, the optimal subset of reinforcement actions, the candidate job entry flag, or the job ranking weight.

[0022] In a preferred embodiment, step 1 specifically includes:

[0023] Step 11: Pre-build the set of capability dimensions 'm' represents the total number of capability dimensions; these capability dimensions include professional skills, tool skills, project practice ability, engineering ability, data analysis ability, communication and collaboration ability, certifications and qualifications, internship experience, and portfolio completeness; each capability dimension is represented by a normalized numerical value, with a value range of [0,1] or [0,100]:

[0024]

[0025] Where D represents the difficulty of enhancing the capability dimension. Indicates the first The difficulty of enhancing each capability dimension is determined by historical learning time, task completion cycle, certificate acquisition cycle, expert rules, or historical sample statistics.

[0026] Step 12: Obtain Students Current capability vector and positions Capability requirement vector The student's current competency vector comes from resume analysis, course grades, project experience, competition experience, certificate information, internship experience, portfolio information, or manually entered competency evaluation results; the job competency requirement vector comes from job description analysis, job competency model, job tag library, or manual configuration.

[0027]

[0028]

[0029] in, Students In the Current ability value in each ability dimension Indicates job position For the The required values ​​for each capability dimension, k∈(1,m).

[0030] In a preferred embodiment, step 2 specifically includes: based on the student's current ability vector and job competency requirements vector Calculate the student relative to the position Competency gaps in each competency dimension; a competency gap represents the portion of a job requirement that exceeds a student's current competency; if a student's current competency already meets or exceeds the job requirement, then the gap in that dimension is 0;

[0031]

[0032] in, This represents the skill gap between student u and job j in the k-th skill dimension;

[0033]

[0034] in, This represents the competency gap vector of student u relative to job j, composed of competency gaps across various competency dimensions. Students In the Current ability value in each ability dimension Indicates job position For the The required values ​​for each capability dimension, k∈(1,m); Step 2 decomposes the reasons for job mismatch from a whole matching into multiple calculable capability gaps, providing input for subsequent reinforcement action selection.

[0035] In a preferred embodiment, step 3 specifically includes: presetting a set of reinforcement actions. ; This represents the nth reinforcement action; each reinforcement action This represents a course learning session, project practice, certificate acquisition, internship training, competition training, scientific research practice, or portfolio construction, where i∈(1,n); each reinforcement action must include at least the action type, action name, capability gain vector, reinforcement cost, and applicable conditions.

[0036]

[0037]

[0038] in, Indicates the action type. Indicates the action name. Represents the capability gain vector. Indicates a supplementary action For the The expected improvement in each capability dimension, k∈(1,m); Indicates a supplementary action The cost of reinforcement is normalized from time cost, difficulty cost, resource cost, or comprehensive cost; This indicates the applicable conditions for the reinforcement action;

[0039] Explanation of the calibration method, accuracy requirements, and update mechanism for the capability gain vector:

[0040] The capability gain vector maintains the same dimension as the capability dimension set; when the capability dimension set contains m capability dimensions, the capability gain vector of the i-th enhancement action is represented as:

[0041]

[0042] in, This represents the expected improvement of the k-th capability dimension by the i-th reinforcement action. With the student's current ability value and job competency requirements The same normalization scale is used. When student ability values ​​and job requirement values ​​are represented using [0,1], It also uses [0,1] to represent; when student ability values ​​and job ability requirement values ​​are represented using [0,100], It is also represented by [0,100]. The following example uses the [0,1] scale.

[0043] Capability gain vector The results are obtained through at least one of the following methods: historical sample statistics, expert rule calibration, and predictive model output. The system prioritizes the use of historical sample statistics; when the number of historical samples is insufficient, it supplements the results with expert rule calibration or predictive model output.

[0044] When historical sample data exists, the system statistically completes the reinforcement action a. i The change in capability before and after. Let the s-th historical sample complete reinforcement action a. i The k-th dimension capability value is A. before,s,k Complete the reinforcement action a i The k-th dimension capability value is A. after,s,k The number of historical samples that completed this reinforcement action is N. i Then the historical statistical gain of this reinforcement action on the k-th dimension capability is:

[0045]

[0046] in, To avoid the negative gain effect caused by fluctuations in ability assessment; when the number of historical samples reaches the preset sample threshold N0, the system adopts historical statistical gain; N0 is set according to the system sample size, and in this embodiment, N0=30.

[0047] When the number of historical samples is insufficient, the system employs expert rules for calibration. Experts, based on course syllabi, project task requirements, certificate assessment content, internship requirements, or competition training content, calibrate the reinforcement action a. i An expert score E is given for the degree of influence of the k-th ability dimension. ik If the expert rating ranges from 0 to 5 points, then the expert calibration gain is expressed as:

[0048] ΔE ik =M k ×E ik / 5

[0049] Among them, M k M represents the maximum allowable gain for the k-th capability dimension under a single enhancement action, used to prevent the enhancement amount of a single action to a certain capability dimension from being overstated. k The system administrator sets the value based on the capability dimension type, for example, 0.30 or 0.40.

[0050] When configuring the prediction model, the system takes the action type, action duration, task difficulty, assessment method, student's basic ability, and historical performance as inputs, and the prediction model outputs the prediction gain ΔP. ik The output of the prediction model is also normalized to the same scale as the capability value.

[0051] The system obtains the final capability gain value by fusing the data in the following manner:

[0052] Gain i,k =clip(λH×ΔH ik +λE×ΔE ik +λP×ΔPik ,0,1)

[0053] λH+λE+λP=1

[0054] Where λH, λE, and λP represent the weights of historical sample statistics, expert rule calibration, and predictive model output, respectively. If a certain type of calibration data is missing, its corresponding weight is reset to 0, and the remaining weights are renormalized. The `clip` function is used to limit the gain value to the capability value scale. Capability gain values ​​are stored with a granularity of 0.01. The system retains at least four decimal places during intermediate calculations and two or four decimal places as needed during output.

[0055] The system updates the capability gain vector Gain according to a preset period. i The update cycle is one month, one semester, or one training batch. A rolling average method is used for updates.

[0056] Gain i (t+1)=(1-η)Gain i (t)+ηGain i,recent

[0057] Among them, Gain i (t) represents the capability gain vector before the update, Gain i,recent This represents the capability gain vector obtained in the most recent statistical period, and η represents the update coefficient, with a value ranging from 0 to 1. Through this method, the set of reinforcement actions Ω is updated as the course content, job requirements, and historical training results change.

[0058] The numerical source of the enhancement action capability gain vector in the example:

[0059] To illustrate the source of the specific values ​​in the capability gain vector, let's take the "back-end integrated project" reinforcement action a as an example. i For example, the system extracts student samples who have completed the project training from historical student growth records, and records the average ability values ​​of four dimensions before and after completing the project training: Java backend capabilities, database capabilities, engineering capabilities, and internship experience.

[0060] A before (a i =(0.37, 0.34, 0.23, 0.10)

[0061] A after (a i =(0.62,0.39,0.43,0.10)

[0062] G1=A after (a i )−A before (ai =(0.25,0.05,0.20,0.00)

[0063] Therefore, G1=(0.25,0.05,0.20,0.00) in the example is not an arbitrarily given value, but is obtained from the historical average change in capabilities before and after completing the "backend comprehensive project". This vector indicates that this action mainly improves Java backend capabilities and engineering capabilities, with a slight improvement in database capabilities, and no direct improvement in internship experience.

[0064] Similarly, the capability gain vectors of other reinforcement actions are also obtained from the historical average capability change before and after the corresponding action is completed, the expert rule calibration results, or the prediction model output results, and are uniformly mapped to the [0,1] capability value scale.

[0065] Regarding reinforcement costs i Normalization method:

[0066] reinforcement cost i It is obtained by weighting time cost, difficulty cost, and resource cost. Let the original time cost of the i-th reinforcement action be T. i The original difficulty cost is L. i The original resource cost is R. i First, normalize each cost component to the range [0,1].

[0067]

[0068]

[0069]

[0070] in, This represents the normalized time cost. This represents the normalized difficulty cost. T represents the normalized resource cost. max L represents the maximum acceptable time cost preset by the system. min and L max R represents the minimum and maximum difficulty scores, respectively. max This represents the maximum acceptable resource cost preset by the system, and ε is a constant to prevent the denominator from being zero.

[0071] reinforcement action a i The comprehensive reinforcement cost is calculated as follows:

[0072]

[0073] wT+wL+wR=1

[0074] Here, wT, wL, and wR represent the weights of time cost, difficulty cost, and resource cost, respectively. In the student career planning scenario, time investment has a significant impact on feasibility, so the system sets wT=0.50, wL=0.30, and wR=0.20.

[0075] By using the above normalization method, we can strengthen the cost. i It is a dimensionless numerical value and is mapped to the range [0,1]. Capability gain vector Gain i Used to represent the improvement to a stat by a reinforcement action; reinforcement cost. i Used to represent the cost required to implement reinforcement actions, the two are used as benefit items and cost items respectively in the subsequent growth attainable points calculation, avoiding the direct addition of data with different dimensions.

[0076] Regarding applicable conditions Cond i Supplement to the definition, storage, and matching:

[0077] The applicable conditions for each reinforcement action (Cond) i This includes at least one of the following categories: prerequisite competency requirements, job gap matching requirements, student status requirements, and exclusion criteria, represented as:

[0078] Cond i ={Pre i ,Target i User i ,Exclude i}

[0079] Among them, Pre i Indicates prerequisite capability conditions, Target i Indicates the job vacancy matching criteria, User i Indicates student status conditions, Exclude i This indicates exclusion conditions. Applicable conditions are stored using rule tables, key-value fields, database fields, or logical expressions.

[0080] Prerequisite competency requirements determine whether a student possesses the fundamental skills to perform the reinforcement exercise. For example, an advanced backend comprehensive project requires a Java backend skill score of at least 0.30 and a database skill score of at least 0.20. Job gap matching criteria determine whether the reinforcement exercise can cover the competency gaps for the target position. Student status criteria determine whether a student meets the requirements for participating in the reinforcement exercise, such as their major, year of study, completed courses, available time, eligibility for certifications, or internship window. Exclusion criteria prevent duplicate recommendations of exercises already completed by the student or exclude exercises that are not relevant to the student's current stage.

[0081] The system performs reinforcement action ai When making an applicability judgment, calculate:

[0082] I i (u,j)=1, when Pre i Target i User i All satisfy and Exclude i Not satisfied

[0083] I i (u,j)=0, when any of the above conditions are not met.

[0084] When I i When (u,j)=1, it means that student u satisfies the reinforcement action a for job j. i Applicable conditions; when I i When (u,j)=0, it means that the applicable conditions are not met.

[0085] The job gap matching criteria are determined based on the effective coverage of the capability gaps by the reinforcement actions:

[0086]

[0087] when When, it indicates a supplementary action a. i This should at least cover part of the skill gaps for the target position. Based on this, the system generates a set of candidate reinforcement actions for student u and position j:

[0088]

[0089] Therefore, the set of reinforcement actions Ω is first constructed from data such as courses, projects, certificates, practical training, internships, competitions, scientific research, portfolio construction, historical growth records and expert rules. Then, it is structured and filtered through capability gain vectors, reinforcement costs and applicable conditions. Based on this, technical personnel generate candidate sets of reinforcement actions.

[0090] In a preferred embodiment, step 4 specifically includes: for any subset of candidate reinforcement actions Calculate the total capability gain of the action subset in each capability dimension, and further calculate the remaining gap after reinforcement;

[0091]

[0092]

[0093] in, Indicates a supplementary action The expected capability gain for the k-th capability dimension; The total capability gain of the candidate reinforcement action subset Δ in the k-th capability dimension is obtained by summing the capability gains of each reinforcement action in the subset in that capability dimension. This represents a set of reinforcement actions;

[0094]

[0095] This represents the remaining capability gap of student u in the k-th capability dimension that is still not covered after adopting the candidate reinforcement action subset Δ for job j; This represents the skill gap between student u and job j in the k-th skill dimension;

[0096]

[0097] in, This represents the coverage of the competency gap of student u relative to job j by the subset of candidate reinforcement actions Δ. Indicates the first The difficulty of enhancing each capability dimension This represents a set of reinforcement actions. To prevent constants with a denominator of zero; when students do not have a skill gap for position j, Set it to 1.

[0098] In a preferred embodiment, step 5 specifically includes: cost calculation of the reinforcement action set.

[0099]

[0100]

[0101] in, This represents the total reinforcement cost of the subset Δ of candidate reinforcement actions. Indicates a supplementary action The cost of reinforcement To preset the maximum allowable reinforcement cost or the maximum cost in the candidate set, Used to incorporate cost constraints into the calculation of subsequent growth attainable points. To prevent constants with a denominator of zero.

[0102] In a preferred embodiment, step 6 specifically includes: calculating job growth attainable points based on gap coverage and reinforcement costs;

[0103]

[0104]

[0105] in, Students Relative to the position In the set of reinforcement actions The growth rate can reach points, This represents the coverage of the competency gap of student u relative to job j by the subset of candidate reinforcement actions Δ; Students With position The current matching degree is calculated from existing matching models, rule matching, or vector similarity. Used to incorporate cost constraints into the calculation of subsequent growth attainable points; These are the weight parameters.

[0106] Specifically, the calculation of the current matching degree based on the capability vector system includes:

[0107] The current matching degree Match(u,j) is directly based on the student's current ability vector A that has already been constructed in this application. u Job competency requirement vector R j and capability gap vector Gap u,j,k Calculation. Thus, Match(u,j) forms a direct interface with the capability vector system in this application, rather than an external matching score calculated independently of the capability vector system.

[0108] When the capability value is represented by [0,1], the current matching degree based on the capability vector system is expressed as:

[0109]

[0110] Among them, A u,k R represents the current ability value of student u in the k-th ability dimension. j,k This represents the capability requirement value for job j in the k-th capability dimension. D represents a capacity gap. k ε represents the difficulty or weight of reinforcing the k-th capability dimension, and ε is a constant to prevent the denominator from being zero.

[0111] because:

[0112]

[0113] Therefore, the input for Match(u,j) is entirely derived from the capability vector system of this application. Specifically, and Determine the capacity gaps in each dimension , This is used to reflect the weight of the difficulty of strengthening different capability dimensions in the current matching degree calculation, and finally obtains the current matching degree on the same scale as the capability gap.

[0114] When students meet or exceed job requirements in all competency dimensions, the competency gap in each dimension is 0, that is:

[0115]

[0116] At this point, Match(u,j)=1 indicates that the student's current abilities meet the job requirements.

[0117] The above calculation method makes Match(u,j) represent the degree to which the student's current ability meets the job requirements, while This indicates the degree to which the set of reinforcement actions Δ covers the capability gap. The interface relationship between the two is: Match(u,j) represents the capability gap before reinforcement. , Both methods utilize the remaining gaps after reinforcement and are based on the same set of ability dimensions and the same student's current ability vector. Vector of competency requirements for the same position Same capacity gap It is calculated using the same normalized scale.

[0118] Therefore, Match(u,j) is not an external matching result separate from the technical solution of this application, but rather the current satisfaction calculation result within the capability vector system of this application. This result, along with the subsequent coverage (u,j,Δ) and normalized cost CostNorm(Δ), is included in the growth attainability score calculation.

[0119] Reach(u,j,Δ) = α × Cover(u,j,Δ) + β × Match(u,j) - (1 - α - β) ×CostNorm(Δ)

[0120] 0≤α≤1, 0≤β≤1, α+β≤1

[0121] Here, Cover(u,j,Δ), Match(u,j), and CostNorm(Δ) are all dimensionless values ​​in the range [0,1]. Therefore, the three can be weighted and calculated in the same growth attainable fraction formula, and there is no problem of directly adding different dimensions.

[0122] Coverage (u,j,Δ) is used to measure the degree to which the set of reinforcement actions covers the capability gap. It is the core evaluation indicator of the technical solution in this application and is therefore given the highest weight.

[0123] The current matching degree Match(u,j) is used to reflect the degree to which the student's current ability meets the job requirements. It can reflect the student's basic ability level before the reinforcement action, but it cannot directly reflect the ability improvement effect after the reinforcement. Therefore, it is given the second highest weight.

[0124] The normalized cost CostNorm(Δ) measures the time, difficulty, and resource costs required to implement reinforcement. Cost factors are used to constrain excessively costly reinforcement schemes and are therefore included as a deduction in the calculation.

[0125] Since the technical objective of this application is to evaluate whether a position can be achieved through reasonable reinforcement, coverage is more important than the current matching degree, and the current matching degree is more important than the reinforcement cost. Based on the above principles, the following settings are made:

[0126]

[0127] In this embodiment, the following values ​​are taken:

[0128] α=0.55, β=0.25,

[0129] The goal is to prioritize the effectiveness of remedial actions in addressing skill gaps, while also taking into account students' current skill levels and the cost constraints of remedial measures.

[0130] In a preferred embodiment, step 7 specifically includes: a set of reinforcement actions. Select reinforcement action subset The selection objective is to minimize reinforcement costs or maximize growth potential while satisfying the remaining gap or coverage constraints.

[0131]

[0132] Students Relative to the position In the set of reinforcement actions The growth potential is achievable; Δ represents a subset of candidate reinforcement actions. This represents a set of reinforcement actions;

[0133] The set of reinforcement actions is selected using a combination of coverage constraints and cost minimization.

[0134]

[0135]

[0136] in, This represents the total reinforcement cost of the subset Δ of candidate reinforcement actions. This indicates that after student u adopts the candidate reinforcement action subset Δ for position j, The allowable weighted residual gap threshold; if a certain set of reinforcement actions makes the weighted residual gap less than or equal to Then it is considered that the set of actions can enable the job Reaching a state where reinforcement is possible; among multiple sets of actions that satisfy the constraints, selecting the set of actions with the minimum total cost as the optimal set. ; Indicates constraints;

[0137] A greedy selection approach is used to reduce computational load: First, the unit cost gap coverage benefit of each reinforcement action is calculated, and then the actions are added to the action set one by one from high to low benefit until the coverage reaches a preset threshold or the maximum allowable cost is reached. :

[0138] .

[0139] in, Indicates the first The difficulty of enhancing each capability dimension Indicates a supplementary action The expected capability gain for the k-th capability dimension; This represents the skill gap between student u and job j in the k-th skill dimension; Indicates a supplementary action The cost of reinforcement; To prevent constants with a denominator of zero.

[0140] In a preferred embodiment, the candidate positions are entered or ranked with weights.

[0141]

[0142] Where Δ* represents the optimal subset of reinforcement actions determined according to the principles of coverage constraints, cost constraints, or maximizing growth attainability. This indicates the job growth score that student u can achieve for job j after adopting the optimal reinforcement action subset Δ*; This represents the total reinforcement cost of the optimal reinforcement action subset Δ*;

[0143]

[0144] in, For growth to reach a threshold, To the maximum reinforcement cost threshold, For students The set of candidate positions, The base recommendation score output by the existing recommendation system. To introduce ranking weights for growth reachability, and These are weight parameters, used to adjust the influence of the base recommendation score and the growth attainable score on the ranking weights, respectively.

[0145] This invention also provides a job development attainability calculation system based on competency gap coverage and the selection of a set of reinforcement actions, comprising:

[0146] Ability Input Module: Used to obtain the student's current ability vector. Job competency requirements vector and the difficulty of improving capabilities ;

[0147] Capacity gap calculation module: used to calculate based on Computing capacity gap; This represents the skill gap between student u and job j in the k-th skill dimension; Students In the Current ability value in each ability dimension Indicates job position For the The required values ​​for each capability dimension, k∈(1,m);

[0148] Reinforcement Action Modeling Module: Used to store and retrieve reinforcement action sets. And configure a capability gain vector for each reinforcement action. and reinforcement costs ;

[0149] Gap Coverage Calculation Module: Used to calculate the set of reinforcement actions. Coverage of capacity gaps ;

[0150] Cost Constraint Selection Module: Used to select a set of reinforcement actions based on cost minimization, coverage threshold, or maximization of growth attainability. ;

[0151] Growth Score Calculation Module: Used to calculate based on coverage, reinforcement cost, and optional current matching score. ;

[0152] Candidate output module: used to output growth attainable score, set of reinforcement actions, candidate position entry mark or ranking weight;

[0153] The system operates as described in the method for calculating job growth attainability based on capability gap coverage and the selection of a set of reinforcement actions.

[0154] Compared with the prior art, the present invention has the following beneficial effects:

[0155] First, it reduces the false negative rate for growth-oriented positions. Compared to solutions that rely solely on current match, text similarity, or tag overlap for recommendations, this application identifies positions that are "not currently a perfect match but can be improved through reasonable enhancements" by assessing skill gap coverage and growth attainability scores. This reduces the probability of growth-oriented positions being filtered out during the recall phase. This advantage stems from the application's use of a continuous calculation mechanism, including Gap, Remain, Cover, and Reach, rather than simply outputting the current match score.

[0156] Second, this application improves the interpretability of job recommendation auxiliary calculations. Existing matching methods typically output an overall matching score, which is insufficient to explain why a user is not matched or how to improve their skills. This application breaks down the gap between job requirements and students' current abilities into gaps across multiple ability dimensions and further outputs a set of improvement actions, thus explaining the source of the job growth attainable score.

[0157] Third, this application transforms reinforcement suggestions from textual prompts into measurable objects. Existing systems often suggest "improving database capabilities" or "increasing project experience," but cannot calculate the coverage effects and costs of different reinforcement actions. This application represents reinforcement actions as... and This allows supplementary activities such as courses, projects, certificates, and practical training to participate in combination selection and cost constraint calculation.

[0158] Fourth, it is easy to integrate as an intermediate module into existing recommendation systems. This application outputs growth attainability scores, a set of reinforcement actions, candidate entry tags, or ranking weights, without requiring replacement of existing resume parsing, job profiling, or basic ranking models. Therefore, it can be used as an auxiliary calculation module in existing job recommendation systems, reducing system modification costs.

[0159] Fifth, it achieves a balance between computational complexity and result quality. This application supports both constraint-based optimization for selecting the set of reinforcement actions and cost-based coverage of benefits. A greedy selection method can be used to reduce the computational complexity of online computation when there are a large number of reinforcement actions. Attached Figure Description

[0160] Figure 1 This is a flowchart illustrating the calculation of job advancement attainability according to a preferred embodiment of the present invention.

[0161] Figure 2 This is a schematic diagram of the system modules and data flow according to a preferred embodiment of the present invention;

[0162] Figure 3 This is a schematic diagram of the reinforcement action data structure of a preferred embodiment of the present invention;

[0163] Figure 4This is a schematic diagram illustrating the capability gap coverage of a preferred embodiment of the present invention. Detailed Implementation

[0164] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0165] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0166] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations according to this application; as used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise; furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0167] refer to Figure 1-4 This paper presents a method for calculating job growth attainability based on competency gap coverage and the selection of a set of reinforcement actions. The method's inputs include the student's current competency vector, job competency requirement vector, competency dimension reinforcement difficulty, set of reinforcement actions, competency gain vector for each reinforcement action, and reinforcement cost. The outputs include the target job's growth attainability score, the set of reinforcement actions used to cover competency gaps, and job candidate entry markers or job ranking weights.

[0168] Unlike methods that only calculate the current match between students and positions, this application further calculates whether the skill gap can be covered by several reinforcement actions when the student's current abilities do not fully meet the position requirements. The remaining gap and reinforcement costs are then included in the growth attainability score calculation. Therefore, positions with a low current match but which can be met through reasonable reinforcement can be identified by the system as growth attainable positions. Specifically, this includes the following steps:

[0169] Step 1: Obtain the student's current ability vector, job ability requirement vector, and difficulty of improving ability dimensions;

[0170] Step 11: Constructing the Capability Dimension Space

[0171] The system pre-builds a set of capability dimensions. The competency dimensions may include professional skills, tool skills, project practice ability, engineering ability, data analysis ability, communication and collaboration ability, certifications and qualifications, internship experience, and portfolio completeness. Each competency dimension is represented by a normalized numerical value, with a range of [0,1] or [0,100]. For ease of explanation, [0,1] will be used as an example below.

[0172]

[0173] in, Indicates the first The difficulty of enhancing each capability dimension can be determined by factors such as historical learning time, task completion cycle, certificate acquisition cycle, expert rules, or historical sample statistics.

[0174] Step 12: Obtaining the student's current ability vector and the job requirement vector

[0175] System obtains students Current capability vector and positions Capability requirement vector A student's current competency vector can be derived from resume analysis, course grades, project experience, competition experience, certificate information, internship experience, portfolio information, or manually entered competency evaluation results. A job competency requirement vector can be derived from job description analysis, job competency models, job tag libraries, or manually configured.

[0176]

[0177]

[0178] in, Students In the Current ability value in each ability dimension Indicates job position For the The required values ​​for each capability dimension.

[0179] Step 2: Calculate the skill gap based on the student's current skill vector and the job skill requirement vector:

[0180] The system is based on the student's current ability vector and job competency requirements vector Calculate the student relative to the position Competency gaps across various competency dimensions. A competency gap represents the portion of a job requirement that exceeds a student's current competency; if a student's current competency already meets or exceeds the job requirements, then the gap in that dimension is 0.

[0181]

[0182]

[0183] This step breaks down the reasons for job mismatch from a holistic matching into multiple calculable capability gaps, providing input for subsequent reinforcement actions.

[0184] Step 3: Obtain the set of reinforcement actions, wherein each reinforcement action in the set of reinforcement actions includes a capability gain vector and a reinforcement cost;

[0185] System preset reinforcement action set Each reinforcement action This can represent a course learning session, project practice, certificate acquisition, internship training, competition training, scientific research practice, or portfolio construction. Each reinforcement action should include at least the action type, action name, capability gain vector, reinforcement cost, and applicable conditions.

[0186]

[0187]

[0188] in, Indicates a supplementary action For the The projected improvement in each capability dimension; Indicates a supplementary action The cost of reinforcement can be normalized from time cost, difficulty cost, resource cost, or comprehensive cost; This indicates the applicable conditions for the reinforcement action.

[0189] Step 4: For the candidate reinforcement action subset, calculate the coverage of the candidate reinforcement action subset on the capability gap based on the capability gap, the capability dimension reinforcement difficulty, and the capability gain vector of each reinforcement action in the candidate reinforcement action subset;

[0190] For any subset of candidate reinforcement actions The system calculates the total capability gain of the action subset in each capability dimension, and further calculates the remaining gap after reinforcement.

[0191]

[0192]

[0193]

[0194]

[0195] in, To prevent constants with a denominator of zero, when students do not have a skill gap for position j, they can... Set it to 1.

[0196] Step 5: Calculate the reinforcement cost of the candidate reinforcement action subset based on the reinforcement cost of each reinforcement action in the candidate reinforcement action subset;

[0197]

[0198]

[0199] in, To preset the maximum allowable reinforcement cost or the maximum cost in the candidate set, This is used to incorporate cost constraints into the calculation of subsequent growth attainable points.

[0200] Step 6: Calculate the job growth achievable score based on the coverage and reinforcement costs;

[0201] The system calculates the job growth attainability score based on gap coverage and reinforcement costs. The growth attainability score considers not only whether the skill gap can be covered by reinforcement actions, but also the cost required to achieve that coverage.

[0202]

[0203]

[0204] in, Students Relative to the position In the set of reinforcement actions The growth rate can reach points; Students With position The current matching degree can be calculated from existing matching models, rule matching, or vector similarity. These are weighting parameters. Authorization is being prioritized during implementation. It can be set to a larger value so that the growth attainable score mainly reflects the degree of gap coverage.

[0205] In another implementation, the growth attainable score can be calculated directly based on the gap coverage and cost without incorporating the current matching degree:

[0206]

[0207] Step 7: Based on the job growth attainable score, select the optimal subset of reinforcement actions from the set of reinforcement actions, and output at least one of the job growth attainable score, the optimal subset of reinforcement actions, the candidate job entry flag, or the job ranking weight.

[0208] Step 71: The system retrieves the set of reinforcement actions. Select reinforcement action subset The selection objective is to minimize reinforcement costs or maximize growth potential while satisfying the remaining gap or coverage constraints.

[0209]

[0210] Alternatively, the system selects the set of reinforcement actions using coverage constraints and cost minimization:

[0211]

[0212]

[0213] in, This is the permissible weighted residual gap threshold. If a certain set of reinforcement actions can make the weighted residual gap less than or equal to... Then it is considered that the set of actions can enable the job Reaching a state where reinforcement is possible; among multiple sets of actions that satisfy the constraints, selecting the set of actions with the minimum total cost as the optimal set. .

[0214] In another implementation, the system uses a greedy selection method to reduce the amount of computation: first, the unit cost gap coverage benefit of each reinforcement action is calculated, and then the actions are added to the action set one by one from high to low according to the benefit, until the coverage reaches a preset threshold or the maximum allowable cost is reached.

[0215]

[0216] Step 71: Output the weights for candidate positions entering or ranking them.

[0217]

[0218]

[0219] in, For growth to reach a threshold, To the maximum reinforcement cost threshold, For students The set of candidate positions, The base recommendation score output by the existing recommendation system. To introduce ranking weights based on growth attainability, this application can be integrated as an intermediate calculation module into existing job recommendation systems without requiring modifications to the entire recommendation platform.

[0220] The system includes:

[0221] Ability Input Module: Used to obtain the student's current ability vector. Job competency requirements vector and the difficulty of improving capabilities .

[0222] Capacity gap calculation module: used to calculate based on Computing capacity gap.

[0223] Reinforcement Action Modeling Module: Used to store and retrieve reinforcement action sets. And configure a capability gain vector for each reinforcement action. and reinforcement costs .

[0224] Gap Coverage Calculation Module: Used to calculate the set of reinforcement actions. Coverage of capacity gaps .

[0225] Cost Constraint Selection Module: Used to select a set of reinforcement actions based on cost minimization, coverage threshold, or maximization of growth attainability. .

[0226] Growth Score Calculation Module: Used to calculate based on coverage, reinforcement cost, and optional current matching score. .

[0227] Candidate output module: used to output growth attainable score, set of reinforcement actions, candidate job entry mark or sorting weight.

[0228] The following is an exemplary embodiment to help understand the technical solution of this application. The numerical values ​​in this embodiment are only used to illustrate the calculation process and do not limit the scope of protection of this application.

[0229] The competency dimensions include four aspects: Java backend skills, database skills, engineering skills, and internship experience. (Students) The current capability vector is:

[0230]

[0231] The capability requirement vector for target position j is:

[0232]

[0233] The capacity gap is:

[0234]

[0235] The enhancement action library includes three actions: For backend integrated projects, , ; For database course projects, , ; For corporate training, If you choose The remaining gap after reinforcement is:

[0236]

[0237] If the difficulty of each dimension is 1, then the set of reinforcement actions... It covers the vast majority of skill gaps; when costs are below a preset threshold and growth potential exceeds the threshold, the system will target specific positions. Mark the positions as attainable through growth and output a set of reinforcement actions {backend comprehensive project, database course project, enterprise training}.

[0238] The following is an exemplary embodiment to illustrate the complete calculation process of the student's current ability vector, job ability requirement vector, ability gap, set of reinforcement actions, coverage, reinforcement cost, current matching degree, growth attainable score, optimal subset of reinforcement actions, and candidate entry judgment in this application. The numerical values ​​in this embodiment are for illustrative purposes only and do not limit the scope of protection of this application.

[0239] Step A: Source and Calculation of Student's Current Ability Vector

[0240] Let the capability dimensions include four dimensions: Java backend capability, database capability, engineering capability, and internship experience, i.e., m=4. The current capability vector of student u. Instead of being given a direct value, the score is based on evidence from the student's courses or certificates. Project or practice evidence scoring Assessment or human evaluation scoring The scores are obtained by weighted normalization. Each score is first mapped to the range [0,1].

[0241] Among them, course or certificate evidence scoring It is derived from the mapping of course grades, course hours, certificate level, or certificate pass rate related to the k-th competency dimension; project or practice evidence score. It is derived from the number of projects, project difficulty, student role, project completion quality, or practical results related to the k-th ability dimension; assessment or manual evaluation score. The scores are derived from competency assessment scores, teacher evaluations, corporate mentor evaluations, or manual review results. All three types of scores are uniformly normalized to the range [0,1].

[0242] In one implementation, course grades can be normalized by dividing the course grade by 100; certificate levels can be normalized by certificate difficulty or pass / fail levels; project or practical evidence can be weighted by project completion, task complexity, and student contribution; and assessments or human evaluations can be obtained by mapping assessment scores or evaluation levels. The system obtains the corresponding values ​​for each ability dimension based on a preset correspondence between ability dimensions and evidence sources. , and .

[0243] The student's current ability value in the k-th ability dimension is calculated as follows:

[0244] A u,k =0.50× +0.30× +0.20×

[0245] The weights 0.50, 0.30, and 0.20 are preset evidence fusion weights, corresponding to course or certificate evidence, project or practical evidence, and assessment or human evaluation evidence, respectively. These weights are determined based on the verifiability, data stability, and direct representation of the competency dimension of different evidence sources. Course or certificate evidence typically has clear grades, credit hours, assessment results, or certificate levels, making it highly verifiable; therefore, its weight is set at 0.50. Project or practical evidence reflects students' ability to solve practical problems, but is affected by project scale, task difficulty, and completion quality; therefore, its weight is set at 0.30. Assessments or human evaluations supplement information not covered by other evidence, but are significantly influenced by the evaluation subject and evaluation standards; therefore, as supplementary evidence, its weight is set at 0.20. The sum of the three weights is 1, used to fuse student competency evidence from different sources into a current competency value under the same competency dimension.

[0246] In this embodiment, the evidence scores corresponding to each capability dimension are shown in Table 1.

[0247] Table 1:

[0248] Taking Java backend capabilities as an example, the system obtains course or certificate evidence scores from students' Java course grades, backend-related certificates, or course records:

[0249] =0.50

[0250] Scoring is based on project or practice evidence obtained from students' participation in backend projects, API development tasks, or code practice records:

[0251] =0.40

[0252] Assessment or manual evaluation scores are obtained from competency assessment results, teacher evaluations, or manual review results:

[0253] =0.40

[0254] Therefore, the current capability value for the Java backend capability dimension is:

[0255]

[0256] Similarly, database skills, engineering skills, and internship experience are calculated in the same way: .

[0257] Therefore, student u's current ability vector is:

[0258] (0.45, 0.40, 0.30, 0.10)

[0259] The values ​​of 0.45, 0.40, 0.30, and 0.10 in the above vectors are obtained by weighted normalization of the student's course or certificate records, project or practice records, and assessment or human evaluation results in the four competency dimensions, and are therefore not arbitrary values.

[0260] Step B: Source and Calculation of Job Competency Requirement Vector

[0261] The target position j is a Java backend developer. The required skill vector R is... j Job Description Analysis and Rating P j,k Job competency model configuration score Q j,k The scores are obtained by weighting. The job description analysis score is derived from the skill keywords, task descriptions, and job requirements in the job text; the job competency model configuration score is derived from the job competency tag library or a manually configured job competency model. Both are normalized to the range [0,1].

[0262] In one implementation, the job description parsing score P j,k The job competency model configuration score can be obtained by mapping the frequency of keywords related to the k-th competency dimension in the job description, the intensity of requirements, the importance of job conditions, or the relevance to job tasks; j,k The scores can be obtained based on the system's preset job competency model, job tag library, industry job standards, or manual configuration rules. The system maps the two types of scores to a unified [0,1] range and then performs a weighted fusion.

[0263] The required value for the position in the k-th competency dimension is calculated as follows:

[0264] R j,k =0.50×P j,k +0.50×Q j,k

[0265] In this embodiment, the job requirement scores corresponding to each capability dimension are shown in Table 2.

[0266] Table 2:

[0267] Taking Java backend capabilities as an example, the required job skills are as follows:

[0268]

[0269] Therefore, the capability requirement vector for target position j is:

[0270] =(0.70,0.65,0.55,0.40)

[0271] The values ​​of 0.70, 0.65, 0.55, and 0.40 in the above vectors are obtained by weighted normalization of the job text parsing results and the job competency model configuration results, and therefore are not arbitrary values.

[0272] Step C: Capacity Gap Calculation

[0273] The system is based on the student's current ability vector Vector of target job competency requirements Computing capacity gap:

[0274] Gap u,j,k =max( - ,0)

[0275] Substituting the values ​​from this embodiment, we get:

[0276] Gap u,j =max( - ,0)

[0277] Gap u,j =max((0.70, 0.65, 0.55, 0.40)−(0.45, 0.40, 0.30, 0.10),0)

[0278] Gap u,j =(0.25,0.25,0.25,0.30)

[0279] The four values ​​represent the skill gaps of student u relative to target position j in four dimensions: Java backend skills, database skills, engineering skills, and internship experience.

[0280] Step D: Reinforcement action set and its capability gains, cost sources

[0281] The set of reinforcement actions Ω in this embodiment includes three reinforcement actions: For backend integrated projects, For database course projects, For enterprise training. The capability gain vector of each reinforcement action is obtained from the historical capability changes before and after completion or the results of expert rule calibration, and the cost is obtained by normalized weighting of time, difficulty and resource costs.

[0282] The capability gain vectors for each enhancement action are shown in Table 3.

[0283] Table 3:

[0284] The four components of the capability gain vector correspond to Java backend capabilities, database capabilities, engineering capabilities, and internship experience, respectively. For example:

[0285] =(0.25, 0.05, 0.20, 0.00) indicates the completion of the back-end integrated project. Subsequently, it is estimated that students will improve by 0.25 in Java backend capabilities, 0.055 in database capabilities, and 0.20 in engineering capabilities, with no direct improvement in internship experience.

[0286] In one implementation, the capability gain vector can be obtained from the historical change in capability before and after completion. For example, if the historical average capability vector before completing the final integrated project is:

[0287]

[0288] The historical average capability vector after completing the integrated backend project is:

[0289]

[0290] The capability gain vector of this reinforcement action is:

[0291]

[0292] Therefore, the capability gain vector in this embodiment is not an arbitrary given value, but can be obtained from historical sample statistics, expert rule calibration, or prediction model output.

[0293] The cost of reinforcement is obtained by weighting time cost, difficulty cost, and resource cost:

[0294]

[0295] in, This represents the time cost of the i-th reinforcement action. Let represent the difficulty cost of the i-th reinforcement action. This represents the resource cost of the i-th reinforcement action. This indicates the system's preset maximum time cost. and These represent the minimum and maximum difficulty scores, respectively. This indicates the maximum resource cost preset by the system.

[0296] Cost normalization parameter selection Hour, , , Weighting , , .

[0297] The cost calculations for each reinforcement action are shown in Table 4:

[0298] Table 4:

[0299] For example, The cost is calculated as follows:

[0300]

[0301] Similarly, , .

[0302] Set the maximum allowable reinforcement cost The normalized cost of the candidate action subset Δ is:

[0303]

[0304] Step E: Calculate the current matching degree Match(u,j)

[0305] This embodiment uses a current matching degree calculation method based on a capability vector system. Let the capability dimension enhancement difficulty vector be:

[0306] D=(1.00,1.00,1.00,1.00)

[0307] The four components correspond to Java backend capabilities, database capabilities, engineering capabilities, and internship experience, respectively. In this embodiment, the difficulty level for each dimension is the same, so all are set to 1.00. In other implementations, different difficulty levels can be set for different capability dimensions based on historical learning time, task completion cycle, certificate acquisition cycle, expert rules, or historical sample statistics.

[0308] The current matching degree is calculated using the following formula:

[0309]

[0310] Here, ε is a constant to prevent the denominator from being zero. Since the denominator in this embodiment is not zero, the effect of ε on the calculation result can be ignored during the numerical demonstration.

[0311] In this embodiment:

[0312]

[0313] at the same time:

[0314]

[0315] therefore:

[0316]

[0317] Therefore, the current matching degree Match(u,j) is directly derived from the student's current ability vector. Job competency requirements vector Capacity gap The reinforcement difficulty vector D is calculated and has a direct interface relationship with the capability vector system of this application.

[0318] Step F: Calculate the coverage, cost, and growth attainability score of the candidate reinforcement action subset.

[0319] For any candidate reinforcement action subset Δ, the system first calculates the total capability gain TotalGain(Δ) of that subset, and then calculates the remaining gap after reinforcement. Then, the coverage (Cover(u,j,Δ)) is calculated.

[0320] The total capability gain of the candidate reinforcement action subset Δ in the k-th capability dimension is:

[0321]

[0322] The remaining gap after reinforcement is:

[0323]

[0324] Coverage is:

[0325]

[0326] In this embodiment, because:

[0327]

[0328] and:

[0329]

[0330] Therefore, the coverage can be written as:

[0331]

[0332] In this embodiment, the following is taken: Therefore, the weight of the cost item is:

[0333] The formula for calculating growth attainable points is:

[0334]

[0335] Substituting, we get:

[0336]

[0337] Since Match(u,j)=0.5435, therefore

[0338]

[0339] The calculations for each candidate reinforcement action subset are shown in Table 5.

[0340] Table 5:

[0341] With candidate subsets:

[0342]

[0343] For example, the calculation process is as follows.

[0344] First, calculate the total capability gain:

[0345]

[0346] Then calculate the remaining gap:

[0347]

[0348] Therefore, the coverage is:

[0349]

[0350] The total reinforcement cost for this subset is:

[0351] The normalized cost is:

[0352]

[0353] Growth can be divided into:

[0354]

[0355] Step G: Optimal reinforcement action subset selection and candidate entry judgment

[0356] This embodiment selects the optimal subset of reinforcement actions by maximizing the growth attainability score.

[0357]

[0358] Simultaneously set the maximum normalized cost threshold.

[0359] The calculation results in step F show that, when the following conditions are met:

[0360]

[0361] The candidate reinforcement action subset, The growth rate can reach the highest score, 0.5112, and Therefore, the system determines the optimal subset of reinforcement actions as follows:

[0362]

[0363] Right now:

[0364] The system further sets the growth achievable threshold. Maximum normalized cost threshold The candidate entry judgment rule is as follows:

[0365]

[0366] because:

[0367]

[0368] and:

[0369]

[0370] The system marks the target position j as a growth-attainable position and outputs the growth-attainable score of 0.5112, the optimal subset of reinforcement actions {backend comprehensive project, database course project, enterprise training}, and the candidate position entry flag Enter(u,j)=1.

[0371] Therefore, this embodiment fully demonstrates the source of the student's current ability vector and job ability requirement vector, the calculation of ability gap, the source of the ability gain vector and cost of reinforcement actions, the assignment of the reinforcement difficulty vector D, the numerical calculation process of Cover, Normalized CostNorm, Current Match, Growth Reach, as well as the optimal reinforcement action subset Δ* and candidate entry judgment process.

Claims

1. A method for calculating job development attainability based on competency gap coverage and the selection of a set of reinforcement actions, characterized in that, Includes the following steps: Step 1: Obtain the student's current ability vector, job ability requirement vector, and difficulty of improving ability dimensions; Step 2: Calculate the capability gap based on the student's current capability vector and the job capability requirement vector; Step 3: Obtain the set of reinforcement actions, wherein each reinforcement action in the set of reinforcement actions includes a capability gain vector and a reinforcement cost; Step 4: For the candidate reinforcement action subset, calculate the coverage of the candidate reinforcement action subset on the capability gap based on the capability gap, the capability dimension reinforcement difficulty, and the capability gain vector of each reinforcement action in the candidate reinforcement action subset; Step 5: Calculate the reinforcement cost of the candidate reinforcement action subset based on the reinforcement cost of each reinforcement action in the candidate reinforcement action subset; Step 6: Calculate the job growth achievable score based on the coverage and reinforcement costs; Step 7: Based on the job growth attainable score, select the optimal subset of reinforcement actions from the set of reinforcement actions, and output at least one of the job growth attainable score, the optimal subset of reinforcement actions, the candidate job entry flag, or the job ranking weight.

2. The method for calculating job development attainability based on capability gap coverage and the selection of a set of reinforcement actions, as described in claim 1, is characterized in that... Step 1 specifically includes: Step 11: Pre-build the set of capability dimensions 'm' represents the total number of capability dimensions; these capability dimensions include professional skills, tool skills, project practice ability, engineering ability, data analysis ability, communication and collaboration ability, certifications and qualifications, internship experience, and portfolio completeness; each capability dimension is represented by a normalized numerical value, with a value range of [0,1] or [0,100]: Where D represents the difficulty of enhancing the capability dimension. Indicates the first The difficulty of enhancing each capability dimension is determined by historical learning time, task completion cycle, certificate acquisition cycle, expert rules, or historical sample statistics. Step 12: Obtain Students Current capability vector and positions Capability requirement vector The student's current competency vector comes from resume analysis, course grades, project experience, competition experience, certificate information, internship experience, portfolio information, or manually entered competency evaluation results; the job competency requirement vector comes from job description analysis, job competency model, job tag library, or manual configuration. in, Students In the Current ability value in each ability dimension Indicates job position For the The required values ​​for each capability dimension, k∈(1,m).

3. The method for calculating job development attainability based on capability gap coverage and the selection of a set of reinforcement actions, as described in claim 1, is characterized in that... Step 2 specifically includes: based on the student's current ability vector and job competency requirements vector Calculate the student relative to the position Competency gaps in each competency dimension; a competency gap represents the portion of a job requirement that exceeds a student's current competency; if a student's current competency already meets or exceeds the job requirement, then the gap in that dimension is 0; in, This represents the skill gap between student u and job j in the k-th skill dimension; in, This represents the competency gap vector of student u relative to job j, composed of competency gaps across various competency dimensions. Students In the Current ability value in each ability dimension Indicates job position For the The required values ​​for each capability dimension, k∈(1,m); Step 2 decomposes the reasons for job mismatch from a whole matching into multiple calculable capability gaps, providing input for subsequent reinforcement action selection.

4. The method for calculating job development attainability based on capability gap coverage and the selection of a set of reinforcement actions, as described in claim 1, is characterized in that... Step 3 specifically includes: a preset set of reinforcement actions. ; This represents the nth reinforcement action; each reinforcement action This represents a course learning session, project practice, certificate acquisition, internship training, competition training, scientific research practice, or portfolio construction, where i∈(1,n); each reinforcement action must include at least the action type, action name, capability gain vector, reinforcement cost, and applicable conditions. in, Indicates the action type. Indicates the action name. Represents the capability gain vector. Indicates a supplementary action For the The expected improvement in each capability dimension, k∈(1,m); Indicates a supplementary action The cost of reinforcement is normalized from time cost, difficulty cost, resource cost, or comprehensive cost; This indicates the applicable conditions for the reinforcement action.

5. The method for calculating job growth attainability based on capability gap coverage and selection of reinforcement action sets according to claim 1, characterized in that, Step 4 specifically includes: for any subset of candidate reinforcement actions Calculate the total capability gain of the action subset in each capability dimension, and further calculate the remaining gap after reinforcement; in, Indicates a supplementary action The expected capability gain for the k-th capability dimension; The total capability gain of the candidate reinforcement action subset Δ in the k-th capability dimension is obtained by summing the capability gains of each reinforcement action in the subset in that capability dimension. This represents a set of reinforcement actions; This represents the remaining capability gap of student u in the k-th capability dimension that is still not covered after adopting the candidate reinforcement action subset Δ for job j; This represents the skill gap between student u and job j in the k-th skill dimension; in, This represents the coverage of the competency gap of student u relative to job j by the subset of candidate reinforcement actions Δ. Indicates the first The difficulty of enhancing each capability dimension This represents a set of reinforcement actions. To prevent constants with a denominator of zero; when students do not have a skill gap for position j, Set it to 1.

6. The method for calculating job growth attainability based on capability gap coverage and selection of reinforcement action sets according to claim 1, characterized in that, Step 5 specifically includes: cost calculation of the reinforcement action set: in, This represents the total reinforcement cost of the subset Δ of candidate reinforcement actions. Indicates a supplementary action The cost of reinforcement To preset the maximum allowable reinforcement cost or the maximum cost in the candidate set, Used to incorporate cost constraints into the calculation of subsequent growth attainable points. To prevent constants with a denominator of zero.

7. The method for calculating job growth attainability based on capability gap coverage and selection of reinforcement action sets according to claim 1, characterized in that, Step 6 specifically includes: calculating the job growth achievable score based on the gap coverage and reinforcement costs; in, Students Relative to the position In the set of reinforcement actions The growth rate can reach points, This represents the coverage of the competency gap of student u relative to job j by the subset of candidate reinforcement actions Δ; Students With position The current matching degree is calculated from existing matching models, rule matching, or vector similarity. Used to incorporate cost constraints into the calculation of subsequent growth attainable points; These are the weight parameters.

8. The method for calculating job development attainability based on capability gap coverage and the selection of a set of reinforcement actions, as described in claim 1, is characterized in that... Step 7 specifically includes: a set of reinforcement actions. Select reinforcement action subset The selection objective is to minimize reinforcement costs or maximize growth potential while satisfying the remaining gap or coverage constraints. Students Relative to the position In the set of reinforcement actions The growth potential is achievable; Δ represents a subset of candidate reinforcement actions. This represents a set of reinforcement actions; The set of reinforcement actions is selected using a combination of coverage constraints and cost minimization. in, This represents the total reinforcement cost of the subset Δ of candidate reinforcement actions. This indicates that after student u adopts the candidate reinforcement action subset Δ for position j, The allowable weighted residual gap threshold; if a certain set of reinforcement actions makes the weighted residual gap less than or equal to Then it is considered that the set of actions can enable the job Reaching a state where reinforcement is possible; among multiple sets of actions that satisfy the constraints, selecting the set of actions with the minimum total cost as the optimal set. ; Indicates constraints; A greedy selection approach is used to reduce computational load: First, the unit cost gap coverage benefit of each reinforcement action is calculated, and then the actions are added to the action set one by one from high to low benefit until the coverage reaches a preset threshold or the maximum allowable cost is reached. : in, Indicates the first The difficulty of enhancing each capability dimension Indicates a supplementary action The expected capability gain for the k-th capability dimension; This represents the skill gap between student u and job j in the k-th skill dimension; Indicates a supplementary action The cost of reinforcement; To prevent constants with a denominator of zero.

9. The method for calculating job development attainability based on capability gap coverage and the selection of a set of reinforcement actions, as described in claim 8, is characterized in that... Candidate position entry or ranking weight output Where Δ* represents the optimal subset of reinforcement actions determined according to the principles of coverage constraints, cost constraints, or maximizing growth attainability. This indicates the job growth score that student u can achieve for job j after adopting the optimal reinforcement action subset Δ*; This represents the total reinforcement cost of the optimal reinforcement action subset Δ*; in, For growth to reach a threshold, To the maximum reinforcement cost threshold, For students The set of candidate positions, The base recommendation score output by the existing recommendation system. To introduce ranking weights for growth reachability, and These are weight parameters, used to adjust the influence of the base recommendation score and the growth attainable score on the ranking weights, respectively.

10. A job development attainability calculation system based on competency gap coverage and selection of reinforcement action sets, characterized in that, include: Ability Input Module: Used to obtain the student's current ability vector. Job competency requirements vector and the difficulty of improving capabilities ; Capacity gap calculation module: used to calculate based on Computing capacity gap; This represents the skill gap between student u and job j in the k-th skill dimension; Students In the Current ability value in each ability dimension Indicates job position For the The required values ​​for each capability dimension, k∈(1,m); Reinforcement Action Modeling Module: Used to store and retrieve reinforcement action sets. And configure a capability gain vector for each reinforcement action. and reinforcement costs ; Gap Coverage Calculation Module: Used to calculate the set of reinforcement actions. Coverage of capacity gaps ; Cost Constraint Selection Module: Used to select a set of reinforcement actions based on cost minimization, coverage threshold, or maximization of growth attainability. ; Growth Score Calculation Module: Used to calculate based on coverage, reinforcement cost, and optional current matching score. ; Candidate output module: used to output growth attainable score, set of reinforcement actions, candidate position entry mark or ranking weight; The system operates a job growth attainability calculation method based on capability gap coverage and reinforcement action set selection, as described in any one of claims 1 to 9.

Images