Post recommendation method and device, storage medium and product

Abstract

The invention provides a post recommendation method and device, a storage medium and a product, and relates to the technical field of software, and the method comprises the steps: receiving a post recommendation request message sent by a client; acquiring employee data of the to-be-recommended employees from a database based on the employee identifiers; determining an adaptation index of at least one candidate post based on the employee data and the post demand data; generating matching scores of the candidate posts based on the adaptation indexes of the candidate posts; determining a target post from the at least one candidate post based on the matching score of each candidate post; and sending a post recommendation response message to the client. According to the method, on the premise of not additionally increasing the data acquisition cost and being compatible with the existing human resource system architecture of the enterprise, efficient adaptive recommendation of flexible employment posts and employees in the enterprise is realized, and the recommendation accuracy and employment configuration efficiency are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of software technology, and in particular to a job recommendation method, device, storage medium, and product. Background Technology

[0002] With the increasing popularity of flexible employment models in enterprises, the precise matching of internal temporary positions with employees has become a core requirement for improving employment efficiency.

[0003] In current technologies, the recommendation of flexible staffing positions within enterprises relies heavily on manual screening or simple keyword matching, which is inefficient, especially when there are many employees and job requirements change frequently, making it difficult to respond quickly. In addition, existing technologies are unable to respond to the dynamic needs of enterprises for flexible staffing, thus affecting the efficiency of staffing and the effectiveness of project execution. Summary of the Invention

[0004] This application provides a job recommendation method, device, storage medium, and product that can achieve efficient matching and recommendation of flexible employment positions and employees within an enterprise without increasing data collection costs or compatibility with the enterprise's existing human resources system architecture, thereby significantly improving recommendation accuracy and staffing efficiency.

[0005] To address the above problems, the embodiments of this application provide the following technical solutions: Firstly, this application provides a job recommendation method, which includes: Receive a job recommendation request message from the client; retrieve employee data of the employees to be recommended from the database based on employee identifiers; determine the suitability indicators for at least one candidate job based on the employee data and job requirement data; generate matching scores for the candidate jobs based on the suitability indicators; determine the target job from at least one candidate job based on the matching scores of each candidate job; and send a job recommendation response message to the client.

[0006] The job requirement data includes the requirement information for at least one candidate job; employee data includes basic employee information, skill data, and performance data; the matching score is used to characterize the suitability between the employee to be recommended and the candidate job; and the job recommendation response message includes the job information for the target job.

[0007] Based on the aforementioned technical means, a suitability assessment system is constructed by integrating the basic information, skill data, and performance data of employees to be recommended, combined with the explicit requirements of candidate positions. This avoids the bias in suitability judgments caused by a single data dimension. Suitability indicators quantify the degree of fit between employees and positions, generating objective and comparable matching scores. This allows for precise measurement of suitability and reduces interference from subjective decision-making. Target positions are selected based on the matching scores and fed back to the client, ensuring that the recommended results not only match the current capabilities of employees but also align with the actual needs of the positions. This significantly improves the accuracy and efficiency of job recommendations, effectively reduces the cost of human resource mismatch for enterprises, and adapts to diverse job configuration scenarios such as flexible employment and cross-departmental collaboration within enterprises.

[0008] In one possible implementation, the suitability metrics for candidate positions include: feature similarity, filtering score, and prediction score. Feature similarity is used to characterize the degree of fit between the employee to be recommended and the candidate position. Filtering score is used to characterize the degree of preference of similar employees for the candidate position. Similar employees are other employees within the company whose employee data and the employee data of the employee to be recommended have a similarity that meets a preset standard. Prediction score is used to characterize the degree to which the employee to be recommended meets the performance standards for the candidate position.

[0009] Based on the aforementioned technical means, by limiting the matching indicators to feature similarity, filtering score, and prediction score, the limitations of judging by a single indicator are avoided. Feature similarity ensures a direct match between employees and the hard requirements of the position. Filtering score provides practical reference by leveraging the real preferences and performance verification of employees with high similarity. Prediction score uses the model to predict the potential of employees to meet job performance standards. This allows the matching assessment to not only match customer characteristics but also improve the comprehensiveness and reliability of the matching score, thereby making the job recommendation results more in line with the employee's abilities and the job requirements.

[0010] In one possible implementation, based on employee data and job requirement data, at least one suitability index for a candidate job is determined. This can be specifically implemented as follows: feature extraction is performed on the employee data and the requirement information of the candidate jobs to obtain the features of the employees to be recommended and the features of the candidate jobs; based on the features of the employees to be recommended and the features of the candidate jobs, a filtering score and a prediction score are determined; based on the features of the employees to be recommended and the features of similar employees, the feature similarity is determined.

[0011] Based on the aforementioned technical methods, feature extraction is first performed on employee data and job requirement information to transform fragmented data into standardized, quantifiable features. This lays a unified and objective foundation for calculating matching indicators. Then, calculation inputs are allocated according to a targeted logic of filtering score, prediction score, and feature similarity. The first two indicators are determined based on employee and job characteristics, while feature similarity is determined based on the characteristics of the employee to be recommended and similar employees. This ensures that the calculation of each indicator aligns with the core requirements and avoids biases caused by input confusion. This process design makes the determination logic of matching indicators clear and highly targeted, effectively improving the accuracy and reliability of each indicator's results and enhancing the accuracy of job recommendations.

[0012] In one possible implementation, the filtering score is determined as follows: Target similar employees whose similarity meets a preset standard are selected from similar employees, where the preset standard is used to determine whether the feature fit of the recommended employee meets the requirements; job selection data of the target similar employees is extracted, which consists of candidate job data where the target similar employees actively applied for or accepted recommendations, and whose performance scores after starting work meet a preset scoring standard, where the preset scoring standard is used to determine whether the employee's work performance after starting work meets the requirements; and the candidate job filtering score is determined based on the job selection data.

[0013] Based on the aforementioned technical means, by first screening target similar employees whose employee similarity meets the preset standard, it is ensured that the characteristics of the reference group and the employees to be recommended are highly consistent, avoiding interference from low-related employee data. Then, the job selection data of the target similar employees is extracted. This data has both real preferences and practical verification attributes, eliminating invalid choices and low-performance cases, ensuring the reliability of the reference basis. Based on this high-quality data, a filtering score is determined, avoiding blind preferences and interference from invalid data, accurately reflecting the actual suitability value of the candidate positions, strengthening the reference significance of the filtering score, complementing other suitability indicators, and improving the comprehensiveness and accuracy of the overall suitability assessment.

[0014] In one possible implementation, the predicted score is determined as follows: the features of the employee to be recommended are concatenated with the features of the candidate positions to obtain combined features; the combined features are input into the prediction model, and a prediction is made based on the prediction model to obtain a predicted score; the prediction model predicts the degree to which the employee to be recommended will meet the performance standards in the candidate positions based on the combined features, and outputs the corresponding quantitative predicted score.

[0015] Based on the aforementioned technical means, by splicing the characteristics of the employee to be recommended with the characteristics of the candidate position, the core traits of both are integrated to form a high-dimensional combined feature, providing a comprehensive and relevant input basis for predicting job performance potential. This avoids the lack of information from a single feature dimension. The combined feature is quantitatively analyzed using a predictive model to accurately predict the employee's performance level in the target position, outputting a standardized predictive score. This replaces subjective experience judgment, improves the objectivity and accuracy of the assessment, and transforms the correlation between employee ability and job requirements into a calculable quantitative result. This complements feature similarity and filtering scores, enriching the dimensions and depth of the matching assessment, making the matching score more reflective of actual job performance potential, thereby enhancing the accuracy of job recommendations and effectively reducing employment risks caused by mismatch in job performance capabilities.

[0016] In one possible implementation, generating a matching score for a candidate position based on its suitability indicators can be achieved as follows: Historical matching data for the candidate position is retrieved from a database. This historical matching data includes: job requirement data, matching records, suitability indicators, and employee performance ratings after onboarding for similar positions. The historical recommendation accuracy rate of the candidate position's suitability indicators is determined based on this data. This historical recommendation accuracy rate characterizes the predictive reliability of the suitability indicators for similar positions. Weighting coefficients for the candidate position's suitability indicators are determined based on the historical recommendation accuracy rate. Finally, a weighted calculation is performed based on the weighting coefficients and the suitability indicators to obtain the candidate position's matching score.

[0017] Based on the aforementioned technical means, by acquiring historical matching data of similar positions for candidate positions, data support is provided for the reliability assessment of matching indicators. The historical recommendation accuracy of each matching indicator is calculated based on the historical matching data, which can dynamically reflect the predictive reliability of the indicator for similar positions. Then, the weight coefficient is determined according to the historical recommendation accuracy, so that highly reliable indicators receive higher weights, avoiding matching bias caused by fixed weights, and further enhancing the accuracy of job recommendations.

[0018] In one possible implementation, the target position is determined from at least one candidate position based on the matching score of each candidate position, including: identifying candidate positions with matching scores greater than a score threshold as preliminary screening positions, where the score threshold is used to determine whether the candidate positions meet the basic suitability requirements; verifying the preliminary screening positions based on candidate conditions, and identifying the preliminary screening positions that pass the verification as target positions, wherein the candidate conditions are used to verify whether the preliminary screening positions match the actual work status of the employees to be recommended.

[0019] Based on the aforementioned technical methods, preliminary job postings are screened using scoring thresholds to ensure that candidate positions meet basic suitability requirements and filter out obviously mismatched options. Then, based on the candidate criteria, the actual work status of employees is verified to eliminate situations where manpower capacity does not match. This ensures that the recommended results not only meet the hard suitability standards but also align with the actual work conditions of employees, significantly improving the accuracy of target job postings. This effectively avoids employment mismatch issues caused by conflicting work statuses, making job recommendations more closely aligned with the company's actual staffing needs and reducing manpower allocation risks.

[0020] In one possible implementation, the job recommendation method provided in this application can also be specifically implemented as follows: based on the updated employee data and the updated job requirement data, the target job is updated. The updated employee data includes: updated skill data, performance change data, and basic information change data of the employees to be recommended. The updated job requirement data includes: adjusted requirement data, urgency change data, and status change data of the candidate job.

[0021] Based on the aforementioned technical means, by responding to the dynamic changes in employee data and job requirement data, the target job is updated based on the updated skill performance information and job requirement adjustments, etc., to avoid the recommendation results from being out of touch with actual needs due to data lag, so that job recommendations continuously match the current ability of employees and the actual requirements of the job, effectively avoiding the risk of human resource mismatch.

[0022] In one possible implementation, the target position is updated based on the updated employee data and the updated job requirement data. Specifically, this can be achieved by: determining the update matching score based on the updated employee data and the updated job requirement data; and updating the target position based on the update matching score.

[0023] Based on the aforementioned technical methods, an updated matching score is recalculated using updated employee data and job requirement data to ensure the timeliness and accuracy of the suitability assessment. The target job is then updated based on the updated matching score, preventing recommendations from becoming out of touch with actual needs due to data changes, and ensuring that job recommendations continuously align with employees' current capabilities and the dynamic requirements of the job. This approach enhances the dynamic adaptability of the recommendation results, effectively mitigating the risk of human resource mismatch caused by lag, improving the flexibility and accuracy of corporate human resource allocation, and adapting to dynamic adjustment scenarios for flexible employment.

[0024] Secondly, this application provides a job recommendation device, which includes: an acquisition module, a determination module, and a sending module.

[0025] The acquisition module is used to receive job recommendation request messages sent by the client. The job recommendation request message includes at least the employee identifier of the employee to be recommended.

[0026] The acquisition module is also used to retrieve employee data of employees to be recommended from the database based on employee identifiers.

[0027] The determination module is used to determine the suitability indicators for at least one candidate position based on employee data and job requirement data. The employee data includes basic employee information, skill data, and performance data; the job requirement data includes the requirement information for at least one candidate position.

[0028] The determination module is also used to generate a matching score for candidate positions based on the suitability indicators of the candidate positions; the matching score is used to characterize the degree of suitability between the recommended employees and the candidate positions.

[0029] The determination module is also used to determine the target position from at least one candidate position based on the matching score of each candidate position. Sending module: Used to send job recommendation response messages to the client. The job recommendation response message includes the job information of the target job.

[0030] The determination module is also used to determine the suitability indicators for candidate positions. These indicators include feature similarity, filtering score, and prediction score. Feature similarity is used to characterize the degree of fit between the employee to be recommended and the candidate position. The filtering score is used to characterize the degree of preference of similar employees for the candidate position. Similar employees are other employees within the company whose employee data and the employee data of the employee to be recommended have a similarity that meets a preset standard. The prediction score is used to characterize the degree to which the employee to be recommended meets the performance standards for the candidate position.

[0031] The determination module is also used to extract features from employee data and candidate job requirements information to obtain the features of the employees to be recommended and the features of the candidate jobs. Based on the features of the employees to be recommended and the features of the candidate jobs, the filtering score and the prediction score are determined. Based on the features of the employees to be recommended and the features of similar employees, the feature similarity is determined.

[0032] The determination module is also used to determine the filtering score, including: screening target similar employees whose employee similarity reaches a preset standard from similar employees, wherein the preset standard is used to judge whether the feature fit of the recommended employees meets the requirements; extracting job selection data of target similar employees, which is the data of candidate positions that target similar employees actively applied for or accepted recommendations for and whose performance scores after taking up the post meet the preset scoring standard, wherein the preset scoring standard is used to judge whether the employee's work performance after taking up the post meets the requirements; and determining the candidate position filtering score based on the job selection data.

[0033] The determination module is also used to determine the prediction score, including: concatenating the features of the employee to be recommended with the features of the candidate position to obtain the combined features; inputting the combined features into the prediction model, making a prediction based on the prediction model to obtain the prediction score; the prediction model predicts the degree of performance of the employee to be recommended in the candidate position based on the combined features, and outputs the corresponding quantitative prediction score.

[0034] The determination module is also used to retrieve historical matching data for candidate positions from the database. This historical matching data includes: job requirement data, matching records, fit indicators, and employee performance ratings after onboarding for similar positions. Based on this historical matching data, the module determines the historical recommendation accuracy rate of the fit indicators for candidate positions. The historical recommendation accuracy rate characterizes the reliability of the fit indicators' predictions for similar positions. Based on the historical recommendation accuracy rate, the module determines the weighting coefficients of the fit indicators for candidate positions. Finally, a weighted calculation is performed based on the weighting coefficients and the fit indicators to obtain the matching score for the candidate position.

[0035] The determination module is also used to identify candidate positions with matching scores greater than a score threshold as preliminary screening positions. The score threshold is used to determine whether the candidate positions meet the basic suitability requirements. Based on the candidate conditions, the preliminary screening positions are verified, and the preliminary screening positions that pass the verification are identified as target positions. The candidate conditions are used to verify whether the preliminary screening positions match the actual work status of the employees to be recommended.

[0036] The job recommendation method provided in this application also includes an update module, which is used to update the target job based on the updated employee data and the updated job requirement data. The updated employee data includes: skill update data, performance change data, and basic information change data of the employees to be recommended. The updated job requirement data includes: requirement adjustment data, urgency change data, and status change data of the candidate job.

[0037] The update module is also used to determine the update matching score based on the updated employee data and the updated job requirement data; and to update the target job based on the update matching score.

[0038] Thirdly, this application provides an electronic device comprising a processor and a memory. The memory stores processor-executable instructions, and when the processor is configured to execute the instructions, the electronic device implements the method described in the first aspect above.

[0039] Fourthly, this application provides a readable storage medium comprising software instructions. When the software instructions are executed in an electronic device, they cause the electronic device to implement the method described in the first aspect above.

[0040] Fifthly, this application provides a computer program product comprising computer instructions. When the computer instructions are executed on an electronic device, the electronic device performs the method described in the first aspect.

[0041] The beneficial effects of the second to fifth aspects mentioned above can be referred to the first aspect, and will not be repeated here. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of the structure of a job recommendation system provided in an embodiment of this application; Figure 2 This is a schematic diagram of a server cluster structure provided in an embodiment of this application; Figure 3 A flowchart illustrating a job recommendation method provided in an embodiment of this application; Figure 4 A flowchart illustrating yet another job recommendation method provided in this application embodiment; Figure 5 A flowchart illustrating yet another job recommendation method provided in this application embodiment; Figure 6 A flowchart illustrating yet another job recommendation method provided in this application embodiment; Figure 7 This is a schematic diagram of the structure of a job recommendation device provided in an embodiment of this application; Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0044] Hereinafter, the terms "first," "second," and "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first," "second," or "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this embodiment, unless otherwise stated, "a plurality of" means two or more.

[0045] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0046] In addition, for ease of understanding, the technical terms involved in the embodiments of this application will be introduced below.

[0047] For example, as enterprises accelerate their digital transformation, flexible employment models are becoming increasingly popular in internal collaborations, and the demand for cross-departmental temporary projects and short-term tasks continues to increase, highlighting the characteristics of diversified employee skills and dynamic job requirements.

[0048] In related technologies, job recommendation methods often rely on single-dimensional skill surface matching, failing to fully integrate multi-dimensional data such as employee performance and similar employee preferences. They also lack a computational mechanism for dynamic optimization based on historical matching results. Furthermore, they are limited by the computing power of ordinary servers, making it difficult to efficiently process large-scale feature data. This results in low matching accuracy of recommendation results, failing to meet the dynamic configuration needs of enterprises for flexible staffing. Consequently, problems such as job vacancies and human resource mismatch arise, affecting employment efficiency and project execution quality.

[0049] Against this backdrop, how to achieve accurate and real-time matching of positions and employees, and improve the efficiency of flexible staffing within enterprises, has become an urgent problem to be solved.

[0050] Therefore, in order to overcome the above problems, this application provides a job recommendation method. By integrating employee basic information, skill data, performance data and job requirement information, a matching score is generated based on the suitability index to accurately represent the degree of fit between employees and jobs, efficiently screen target jobs that meet the requirements, improve the accuracy and effectiveness of job recommendations, avoid human resource mismatch, and adapt to the flexible employment scenarios of enterprises.

[0051] The job recommendation method provided in this embodiment will be described below, starting with an introduction to the relevant technologies.

[0052] The job recommendation method provided in this application embodiment can be applied to, for example, Figure 1 The job recommendation system shown Figure 1 This is a schematic diagram of the structure of a job recommendation system provided in an embodiment of this application. Figure 1 The job recommendation system shown includes: client 100 and server cluster 101.

[0053] The client 100 and the server cluster 101 communicate with each other through a preset network interface.

[0054] It should be understood that the above communication connection methods can be flexibly adjusted according to the enterprise deployment scenario. For example, private network deployment can be used to improve the intranet transmission efficiency, or cloud-native interfaces can be used to adapt to the hybrid cloud architecture. As long as secure and efficient data interaction between components can be achieved, it is not limited to specific protocols or transmission methods.

[0055] Client 100 can be a terminal device or application used by enterprise employees to initiate job recommendation requests, receive recommendation results, and provide feedback on operational intentions. This includes internal enterprise office software, mobile management software, or applications. It is used to communicate with server cluster 101 to complete the initiation of job recommendation requests, receipt of results, and feedback interaction.

[0056] Specifically, Client 100 allows employees to log in to the company's internal system using an account and password. After logging in, the system automatically associates the employee ID or allows employees to manually enter the employee ID to initiate a job recommendation request. After receiving the job recommendation response message returned by the server, it displays detailed information such as the name of the target job, its business line, skill requirements, working hours, matching score, and urgency in a list format. It provides operation buttons such as accept, reject, and favorite. After an employee triggers an operation, Client 100 will synchronize the operation feedback to the server cluster 101 in real time. At the same time, it also allows employees to view their own skill data, performance data, and historical recommendation records.

[0057] It should be understood that the specific form of Client 100 can be adjusted according to the usage habits of enterprise employees, and is not limited to the types listed above. As long as it can realize communication and interaction with the server cluster and core functions, it can be integrated into the enterprise's existing office system to reduce the usage threshold for employees.

[0058] Server cluster 101 can be a collaborative processing unit composed of multiple dedicated functional servers. It is the core data processing and computing hub of the job recommendation system, used to coordinate the collection, preprocessing, storage scheduling, model calculation and recommendation result generation of raw data related to job recommendation, and provide full-process technical support for client 100.

[0059] It should be understood that the deployment form of the server cluster 100 can be flexibly adjusted. It can be a local cluster composed of physical servers, a distributed cluster built on cloud servers, or a hybrid architecture of physical machines and virtual machines. The hardware configuration of each server can be dynamically expanded according to the enterprise's data scale and computing needs, as long as the functional division and collaborative interaction of each server can be realized to ensure the efficient operation of the system, and it is not limited to a specific deployment form or hardware specifications.

[0060] Specifically, the server cluster 101 includes a human resources HR system server 1001, a data server 1002, and a graphics processing unit (GPU) server 1003.

[0061] like Figure 2 As shown, Figure 2 This is a schematic diagram of a server cluster structure provided in an embodiment of this application.

[0062] The HR system server 1001, data server 1002, and graphics processor (GPU) server 1003 communicate with each other through a preset network interface.

[0063] Among them, HR system server 1001 is a dedicated server for deploying the enterprise human resource management system. It is used to collect original employee data and original job requirement data within the enterprise, perform preprocessing such as cleaning, deduplication, standardization and quantification on the original data, generate structured data and synchronize it to data server 102 for storage. At the same time, it receives historical matching data fed back by data server 1002 to optimize the data collection dimensions.

[0064] It should be understood that the data collection methods of the HR system server 1001 may include manual entry, automatic synchronization with the enterprise attendance system / project management system, etc. The preprocessing rules can be customized and adjusted according to the enterprise's business needs. Its core function is to provide standardized and complete basic data for subsequent recommendation calculations.

[0065] Data server 1002 is a dedicated server for data storage and scheduling. It stores structured employee data, structured job requirement data, and employee-job historical matching data synchronized by HR system server 1001. It responds to computing data retrieval requests from GPU server 1003, provides historical data required for model training and matching calculations, and responds to job information retrieval requests in the recommendation result generation stage, outputting detailed information of the target job.

[0066] It should be understood that the data server 102 can be implemented using a distributed storage architecture or a relational database. The retention period for historical matching data is configurable. Its core function is to serve as the data hub of the system, ensuring reliable data storage and efficient scheduling.

[0067] The GPU server 1003 can be a dedicated computing server equipped with a graphics processor. Its core is used to perform high-density parallel computing tasks. In this application, it is used to train a fusion model based on historical matching data provided by the data server 1002. It receives the extracted employee feature vector and job feature vector, performs calculations through the pre-trained fusion model, generates matching scores for candidate jobs, and sends the calculation results back to the recommendation result determination stage to support the selection of target jobs.

[0068] like Figure 3 As shown, Figure 3 The following is a flowchart illustrating a job recommendation method provided in an embodiment of this application. The specific steps of the job recommendation method provided in this embodiment are as follows: S301: Receives a job recommendation request message sent by the client.

[0069] The job recommendation request message should include at least the employee identifier of the employee to be recommended, and may also include auxiliary information such as the recommendation scenario and the desired job type, depending on actual needs, to help filter candidate jobs.

[0070] The employees to be recommended are current employees within the company who have job matching needs. They can be employees who actively initiate the recommendation request through the client, or employees whose recommendation requests are initiated by department administrators or HR personnel.

[0071] Employee identification includes unique identification information assigned to employees within the company, such as employee ID, unique identity identifier, and corporate email account, which is used to link relevant data of the employee to be recommended.

[0072] S302: Retrieve employee data of the employees to be recommended from the database based on employee identifiers.

[0073] The aforementioned database is a structured database stored by data server 1002 in the server cluster. The employee data in this database is generated synchronously after the HR system server 1001 collects raw data and preprocesses it, and has the characteristics of integrity and standardization.

[0074] The aforementioned employee data includes basic employee information, skills data, and performance data. Specifically, basic employee information includes employee education, length of service, department, and business line affiliation; skills data includes the names of skills possessed by the employee, skill mastery levels, skill certification certificates, and practical project records; and performance data includes ratings of recently participated projects, the percentage of high-performing projects, and experience in executing similar projects.

[0075] S303: Based on employee data and job requirement data, determine the suitability indicators for at least one candidate job.

[0076] The job requirement data includes the requirement information for at least one candidate job position.

[0077] Job requirements information outlines the core recruitment criteria and performance requirements for candidate positions, clarifying the suitability of employees for each position. This includes: skill requirements, basic thresholds, performance conditions, business line affiliation, and a brief description of job responsibilities.

[0078] Specifically, the skill requirements mentioned above may include the skill type and the minimum mastery level for each skill.

[0079] The aforementioned basic thresholds may include educational requirements, minimum years of work experience, and experience working on the same project.

[0080] The aforementioned execution conditions may include working hours, workload intensity, and job urgency.

[0081] S304: Generate matching scores for candidate positions based on their suitability metrics.

[0082] The matching score is used to characterize the degree of fit between the recommended employee and the candidate position.

[0083] The suitability metrics for candidate positions are the core parameters for evaluating the suitability between employees and positions, specifically including: feature similarity, filtering score, and prediction score.

[0084] The aforementioned feature similarity is used to characterize the degree of fit between the recommended employee and the candidate position.

[0085] The filtering scores mentioned above are used to characterize the degree of preference of similar employees for candidate positions.

[0086] The aforementioned similar employees are other employees within the company whose employee data and the employee data of the employee to be recommended meet the preset standard of similarity.

[0087] The aforementioned preset standard is that the cosine similarity of the feature vectors corresponding to the employee data is greater than or equal to a preset threshold, ensuring that the selected similar employees have similar matching potential to the employees to be recommended.

[0088] It should be understood that the preset thresholds here can be customized and adjusted according to the needs of the enterprise, and this application does not impose any restrictions on them.

[0089] The predicted scores are used to characterize the degree to which the recommended employees meet the performance standards for the candidate positions.

[0090] S305: Based on the matching scores of each candidate position, determine the target position from at least one candidate position.

[0091] S306: Send a job recommendation response message to the client.

[0092] The aforementioned job recommendation response message includes job information for the target position, specifically including the job title, department, business line affiliation, detailed skill requirements, basic threshold, working hours, workload intensity, job urgency, matching score, job responsibilities, and a brief description of the job content. This allows employees seeking recommendations to quickly understand the core information of the position and make a decision to accept or decline.

[0093] Based on the above embodiments, this application integrates employee basic information, skills data, performance data, and job requirement information, generates matching scores based on adaptation indicators, accurately represents the degree of employee and job fit, efficiently filters target jobs that meet the requirements, improves the accuracy and effectiveness of job recommendations, avoids human resource mismatch, and adapts to the flexible employment scenarios of enterprises.

[0094] Please see Figure 4This application provides a flowchart illustrating another job recommendation method. The specific steps of the job recommendation method provided in this application are as follows: S401: Receives a job recommendation request message sent by the client.

[0095] S402: Retrieve employee data of the employees to be recommended from the database based on employee identifiers.

[0096] S403: Extract features from employee data and candidate job requirements to obtain the features of employees to be recommended and the features of candidate jobs.

[0097] In one possible implementation, feature extraction is performed on employee data and candidate job demand information to obtain the features of the employees to be recommended and the features of the candidate jobs, including: Step 1: Obtain the feature vector of the employee to be recommended, denoted as The feature vectors of candidate positions are denoted as Both vectors have the same dimension and are standardized numerical vectors. Step II: Calculate the dot product of the two vectors, i.e. Where i is the feature dimension index; Step III: Calculate the L2 magnitude of each vector, i.e. , ; Step IV: Calculate using the cosine similarity formula:

[0098] The results are mapped to a 0-1 score system, with higher scores indicating a stronger fit between the employee and the core characteristics of the position.

[0099] S404: Determine the filtering score and prediction score based on the characteristics of the employees to be recommended and the characteristics of the candidate positions.

[0100] As one possible approach, determining the above filtering score includes the following steps: Step 1: Select target similar employees from similar employees whose employee similarity reaches the similarity threshold.

[0101] Specifically, determining employee similarity includes: preprocessing employee data of employees who are not recommended and employee data of other employees within the company, extracting and quantifying the feature dimensions corresponding to basic information, skills data, and performance data, generating feature vectors of the employee to be recommended and feature vectors of other employees, and calculating the cosine similarity between the two vectors. This cosine similarity is the employee similarity, which is used to characterize the degree of fit between the two in the core feature dimensions.

[0102] The similarity threshold is used to determine whether the feature fit of the employee to be recommended meets the requirements. Specifically, the similarity threshold is the cosine similarity critical value between the employee feature vector and the feature vector of similar employees.

[0103] It should be understood that the above similarity threshold can be calibrated and adjusted based on the accuracy of the company's historical matching data. The logic behind setting the threshold is to screen a group of employees who are highly compatible with the employees to be recommended in terms of skill structure, performance level, and basic conditions, so as to avoid the preference data of employees with low similarity from interfering with the accuracy of the filtering score.

[0104] Step II: Extract job selection data from employees with similar goals.

[0105] Among them, the job selection data consists of candidate jobs that employees with similar goals actively apply for or accept recommendations for, and whose performance scores meet the preset scoring standards after taking up the job.

[0106] The specific data for job selection includes: job identifier, business line, skill requirement list, job type, timestamp of similar employees selected, project performance score after onboarding, and matching degree between job requirements and similar employees' skills.

[0107] The aforementioned pre-set scoring criteria are used to determine whether an employee's work performance meets the requirements after starting work.

[0108] For example, the standard is a unified job performance scoring system within the enterprise, using a 5-point system with 1 point being the lowest and 5 points being the highest. The preset scoring threshold is 4 points, that is, a performance score of ≥4 points is considered to meet the standard; for core business lines or highly complex positions, the threshold can be dynamically adjusted to 4.5 points.

[0109] Step 3: Determine the filtering score for candidate positions based on the job selection data.

[0110] Specifically, a weighted summation algorithm is used to calculate the filtering score, which includes: counting the frequency of employees with similar goals selecting the candidate position or similar positions, and assigning weights according to the frequency ratio; extracting the average performance score of those who achieved the target after selection, and assigning weights according to the mean range, with higher mean values ​​having higher weights; calculating the time difference between the selection timestamp and the current recommendation time, with more recent selections assigned higher weights; and then weighting the above three items and standardizing them to a 0-1 score scale to obtain the filtering score for the candidate position.

[0111] For example, if a candidate position is selected by 10 employees with similar goals, with a frequency of 30%, an average performance score of 4.2 (corresponding to a weight of 35%), and a selection rate of 80% within 6 months (corresponding to a weight of 16%), then the weighted sum is 81%, and the standardized filtering score is 0.81.

[0112] It should be understood that the filtering score provides a reference for the job suitability of recommended employees by using similar employee selection behaviors with actual performance verification. This avoids the limitations of relying solely on model prediction or surface feature matching. The job selection data all come from historical matching records stored on the data server 1002, which has traceability and authenticity, ensuring that the filtering score can objectively reflect the actual suitability value of the candidate job and further improve the accuracy of the recommendation results.

[0113] Based on the above embodiments, this application determines the filtering score by screening target similar employees whose employee similarity reaches a threshold, extracting job selection data of those employees who actively choose the job and whose performance meets the standards after taking up the post, ensuring that the preference reference is based on the effective practices of highly compatible employees, reducing interference from invalid data, making the filtering score more valuable, and thus improving the accuracy of job recommendations.

[0114] As one feasible approach, determining the aforementioned prediction score involves the following steps: Step 1: Combine the characteristics of the employee to be recommended with the characteristics of the candidate positions to obtain the combined features.

[0115] Among them, the combined feature is a high-dimensional feature vector formed by concatenating the employee feature vector of the employee to be recommended with the job feature vector of the candidate job in dimensional order.

[0116] In one possible implementation, the features of the employee to be recommended can be combined with the features of the candidate positions in the following order: employee basic information features - skill features - performance features - basic job requirements features - job skill requirements features - job performance conditions features, ensuring that the feature dimensions correspond and are not redundant.

[0117] Step II: Input the combined features into the prediction model, make predictions based on the prediction model, and obtain the prediction score.

[0118] The prediction model is based on combined features to predict the degree to which the recommended employee will meet the performance standards for the candidate position and outputs the corresponding quantitative prediction score.

[0119] Specifically, the prediction model is a multilayer perceptron (MLP) deep learning model trained on the enterprise's internal employee-job historical matching dataset. The input in the training data is historical combined features, and the label is whether the employee's performance after taking up the post meets the standard. The model extracts feature association information through hierarchical fully connected layers and outputs a score-based fit probability value as the prediction score.

[0120] The multilayer perceptron (MLP) deep learning model includes an input layer, at least three fully connected hidden layers, and an output layer. Each layer is connected by weight parameters. During model training, the weights are optimized using the cross-entropy loss function and iteratively adjusted based on the gradient descent algorithm until the model's prediction accuracy on the validation set reaches a preset value, ensuring that the model has reliable performance prediction capabilities.

[0121] In one possible implementation, the combined features are first standardized and preprocessed to map the feature values ​​of each dimension to the [0,1] interval, eliminating the interference of dimensional differences on model prediction. The preprocessed combined features are then input into the input layer of the MLP model, and feature association information is extracted layer by layer through fully connected hidden layers. The hidden layers use the ReLU (Rectified Linear Unit) activation function to enhance the model's nonlinear fitting ability. After the hidden layer operation, the feature mapping result is converted into a 0-1 probability value through the sigmoid activation function of the output layer. If the probability value is greater than or equal to the preset prediction threshold, it is determined that the probability of fulfilling the duties is relatively high, and the probability value is finally output as the prediction score.

[0122] The ReLU activation function is the modified linear unit activation function, and its mathematical expression is: That is, the part of the input feature value less than 0 is set to 0, and the part greater than 0 remains unchanged.

[0123] In this application, the ReLU activation function is used in the hidden layer of the MLP model, which can effectively solve the gradient vanishing problem in deep network training, reduce computational complexity, quickly extract nonlinear correlation information between employee features and job features, adapt to the complex mapping requirements of high-dimensional combined features, and improve the accuracy of prediction scores.

[0124] Based on the above embodiments, this application forms combined features by splicing employee and job characteristics in a specific order, eliminates dimensional interference through standardized preprocessing, extracts nonlinear correlation information by combining a model trained on the company's internal historical data with the ReLU activation function, and outputs a quantitative prediction score through the sigmoid function, effectively improving the accuracy of job performance prediction, providing a reliable basis for job suitability assessment, and thus optimizing the job recommendation effect.

[0125] S405: Determine feature similarity based on the characteristics of the employee to be recommended and the characteristics of similar employees.

[0126] In one possible implementation, determining feature similarity includes the following steps: Step 1: Obtain the feature vector of the employee to be recommended and the feature vector of similar employees. Both vectors are standardized high-dimensional numerical vectors with consistent dimensions and covering the core feature dimensions corresponding to the employee's basic information, skill data, and performance data.

[0127] Step II: Calculate the cosine similarity between the two vectors to obtain the original calculated value. Standardize the original calculated value and map it to a 0-1 score scale. The higher the score, the stronger the fit between the employee to be recommended and similar employees in the core feature dimension, providing a unified similarity reference for subsequent matching score calculation.

[0128] It should be understood that this step uses the cosine similarity algorithm to accurately quantify the degree of feature fit between employees. The calculation based on the standardized feature vector can eliminate the interference of dimensional differences, ensure the objectivity and comparability of feature similarity results, and provide a reliable basic indicator for multi-dimensional adaptation assessment.

[0129] S406: Generate a matching score for the candidate positions based on the feature similarity, filtering score, and prediction score.

[0130] S407: Based on the matching scores of each candidate position, determine the target position from at least one candidate position.

[0131] S408: Send a job recommendation response message to the client.

[0132] For the above S401, S402, S406, S407 and S408, please refer to S301, S302, S304, S305 and S306, which will not be elaborated on here.

[0133] Based on the above embodiments, by first extracting standardized feature vectors of employees and positions, and then calculating feature similarity reflecting the fit of core features, filtering scores based on effective practices of similar employees with high fit, and prediction scores based on models trained on historical enterprise data, a multi-dimensional adaptation evaluation system is constructed in collaboration among the three to avoid bias in single-dimensional judgments, and thus output recommendation results that are more in line with the needs of employees and positions.

[0134] Please see Figure 5 This application provides a flowchart illustrating another job recommendation method. The specific steps of the job recommendation method provided in this application are as follows: S501: Receives a job recommendation request message sent by the client.

[0135] S502: Retrieve employee data of the employees to be recommended from the database based on employee identifiers.

[0136] S503: Based on employee data and job requirement data, determine the suitability indicators for at least one candidate job.

[0137] S504: Retrieve historical matching data corresponding to candidate positions from the database based on the candidate positions.

[0138] The aforementioned historical matching data includes: job requirements data for similar positions to the candidate positions, matching records, suitability indicators, and employee performance ratings after onboarding.

[0139] The job requirement data for similar positions to the above candidate positions includes: job requirements that are consistent with the business line of the candidate positions, have a similarity in skill requirements greater than the preset skill similarity threshold, and have the same basic threshold. Specifically, this includes the job skill list and level requirements, working hours, execution conditions, urgency, etc.

[0140] The aforementioned preset skill similarity threshold is a critical value for judging the fit of job skill requirements. It is calibrated by the enterprise based on the skill requirement overlap statistics of high-performance matching cases of similar positions in the past. The threshold can be appropriately increased for core business lines or high skill threshold positions to ensure that the skill requirements of similar positions have a strong correlation. For example, the value range is 0.75-0.85 (default value 0.8).

[0141] The matching records mentioned above include: employee identifiers in historical recommendations, identifiers for similar positions, recommendation timestamps, original calculated values ​​of the matching indicators, employee feedback on acceptance / rejection of recommendations, and records of whether they are on duty.

[0142] The original calculated values ​​of the above-mentioned adaptation indicators refer to the initial results obtained in the historical recommendation process according to the corresponding calculation logic without standardization or weight adjustment. Specifically, they include: the original cosine similarity calculation value of feature similarity, the original weighted summation value of filtering score, and the probability value directly output by the model for prediction score. These original values ​​completely retain the calculation process data and provide a true basis for the subsequent statistical analysis of historical recommendation accuracy.

[0143] The performance evaluation of employees after they start work includes: project completion quality evaluation, collaboration and adaptation evaluation, and task achievement rate evaluation. The evaluation data is synchronized from the company's HR system server to the database.

[0144] S505: Historical recommendation accuracy of matching indicators for candidate positions determined based on historical matching data.

[0145] The historical recommendation accuracy rate is used to characterize the reliability of the matching indicator for predicting similar positions. Specifically, it refers to the ratio of the number of cases in historical data where the matching indicator is predicted to meet the standard and the employee's actual performance score reaches the preset performance score standard after taking up the post to the total number of cases where the matching indicator is predicted to meet the standard. The value ranges from 0 to 1. The higher the ratio, the stronger the predictive reliability of the indicator.

[0146] In one possible implementation, the historical recommendation accuracy of the candidate job's suitability index is determined based on historical matching data, including: screening valid cases of similar jobs from historical matching data; for each suitability index, counting the number of cases with index values ​​≥ the corresponding threshold as the numerator; counting the number of cases in the above cases where employee performance scores ≥ a preset performance score standard as the denominator; and calculating the ratio of the denominator to the numerator to obtain the historical recommendation accuracy of the suitability index.

[0147] It should be understood that the historical recommendation accuracy is calculated based on the company's real historical data. It can dynamically reflect the actual predictive effect of each matching indicator in similar positions, avoid the weight allocation bias caused by using a fixed accuracy rate, and make the subsequent weight coefficients more targeted.

[0148] S506: Determine the weighting coefficients of the suitability indicators for candidate positions based on historical recommendation accuracy.

[0149] The weight coefficients of the aforementioned adaptation indicators represent the contribution percentage of each adaptation indicator in the matching score calculation. The sum of the weight coefficients is 1. The higher the historical recommendation accuracy of an adaptation indicator, the larger its corresponding weight coefficient, in order to highlight the decision-making value of highly reliable indicators.

[0150] In one possible implementation, the weight coefficients of the matching indicators for candidate positions are determined based on historical recommendation accuracy, including: obtaining the historical recommendation accuracy of each matching indicator, denoted as P1, P2, and P3 respectively; calculating the sum of the total accuracy S = P1 + P2 + P3; calculating the weight coefficients W1 = P1 / S, W2 = P2 / S, and W3 = P3 / S for each indicator respectively; if the historical recommendation accuracy of an indicator is lower than a preset lower limit, its weight coefficient is set as the minimum weight threshold, and the remaining weights are redistributed according to the proportion of the remaining accuracy, so as to avoid low-reliability indicators from excessively affecting the matching results.

[0151] The aforementioned minimum weight threshold refers to the lower limit of the weight coefficient set in advance to avoid excessive interference from low-reliability matching indicators in the matching score calculation. This threshold is set based on the model training experience of the enterprise's historical matching data. The core logic is to deprive low-accuracy indicators of their dominance while retaining a small amount of their weight to take into account feature diversity and avoid information omission due to complete zeroing. Its specific value can be dynamically adjusted according to the enterprise's recommendation scenario to ensure a balance between filtering invalid interference and retaining potentially effective information.

[0152] S507: The matching score of the candidate positions is obtained by weighting the weight coefficients and the matching indicators.

[0153] In one possible implementation, the weighted calculation uses a linear weighted summation formula:

[0154] W1, W2, and W3 are the weighting coefficients determined by S506. The calculation results are rounded to two decimal places and mapped to a 0-1 score system. All the above adaptation indicators are parameters that have been preprocessed and standardized.

[0155] For example: if the feature similarity is 0.85 (W1=0.35), the filtering score is 0.8 (W2=0.3), and the prediction score is 0.75 (W3=0.35), then the matching score = 0.85×0.35+0.8×0.3+0.75×0.35=0.7975, which is 0.80 rounded to two decimal places.

[0156] It should be understood that by using dynamic weighting coefficients to weight multi-dimensional adaptation indicators, compared to fixed weights, the contribution of each indicator can be adjusted based on the historical prediction results of similar positions, making the matching score more in line with the actual adaptation needs of the position and further improving the accuracy of recommendations.

[0157] S508: Based on the matching scores of each candidate position, determine the target position from at least one candidate position.

[0158] In one possible implementation, candidate positions with matching scores greater than a score threshold are identified as initial screening positions.

[0159] The aforementioned scoring threshold is used to determine whether candidate positions meet the basic suitability requirements. Specifically, the threshold range is set based on the statistical distribution of the historical target position matching scores of the enterprise. The threshold can be raised for core business lines, positions with high urgency or high skill thresholds, and lowered for temporary replacements or low-threshold positions, to ensure that the initially screened positions have basic suitability potential.

[0160] In one possible implementation, based on the preliminary screening of positions, the preliminary screening positions are verified according to the candidate conditions, and the preliminary screening positions that pass the verification are determined as target positions.

[0161] Among them, the candidate criteria are used to verify whether the initially screened positions match the actual work status of the employees to be recommended.

[0162] The aforementioned actual work status refers to the objective conditions affecting the employee's ability to take on the position, such as current workload, available time, and departmental manpower constraints. Specifically, this includes: whether the employee is currently undertaking any unfinished urgent tasks, the daily / weekly working hours available for the target position, whether the department allows the employee to collaborate across departments or participate in temporary projects, and the employee's workload within the current performance cycle.

[0163] S509: Send a job recommendation response message to the client.

[0164] For the above S501, S502, S503, S508 and S509, please refer to S301, S302, S303, S305 and S306, which will not be elaborated on here.

[0165] Based on the above embodiments, by calculating the historical recommendation accuracy of the matching indicators according to historical matching data of similar positions, and dynamically allocating the weight coefficients of each indicator, the matching score is made to fit the actual prediction effect of the position. By combining multi-dimensional matching indicator collaborative evaluation with the actual status verification such as the employee's current workload and available time, it avoids the bias of single-dimensional judgment and eliminates the situation of mismatch between manpower and capacity, which significantly improves the accuracy and feasibility of job recommendations, effectively adapts to the dynamic configuration needs of enterprises for flexible employment, and reduces the problems of job vacancy and manpower mismatch.

[0166] Please see Figure 6 This application provides a flowchart illustrating another job recommendation method. The specific steps of the job recommendation method provided in this application are as follows: S601: Receives a job recommendation request message sent by the client.

[0167] S602: Retrieve employee data of employees to be recommended from the database based on employee identifiers.

[0168] S603: Based on employee data and job requirement data, determine the suitability indicators for at least one candidate job.

[0169] S604: Generate matching scores for candidate positions based on their suitability metrics.

[0170] S605: Based on the matching scores of each candidate position, determine the target position from at least one candidate position.

[0171] S606: Send a job recommendation response message to the client.

[0172] S607: Update the target positions based on the updated employee data and the updated job requirement data.

[0173] The updated employee data includes: skills update data, performance change data, and basic information change data for employees to be recommended.

[0174] The updated job requirement data includes: adjustment data for candidate job requirements, change data for urgency, and change data for status.

[0175] The aforementioned data on adjustments to the requirements for candidate positions includes changes in core requirements dimensions such as additions / removals to the skill requirement list, increases / decreases in skill level requirements, adjustments to basic thresholds, changes in working hours or workload, and modifications to execution conditions. This data is synchronized to the database in real time by the HR system server 1001.

[0176] It should be understood that both employee data and job requirement data are constantly changing. For example, employees may acquire new skills through training, or job requirements may increase due to business adjustments. If the original target positions are maintained based on historical static data, it is easy for the recommended results to become out of touch with actual needs. The HR system server 1001 updates data to the database in real time to ensure that the updated employee data and job requirement data are authentic, complete, and timely. Dynamically adjusting target positions based on such updated data allows the recommended results to continuously match the current capabilities of employees with the actual needs of the positions, avoiding human resource mismatch or job mismatch caused by data lag. This further strengthens the dynamic adaptability of job recommendations and adapts to the dynamic adjustment scenarios of flexible employment in enterprises.

[0177] As one feasible approach, an update matching score is determined based on the updated employee data and the updated job requirement data, and the target job is updated based on the update matching score.

[0178] Among them, the updated matching score refers to the multi-dimensional adaptation weighted score recalculated based on the updated employee data and the updated job requirement data. It is used to accurately represent the real-time adaptability between the recommended employees and the updated candidate jobs. The calculation logic is the same as the original matching score, but the input data is the latest state.

[0179] Specifically, based on the updated employee data and the updated job requirement data, the updated matching score is determined, including: extracting features from the updated employee data and job requirement data to generate updated employee feature vectors and job feature vectors; recalculating feature similarity, filtering score, and prediction score; obtaining the latest historical recommendation accuracy of each matching indicator to determine dynamic weights; and calculating the updated matching score using a linear weighted summation formula, with the result rounded to two decimal places.

[0180] Based on this, the target positions are updated according to the updated matching score, including: identifying candidate positions with updated matching scores greater than a preset score threshold as preliminary updated positions; verifying the preliminary updated positions based on the candidate conditions; if the preliminary updated positions pass the verification and the updated matching score ranking does not change significantly, the original target positions are maintained and the position information is updated synchronously; if the preliminary updated positions fail the verification or the updated matching score ranking changes significantly, the top-ranked positions that pass the verification are re-selected as new target positions; if no candidate positions meet the updated matching score threshold and candidate conditions, a response indicating that there are currently no suitable target positions is returned to the client, and key mismatches in job requirements or employee status are simultaneously indicated.

[0181] Please see Figure 7 , Figure 7 This is a schematic diagram of a job recommendation device provided in an embodiment of this application. The job recommendation device provided in this embodiment includes: an acquisition module 701, a determination module 702, a sending module 703, and an update module 704.

[0182] Module 701: Used to receive job recommendation request messages sent by the client. The job recommendation request message includes at least the employee identifier of the employee to be recommended.

[0183] Module 701: It is also used to retrieve employee data of employees to be recommended from the database based on employee identifiers.

[0184] Module 702: Used to determine the suitability indicators for at least one candidate position based on employee data and job requirement data; wherein, employee data includes basic employee information, skill data and performance data; job requirement data includes the requirement information for at least one candidate position.

[0185] Module 702: It is also used to generate a matching score for the candidate positions based on the suitability indicators of the candidate positions; the matching score is used to characterize the degree of suitability between the recommended employee and the candidate position.

[0186] Module 702: It is also used to determine the target position from at least one candidate position based on the matching score of each candidate position; Sending module 703: Used to send a job recommendation response message to the client. The job recommendation response message includes the job information of the target job.

[0187] Module 702 is further used to determine the suitability indicators for candidate positions. The suitability indicators for candidate positions include: feature similarity, filtering score, and prediction score. Feature similarity is used to characterize the degree of fit between the employee to be recommended and the candidate position. Filtering score is used to characterize the degree of preference of similar employees for the candidate position. Similar employees are other employees within the company whose employee data and the employee data of the employee to be recommended have a similarity that meets a preset standard. Prediction score is used to characterize the degree to which the employee to be recommended meets the performance standards for the candidate position.

[0188] The determination module 702 is also used to extract features from employee data and candidate job requirements information respectively, to obtain the features of the employees to be recommended and the features of the candidate jobs, to determine the filtering score and the prediction score based on the features of the employees to be recommended and the features of similar employees, and to determine the feature similarity based on the features of the employees to be recommended and the features of similar employees.

[0189] Module 702 is further used to determine the filtering score, including: screening target similar employees whose employee similarity reaches a preset standard from similar employees, wherein the preset standard is used to judge whether the feature fit of the recommended employees meets the requirements; extracting job selection data of target similar employees, wherein the job selection data is the data of candidate positions that target similar employees actively apply for or accept recommendations for, and whose performance scores reach the preset scoring standard after taking up the post, wherein the preset scoring standard is used to judge whether the work performance of the employees after taking up the post meets the requirements; and determining the candidate position filtering score based on the job selection data.

[0190] The determination module 702 is also used to determine the prediction score, including: concatenating the features of the employee to be recommended with the features of the candidate position to obtain a combined feature; inputting the combined feature into the prediction model, making a prediction based on the prediction model to obtain a prediction score; the prediction model predicts the degree of performance of the employee to be recommended in the candidate position based on the combined feature, and outputs the corresponding quantitative prediction score.

[0191] Module 702 is further used to retrieve historical matching data corresponding to candidate positions from the database based on the candidate positions. The historical matching data includes: job requirement data, matching records, fit indicators, and employee performance scores after onboarding for similar positions. Based on the historical matching data, the historical recommendation accuracy rate of the fit indicators for candidate positions is determined. The historical recommendation accuracy rate is used to characterize the predictive reliability of the fit indicators for similar positions. Based on the historical recommendation accuracy rate, the weight coefficients of the fit indicators for candidate positions are determined. Based on the weight coefficients and fit indicators, a weighted calculation is performed to obtain the matching score of the candidate positions.

[0192] Module 702: It is also used to determine candidate positions with matching scores greater than the score threshold as preliminary screening positions. The score threshold is used to determine whether the candidate positions meet the basic adaptation requirements. Based on the candidate conditions, the preliminary screening positions are verified, and the preliminary screening positions that pass the verification are determined as target positions. The candidate conditions are used to verify whether the preliminary screening positions are consistent with the actual work status of the employees to be recommended.

[0193] Update module 704: Used to update target positions based on updated employee data and updated job requirement data. Updated employee data includes: skill update data, performance change data, and basic information change data of employees to be recommended. Updated job requirement data includes: requirement adjustment data, urgency change data, and status change data of candidate positions.

[0194] Update module 704: It is also used to determine the update matching score based on the updated employee data and the updated job requirement data; and to update the target job based on the update matching score.

[0195] It should be noted that, Figure 7 The module division shown is illustrative and represents only one logical functional division; in actual implementation, other division methods are possible. For example, two or more functions can be integrated into a single processing module. These integrated modules can be implemented either in hardware or as software functional modules.

[0196] In exemplary embodiments, as described above, the electronic device may specifically be an electronic device with computing processing capabilities, such as a computer or service. In this case, embodiments of this application also provide an electronic device. Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. For example... Figure 8 As shown, the electronic device includes: a processor 10, a memory 20, a communication line 30, a communication interface 40, and an input / output interface 50.

[0197] The processor 10, memory 20, communication interface 40, and input / output interface 50 can be connected via communication line 30.

[0198] Processor 10 is used to execute instructions stored in memory 20 to implement the flight scheduling method provided in the above embodiments of this application. Processor 10 can be a CPU, a network processor (NP), a digital signal processor (DSP), a microprocessor, a microcontroller (MCU) / single-chip microcomputer / microcontroller, a programmable logic device (PLD), or any combination thereof. Processor 10 can also be any other device with processing capabilities, such as a circuit, device, or software module; this application embodiment does not limit this. In one example, processor 10 may include one or more CPUs, for example... Figure 8 CPU0 and CPU1 in the example. As an optional implementation, the electronic device may include multiple processors; for example, in addition to processor 10, it may also include processor 60. Figure 8 (The example shown is a dashed line).

[0199] The memory 20 is used to store instructions. For example, the instructions may be computer programs. Optionally, the memory 20 may be a read-only memory (ROM) or other types of static storage devices that can store static information and / or instructions; it may also be a random access memory (RAM) or other types of dynamic storage devices that can store information and / or instructions; it may also be an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM), or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media, or other magnetic storage devices, etc. The embodiments of this application do not limit this.

[0200] It should be noted that the memory 20 can exist independently of the processor 10 or it can be integrated with the processor 10. The memory 20 can be located inside or outside the electronic device, and this application embodiment does not impose any restrictions on this.

[0201] Communication line 30 is used to transmit information between the components included in the electronic device.

[0202] Communication interface 40 is used to communicate with other devices or other communication networks. These other communication networks can be Ethernet, radio access network (RAN), wireless local area network (WLAN), etc. Communication interface 40 can be a module, circuit, transceiver, or any device capable of enabling communication.

[0203] Input / output interface 50 is used to enable human-computer interaction between users and electronic devices. For example, it enables action interaction or information exchange between users and electronic devices.

[0204] For example, the input / output interface 50 can be a mouse, keyboard, display screen, or touch screen. Action or information interaction between the user and the electronic device can be achieved through a mouse, keyboard, display screen, or touch screen.

[0205] It should be noted that, Figure 8 The structures shown do not constitute a limitation on electronic devices, except... Figure 8 In addition to the components shown, electronic devices may include more or fewer components than illustrated, or combinations of certain components, or different component arrangements.

[0206] In an exemplary embodiment, this application also provides a readable storage medium including software instructions that, when run on an electronic device, cause the electronic device to perform any of the methods provided in the above embodiments.

[0207] In an exemplary embodiment, this application also provides a computer program product containing computer execution instructions, which, when run on an electronic device, causes the electronic device to perform any of the methods provided in the above embodiments.

[0208] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software programs, implementation can be, in whole or in part, in the form of a computer program product. This computer program product includes one or more computer-executable instructions. When these computer-executable instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer-executable instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer-executable instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device containing one or more servers, data centers, etc., that can be integrated with the medium. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

[0209] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or S, and "a" or "an" does not exclude a plurality. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0210] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of this application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from the spirit and scope of this application. Thus, if such modifications and modifications of this application fall within the scope of the claims of this application and their equivalents, this application is also intended to include such modifications and modifications.

[0211] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A job recommendation method, characterized in that, Applied to a server, the method includes: Receive a job recommendation request message sent by the client, wherein the job recommendation request message includes at least the employee identifier of the employee to be recommended; Based on the employee identifier, retrieve the employee data of the employee to be recommended from the database; Based on the employee data and job requirement data, at least one candidate job's suitability indicators are determined; wherein, the employee data includes basic employee information, skill data, and performance data; the job requirement data includes the requirement information for the at least one candidate job. Based on the suitability indicators of the candidate positions, a matching score is generated for the candidate positions; the matching score is used to characterize the degree of suitability between the employee to be recommended and the candidate positions. Based on the matching scores of each of the candidate positions, a target position is determined from the at least one candidate position; A job recommendation response message is sent to the client, the job recommendation response message including the job information of the target job.

2. The job recommendation method according to claim 1, characterized in that, The matching indicators for the candidate positions include: feature similarity, filtering score, and prediction score; wherein, feature similarity is used to characterize the degree of fit between the employee to be recommended and the candidate position; the filtering score is used to characterize the degree of preference of similar employees for the candidate position, and the similar employees are other employees within the enterprise whose employee data and the employee data of the employee to be recommended have a similarity of a preset standard; the prediction score is used to characterize the degree of performance of the employee to be recommended in the candidate position.

3. The job recommendation method according to claim 2, characterized in that, The step of determining the suitability indicators for at least one candidate position based on the employee data and job requirement data includes: Feature extraction is performed on the employee data and the demand information of the candidate positions to obtain the features of the employees to be recommended and the features of the candidate positions. Based on the characteristics of the employees to be recommended and the characteristics of the candidate positions, the filtering score and the prediction score are determined; Based on the characteristics of the employee to be recommended and the characteristics of similar employees, the feature similarity is determined.

4. The job recommendation method according to claim 3, characterized in that, The filtering score is determined in the following way: Select target similar employees from the similar employees whose employee similarity reaches a preset standard, wherein the preset standard is used to determine whether the feature matching degree of the employee to be recommended meets the requirements; Extract the job selection data of the target-similar employees. The job selection data is the data of candidate positions that the target-similar employees actively applied for or accepted recommendations for, and whose performance scores reached the preset scoring standards after taking up the post. The preset scoring standards are used to determine whether the employee's work performance after taking up the post meets the requirements. The candidate job filtering score is determined based on the job selection data.

5. The job recommendation method according to claim 3, characterized in that, The predicted score is determined in the following way: The features of the employees to be recommended are combined with the features of the candidate positions to obtain combined features; The combined features are input into the prediction model, and a prediction is made based on the prediction model to obtain the prediction score. The prediction model predicts the degree of performance of the recommended employee in the candidate position based on the combined features, and outputs the corresponding quantitative prediction score.

6. The job recommendation method according to claim 1, characterized in that, The step of generating a matching score for the candidate positions based on the suitability metrics of the candidate positions includes: Based on the candidate positions, historical matching data corresponding to the candidate positions are obtained from the database. The historical matching data includes: job requirement data, matching records, suitability indicators, and employee performance scores after onboarding for similar positions. Based on the historical matching data, the historical recommendation accuracy of the suitability index for the candidate positions is determined. The historical recommendation accuracy is used to characterize the reliability of the suitability index in predicting similar positions. The weighting coefficients of the suitability indicators for the candidate positions are determined based on the historical recommendation accuracy. The matching score of the candidate position is obtained by weighting the weight coefficients and the matching indicators.

7. The job recommendation method according to claim 1, characterized in that, The step of determining the target position from the at least one candidate position based on the matching score of each candidate position includes: Candidate positions with matching scores greater than a score threshold are identified as preliminary screening positions. The score threshold is used to determine whether the candidate positions meet the basic matching requirements. The preliminary screening positions are verified based on the candidate criteria, and the preliminary screening positions that pass the verification are determined as target positions. The candidate criteria are used to verify whether the preliminary screening positions match the actual work status of the employees to be recommended.

8. The job recommendation method according to claim 1, characterized in that, The method further includes: Based on the updated employee data and the updated job requirement data, the target job is updated. The updated employee data includes: updated skill data, performance change data, and basic information change data of the employees to be recommended. The updated job requirement data includes: adjusted requirement data, urgency change data, and status change data of the candidate job.

9. The job recommendation method according to claim 8, characterized in that, The process of updating the target position based on the updated employee data and the updated job requirement data includes: Based on the updated employee data and the updated job requirement data, an update matching score is determined; The target job position is updated based on the updated matching score.

10. An electronic device, characterized in that, include: Processor and memory; The memory stores instructions that the processor can execute; When the processor is configured to execute the instructions, it causes the electronic device to implement the method as described in any one of claims 1 to 9.

11. A computer-readable storage medium, characterized in that, The storage medium stores a computer program or instructions, which, when executed by a processing device, implement the method of any one of claims 1 to 9.

12. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed on the processing device, the method of any one of claims 1 to 9 is implemented.

Images