Talent whole-process management method and device based on AI, electronic equipment and medium

Abstract

The invention provides an AI-based talent whole-process management method and device, electronic equipment and a medium, and the method comprises the steps: obtaining a recruitment application submitted by a recruiter, building a recruitment post after the recruitment application passes the approval, and generating the post responsibility information of the recruitment post based on a first sub-module of an AI model; obtaining resume information of the applicant, executing a resume information evaluation task based on a second sub-module of the AI model, and calculating a matching degree score between the resume information and the recruitment post; an interview invitation is sent to the applicant based on the matching degree score, an interview evaluation task is executed through a third sub-module of the AI model, the applicant receiving the interview invitation is interviewed, and an evaluation result of the interviewed applicant is obtained; and determining candidates based on the evaluation result, and sending a review notification or an in-job invitation notification to the candidates. According to the invention, the recruitment process is optimized, and the recruitment quality is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method, device, electronic device and medium for full-process management of talents based on AI. Background Art

[0002] With the growth of enterprises' demand for talents, the traditional recruitment methods can no longer meet the requirements of fast, efficient and accurate recruitment. The traditional recruitment process is cumbersome and time-consuming, and it is difficult to quickly respond to market changes. The efficiency of manual resume screening is low, which is easily affected by subjective factors and it is difficult to accurately evaluate the matching degree between candidates and positions. In addition, the data in the recruitment process is scattered and it is difficult to form effective data analysis. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a method, device, electronic device and medium for full-process management of talents based on AI, so as to optimize the recruitment process and improve the recruitment quality.

[0004] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0005] In the first aspect, the present invention provides a method for full-process management of talents based on AI, including: obtaining a recruitment application submitted by a recruiter, and creating a recruitment position after the recruitment application is approved, and generating job responsibility information of the recruitment position based on the first sub-module of the AI model; obtaining the resume information of the applicant, and performing a resume information evaluation task based on the second sub-module of the AI model, and calculating the matching degree score between the resume information and the recruitment position; sending an interview invitation to the applicant based on the matching degree score, and performing an interview evaluation task through the third sub-module of the AI model to interview the applicant who has received the interview invitation, and obtaining the evaluation result of the interviewed applicant; determining a candidate based on the evaluation result, and sending a second-round interview notice or an employment invitation notice to the candidate.

[0006] Optionally, after sending the employment invitation notice to the candidate, it further includes: after the candidate accepts the employment invitation, obtaining the candidate's identity information and adding it to a preset talent pool; judging whether the candidate needs to participate in training based on the job requirements of the recruitment position; if so, sending a training notice to the candidate so that the candidate can participate in the training and take a final exam; obtaining the final exam results of the candidate who participates in the training, and sending an employment notice to the candidate who passes the final exam based on the final exam results.

[0007] Optionally, after sending the employment notice to the candidates who have passed the graduation examination, it further includes: after the new employees start work, conducting performance appraisals on the new employees at a preset cycle, and performing performance appraisal assessment tasks through the fourth sub-module of the AI model to analyze the performance appraisal results of the new employees; generating the career development paths of the new employees based on the analysis results of the performance appraisal results.

[0008] Optionally, the AI model includes multiple sub-models; the method further includes: calculating the adaptation degree of the sub-models to the tasks, and determining the main model and the alternative model based on the adaptation degree; where the tasks at least include: resume information evaluation tasks, interview evaluation tasks, performance appraisal tasks, talent pool maintenance and update tasks; performing tasks based on the main model and the alternative model, and verifying the task execution results of the main model based on the task execution results of the alternative model to obtain the final task execution results.

[0009] Optionally, verifying the task execution results of the main model based on the task execution results of the alternative model to obtain the final task execution results, including: verifying the task execution results of the main model according to a preset verification function based on the task execution results of the alternative model; if the result of the verification function is the task execution results of the main model, then determining the task execution results of the main model as the final task execution results, otherwise, re-determining the main model; where the verification function is:

[0010] Optionally, it further includes: calculating the health scores of the main model and the alternative model; where the health score is obtained by weighted calculation of multiple model evaluation indicators; if the health score of the main model is lower than a preset first critical value, then updating the main model to the alternative model; if the health score of the alternative model is higher than a preset second critical value within a preset time period, then updating the alternative model to the main model; where the second critical value is greater than the first critical value.

[0011] Optionally, it further includes: splitting the tasks into multiple subtasks, and determining the target sub-models for executing the subtasks based on the task requirements of the subtasks and the model performance of the sub-models.

[0012] In a second aspect, the present invention provides an AI-based full-process talent management device, including: a job creation module, configured to obtain a recruitment application submitted by a recruiter, and create a recruitment position after the recruitment application is approved, and generate job responsibility information for the recruitment position based on a first sub-module of an AI model; a resume evaluation module, configured to obtain resume information of an applicant, and perform a resume information evaluation task based on a second sub-module of the AI model, and calculate a matching degree score between the resume information and the recruitment position; an interview evaluation module, configured to send an interview invitation to the applicant based on the matching degree score, and perform an interview evaluation task through a third sub-module of the AI model to interview the applicant who has received the interview invitation, and obtain an evaluation result of the interviewed applicant; a notification module, configured to determine a candidate based on the evaluation result, and send a second-round interview notice or an employment invitation notice to the candidate.

[0013] In a third aspect, the present invention provides an electronic device, including a processor and a memory, where the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the steps of any one of the methods provided in the first aspect above.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, it executes the steps of any one of the methods provided in the first aspect above.

[0015] The present invention brings the following beneficial effects:

[0016] The above-mentioned AI-based full-process talent management method, device, electronic device and medium provided by the present invention first obtain the recruitment application submitted by the recruiter, and after the recruitment application is approved, create a recruitment position, and generate the job responsibilities information of the recruitment position based on the first sub-module of the AI model; then obtain the resume information of the applicant, and perform a resume information evaluation task based on the second sub-module of the AI model to calculate the matching degree score between the resume information and the recruitment position; then send an interview invitation to the applicant based on the matching degree score, and perform an interview evaluation task through the third sub-module of the AI model to interview the applicant who has received the interview invitation to obtain the evaluation result of the interviewed applicant; finally, determine the candidate based on the evaluation result, and send a second-round interview notice or an employment invitation notice to the candidate. In the recruitment process, the above method can generate the job responsibilities information of the recruitment position through the first sub-module of the AI model to optimize the job description; perform resume evaluation through the second sub-module of the AI model to automatically screen resumes and provide the matching degree score between the resume and the position, thereby improving the resume screening efficiency, reducing the influence of subjective factors, and being able to accurately evaluate the matching degree between the candidate and the position according to the matching degree score; in addition, the above method can also interview the applicant who has passed the resume screening through the third sub-module of the AI model and perform an interview evaluation to obtain the evaluation result, and decide whether to conduct a second-round interview or send an employment invitation notice according to the evaluation result, thereby optimizing the recruitment process and improving the recruitment quality.

[0017] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention are achieved and obtained by the structures specifically pointed out in the specification, claims and drawings.

[0018] In order to make the above objectives, features and advantages of the present invention more obvious and understandable, the following specifically gives preferred embodiments and, in conjunction with the accompanying drawings, the detailed description is as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0020] Figure 1 It is a flowchart of an AI-based full-process talent management method provided by an embodiment of the present invention;

[0021] Figure 2 It is a flowchart of another AI-based full-process talent management method provided by an embodiment of the present invention;

[0022] Figure 3 This is a schematic structural diagram of a talent full - process management device based on AI provided by an embodiment of the present invention;

[0023] Figure 4 This is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. Detailed implementation manners

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0025] Currently, the traditional recruitment process is cumbersome and time - consuming, and it is difficult to quickly respond to market changes. Manual resume screening is inefficient, easily affected by subjective factors, and it is difficult to accurately evaluate the matching degree between candidates and positions. In addition, the data in the recruitment process is scattered, making it difficult to form effective data analysis.

[0026] Based on this, a talent full - process management method, device, electronic device, and medium based on AI provided by the embodiments of the present invention can optimize the recruitment process and improve the recruitment quality.

[0027] For the convenience of understanding this embodiment, first, a talent full - process management method based on AI disclosed in the embodiments of the present invention will be introduced in detail. This method can be executed by an electronic device, such as a smart phone, a computer, a tablet computer, etc. Refer to Figure 1 The flowchart of a talent full - process management method shown below shows that this method mainly includes the following steps S101 to step S104:

[0028] Step S101: Obtain a recruitment application submitted by a recruiter, and after the recruitment application is approved, create a recruitment position, and generate job responsibility information for the recruitment position based on the first sub - module of the AI model.

[0029] In one implementation manner, the employing department submits a recruitment application. After the recruitment application is approved, the recruitment requirements are released; then, according to the recruitment requirements, a recruitment position is automatically created or manually released to each recruitment channel, and job responsibility information for the recruitment position, that is, a job description, including responsibilities, qualification conditions, skill requirements, etc., is generated through the first sub - module of the AI model to optimize the job description.

[0030] Step S102: Obtain the resume information of the applicant, and perform a resume information evaluation task based on the second sub-module of the AI model to calculate the matching degree score between the resume information and the recruitment position.

[0031] In one implementation, the applicant can submit a resume through various channels. After obtaining the applicant's resume information (including educational background, work experience, skill list, etc.), the resume can be evaluated through the second sub-module of the AI model (i.e., perform a resume information evaluation task), automatically screen the resume, and provide a matching degree score.

[0032] Specifically, calculating the matching degree score between the resume information and the recruitment position through the second sub-module of the AI model includes the following steps:

[0033] (1) Obtain the job responsibilities information of the recruitment position and the resume information of the applicant.

[0034] Among them, obtain the detailed requirements of the position (i.e., the job responsibilities information of the recruitment position) from the recruitment website or within the company, including responsibilities, qualification conditions, skill requirements, etc.; collect the applicant's resume, including educational background, work experience, skill list, etc.

[0035] (2) Preprocess the obtained job responsibilities information of the recruitment position and the resume information of the applicant.

[0036] Among them, the preprocessing at least includes: Text cleaning: Remove irrelevant characters, such as punctuation marks, special characters, etc., and convert the text to lowercase.

[0037] Word segmentation: Split the text into words or phrases.

[0038] Stop word removal: Remove common but unhelpful words for analysis, such as "de", "shi", etc.

[0039] Stemming / Lemmatization: Convert words to their basic forms for better comparison.

[0040] (3) Extract features from the preprocessed data.

[0041] Among them, feature extraction at least includes: TF-IDF (Term Frequency-Inverse Document Frequency): Measure the importance of a word in a document.

[0042] Word Embeddings: Use pre-trained models (such as Word2Vec, GloVe) to convert the text into vector representations.

[0043] Skill Matching: Specifically identify and match skill keywords.

[0044] (4) Calculate the matching degree score based on the extracted feature vectors.

[0045] Among them, the similarity between the job responsibilities information of the recruitment position and the resume information can be calculated based on the feature vectors as the matching degree score. Common methods include cosine similarity, Jaccard similarity, etc.

[0046] The supervised learning method can also be adopted, such as logistic regression, support vector machine (SVM), random forest, etc., to train a model to predict the matching degree. Specifically, a historical data set can be used to train the model, and the parameters can be adjusted through means such as cross-validation to optimize the performance. The trained model is applied to the actual resume screening process to give the matching degree score in real time, and a visualization tool helps the HR to understand the results more intuitively.

[0047] Step S103: Send an interview invitation to the applicant based on the matching degree score, and execute the interview assessment task through the third sub-module of the AI model to interview the applicant who has received the interview invitation and obtain the assessment result of the interviewed applicant.

[0048] In one implementation, an interview invitation can be sent to the applicant according to the matching degree score, such as: sending an interview invitation to the applicant whose matching degree score exceeds the matching degree score threshold (which can be 90 points). After receiving the interview invitation, the applicant can be initially screened through the third sub-module of the AI model for an interview.

[0049] In specific implementation, when the recruiter sends an interview invitation, an interview invitation can be automatically sent to the candidate by email or recruitment platform, and an interview link is provided. The applicant signs in for the interview by clicking the interview link and conducts the interview. After the interview, the third sub-module of the AI model is used for interview assessment. Specifically, historical data (such as previous interview records, successful employment cases, etc.) can be used to train the AI model so that it can understand and evaluate the candidate's answers. The AI model can formulate a series of structured questions according to the job requirements of the recruitment position to comprehensively examine the candidate's abilities.

[0050] During the interview process, an AI interviewer is used to interact with the applicant through video conferencing technology. The AI model can capture the applicant's verbal information through speech recognition and understand the content through natural language processing (NLP); at the same time, analyze non-verbal information such as body language and expressions through facial expression recognition algorithms. The AI model can immediately ask follow-up questions or follow-up inquiries during the interview process to understand the applicant's background and abilities more deeply.

[0051] After the interview, the AI model can collect all the data of the candidate during the interview, including the text transcription of the answers, intonation changes, facial expressions, etc. Then, based on the preset criteria and algorithms, the above data is converted into quantitative scores. These criteria may involve multiple dimensions such as professional knowledge, communication skills, and teamwork, and by combining the scores of different dimensions, an overall evaluation result of the candidate is obtained. The AI model can also generate a detailed interview report for each candidate, including various scores, strengths, areas for improvement, and final recommendations. HR or hiring managers can make recruitment decisions with the assistance of the evaluation report provided by the AI model.

[0052] Step S104: Determine the candidates based on the evaluation results, and send a second-round interview notice or an employment invitation notice to the candidates.

[0053] In one implementation, it can be decided whether a second-round interview is needed according to the evaluation results given by the AI model. If a second-round interview is needed, a second-round interview notice can be sent to the candidate. If a second-round interview is not needed, an employment invitation notice can be sent to the candidate.

[0054] The above AI-based full-process talent management method provided by the embodiments of the present invention can, in the recruitment process, generate the job responsibilities information of the recruitment position through the first sub-module of the AI model to optimize the job description; conduct resume evaluation through the second sub-module of the AI model, automatically screen resumes, and provide a matching degree score between the resume and the position, thereby improving the resume screening efficiency, reducing the influence of subjective factors, and being able to accurately evaluate the matching degree between the candidate and the position according to the matching degree score; in addition, the above method can also interview the candidates who pass the resume screening through the third sub-module of the AI model and conduct interview evaluation to obtain the evaluation results, and decide whether to conduct a second-round interview or send an employment invitation notice according to the evaluation results, thereby optimizing the recruitment process and improving the recruitment quality.

[0055] In one implementation, after sending the employment invitation notice to the candidate, the above method further includes: after the candidate accepts the employment invitation, obtaining the candidate's identity information and adding it to the preset talent pool; judging whether the candidate needs to participate in training based on the job requirements of the recruitment position; if so, sending a training notice to the candidate so that the candidate can participate in the training and take a final exam; obtaining the final exam results of the candidates who participate in the training, and based on the final exam results, sending an employment notice to the candidates who pass the final exam.

[0056] In specific implementation, after a candidate accepts an employment invitation, they go through the onboarding procedures and enter their identity information into the talent pool. After the candidate starts work, based on the job requirements, it is determined whether training is needed. If so, a training participation notice is sent to the candidate. The candidate participates in the training and takes a final exam, and based on the candidate's final exam results, it is determined whether the candidate has passed the final exam. After the candidate passes the final exam, an employment notice is sent to them. If the candidate fails the final exam, they may need to retrain or consider other candidates.

[0057] In one implementation, after sending an employment notice to a candidate who has passed the final exam, the above further includes: after a new employee starts work, conducting a performance appraisal of the new employee at a preset cycle, and performing a performance appraisal evaluation task through the fourth sub-module of the AI model to analyze the performance appraisal results of the new employee; generating a career development path for the new employee based on the analysis results of the performance appraisal results.

[0058] In specific implementation, after a new employee starts work, a regular performance appraisal is conducted on them, such as once every three months. The fourth sub-module of the AI model is used for performance appraisal to evaluate the employee's performance, and based on the performance appraisal results, a career development path is provided for the employee. Specifically, before conducting a performance appraisal on a new employee, first determine the performance appraisal objectives and the key performance indicators (KPIs) to be evaluated, including: work quality, work efficiency, teamwork, etc.; then collect historical performance appraisal records and other relevant data, such as attendance rate, project completion status, etc., to train the AI model.

[0059] When conducting a performance appraisal on a new employee, relevant performance-related data can be automatically collected through an enterprise resource planning system (ERP), a customer relationship management system (CRM), or other business systems, and employee self-evaluations and feedback from colleagues and superiors are collected as supplementary data sources. The collected data is input into the AI model to obtain the performance appraisal results of the new employee, and a detailed performance evaluation report is generated for each employee, including scores, areas of strength, areas for improvement, career development paths, etc.

[0060] In the embodiments of the present invention, the AI model can also be used to update and maintain the talent pool, greatly improving efficiency and ensuring the accuracy and timeliness of information. The specific process steps include:

[0061] (1) Data collection and integration

[0062] Use AI technology to automatically collect data on candidates and employees from multiple sources, including social media, professional networks (such as LinkedIn), enterprise internal databases, etc., and clean and standardize the collected information through natural language processing and other AI technologies to remove duplicates and errors.

[0063] (2) Intelligent matching and recommendation

[0064] Adopt an intelligent matching algorithm to analyze job requirements and personal profiles, and find the most suitable candidates for specific positions. Based on the latest activities, skill development of candidates, and market trends, recommend potentially interesting positions or training opportunities in real time.

[0065] (3) Career path prediction and development planning

[0066] Use machine learning algorithms to analyze the career development history of employees, predict future career paths, and provide suggestions to promote their career growth. Based on personal career interests and development goals, combined with the strategic direction of the enterprise, generate personalized learning and development plans.

[0067] (4) Regular updates and interactions

[0068] Set up automated tools to send regular reminders to candidates and employees, encouraging them to update their personal information. Obtain the latest information of candidates and employees through questionnaires, interview feedback, etc., and continuously optimize the content of the talent pool.

[0069] (5) Data protection and compliance

[0070] Ensure that all data processing complies with data protection regulations such as GDPR, and protect the personal privacy of candidates and employees. Obtain clear consent before collecting and using personal information, and provide a transparent policy on how to use this information.

[0071] (6) Performance monitoring and continuous improvement

[0072] Monitor the performance of the AI system to ensure its effectiveness and accuracy. Adjust and optimize the AI model according to the feedback in actual applications and new market demands, and ensure the continuous update and maintenance quality of the talent pool.

[0073] Through the above processes, enterprises can manage and utilize their talent resources more efficiently, and at the same time provide better experiences and services for candidates and employees, which not only helps to discover potential talents, but also promotes the growth and development of existing employees.

[0074] In one implementation, AI models all have certain biases and discriminations and cannot guarantee complete fairness. Relying entirely on a single model may make the talent system face a certain risk of simplification. Based on this, the AI models adopted in the embodiments of the present invention include multiple sub-models, and the main model and alternative models can be determined according to the actual task requirements. Specifically, the above method further includes:

[0075] First, calculate the degree of adaptation of the sub-model to the task, and determine the main model and alternative models based on the degree of adaptation.

[0076] In specific implementation, assume there are n sub-models: M1, M2, ..., M n , and the adaptation degree of each sub-model to task T can be represented by S i . Task T may include technologies for talent management using AI technology, such as: resume information evaluation task, interview evaluation task, performance appraisal task, talent pool maintenance and update task.

[0077] Next, select the sub-model with the highest adaptation degree as the main model:

[0078]

[0079] Subsequently, select K sub-models whose adaptation degree exceeds the adaptation degree threshold as alternative models:

[0080]

[0081] where θ represents the set adaptation degree threshold.

[0082] Then, execute the task based on the main model and the alternative models, and verify the task execution result of the main model based on the task execution results of the alternative models to obtain the final task execution result.

[0083] In specific implementation, based on the task execution results of the alternative models, verify the task execution result of the main model according to a preset verification function; if the result of the verification function is the task execution result of the main model, then determine the task execution result of the main model as the final task execution result, otherwise, re-determine the main model.

[0084] Specifically, use the main model and the alternative models to execute the task. Assume the task execution result of the main model is y 主 , and the task execution results of the alternative models are y j (j = 1, 2, ..., K). Define the verification function as:

[0085] If V(y 主 , {y i}) = y 主 , then the final task execution result is y 主 , otherwise, further analysis may be required to re-determine the main model.

[0086] In the embodiments of the present invention, among numerous models, the main model and the alternative models can be determined based on the adaptation degree of the models to the task, the analysis task can be executed based on the main model and the alternative models, and the result of the main model can be verified based on the results of the alternative models, and the final task execution result can be determined, so as to reduce the impact of the bias and discrimination of a single model on the entire system.

[0087] In some embodiments, in today's data-driven decision-making environment, the applications of machine learning (ML) and artificial intelligence (AI) are becoming increasingly widespread. However, with the changes in the environment and the development of technology, static AI models may become inapplicable or inefficient. Therefore, in the embodiments of the present invention, a dynamic adjustment mechanism can be created that can continuously adjust the selection criteria of the main model and alternative models according to the performance and feedback of the models.

[0088] Specifically, the performance of the model can be measured based on model evaluation metrics. These metrics should include but are not limited to traditional classification metrics such as accuracy, recall, F1 score, etc., as well as application domain-specific metrics such as processing speed, resource consumption, recruiter satisfaction, etc. For different application scenarios, additional factors may also need to be considered, such as fairness, transparency, and interpretability. For example, in a recommendation system, metrics such as CTR (click-through rate), CVR (conversion rate), and AOV (average order value) may be concerned.

[0089] Track the performance of all active models and collect data on their performance. This can be specifically accomplished by integrating logging, API call statistics, and other relevant data sources. The monitoring system can also detect abnormal situations, such as a sudden drop in performance or a deviation in the prediction results. For example, the Z-score is used to identify outliers, and the calculation formula is:

[0090] where X represents the observed value, μ is the mean, and σ is the standard deviation. When |Z| > 3, the observed value can be considered abnormal.

[0091] In a feasible implementation manner, when adjusting the main model and alternative models, the following methods can be adopted including but not limited to: calculating the health scores of the main model and alternative models; where the health score is obtained by weighted calculation of multiple model evaluation metrics; if the health score of the main model is lower than a preset first critical value, then update the main model to an alternative model; if the health score of the alternative model is higher than a preset second critical value within a preset time period, then update the alternative model to the main model; where the second critical value is greater than the first critical value.

[0092] In specific implementation, determine the health score of each sub-model, and this health score can synthesize the results of multiple model evaluation metrics. When the health score of a certain main model is lower than the preset first critical value, it is demoted to an alternative model. Conversely, if a certain alternative model continuously performs excellently (i.e., the health score is higher than the preset second critical value) during its probation period (i.e., within the preset time period), it can be promoted to the main model. In the embodiments of the present invention, a certain number of high-quality alternative models need to be maintained so as to be ready to replace any underperforming main model at any time.

[0093] Specifically, the health score S can be calculated by weighted summation:

[0094] S = ω1·A + ω2·B +... + ω n ·N

[0095] Among them, ω n represents the weight of the nth model evaluation index, and A, B,..., N respectively represent the scores of each model evaluation index. The weights can be adjusted according to the importance of the actual application scenario.

[0096] To improve the flexibility, adaptability, and decision-making quality of the system, a whole can be decomposed into multiple parts to achieve more refined control and higher efficiency. Specifically, the task T can be split into multiple subtasks, and each subtask is allowed to be processed by the most suitable model. Such a method not only increases the adaptability of the system, enabling it to better meet the needs of different recruiters, but also promotes teamwork and knowledge sharing, thereby improving the objectivity and fairness of the evaluation process.

[0097] Based on this, the above method further includes: splitting the task into multiple subtasks, and determining the target submodel for executing the subtask based on the task requirements of the subtask and the model performance of the submodel.

[0098] In specific implementation, when splitting the task, it is necessary to clarify its goals, inputs and outputs, and the key processes involved. For example, in a recruitment platform, the task T may be to recommend personalized job positions to customers. This process involves multiple links such as data collection, user behavior analysis, job classification, and recommendation algorithms. Each link has its unique challenges and technical requirements, so it is very suitable for splitting and processing.

[0099] Next, it is necessary to determine how to reasonably divide the task T. This step needs to be considered in combination with business logic and technical feasibility. For the above example, it can be split according to functional modules or data streams. For example, the task T can be divided into the following subtasks:

[0100] Data preprocessing: Clean, transform, and standardize the original data to ensure the data quality for subsequent steps.

[0101] User profile construction: Create a detailed user profile based on historical interaction records and other relevant information.

[0102] Job feature extraction: Extract features that are helpful for recommendation from multi-source information such as job descriptions and pictures.

[0103] Similarity calculation: Measure the matching degree between the user and the job, which serves as the basis for recommendation.

[0104] Recommendation list generation: Considering multiple factors comprehensively, personalized recommendation results are finally formed.

[0105] Once the sub-task boundaries are determined, the most suitable model can be selected for each sub-task. Different sub-tasks may require different types of technical support, such as deep learning, random forest, collaborative filtering, etc. Factors such as model complexity, training time, and prediction accuracy should be considered during selection. In addition, it is also necessary to evaluate whether the model is easy to interpret, maintain, and support expansion. For example, for the sub-task of "user profile construction", a rule-based method or a shallow neural network can be adopted; while for "similarity calculation", a complex embedded representation method is more suitable.

[0106] For the convenience of understanding, the embodiments of the present invention also provide an AI-based full-process talent management solution, see Figure 2 as shown, including the following steps:

[0107] Step 1), the hiring department submits a recruitment application. After approval, the recruitment requirements are released.

[0108] Step 2), automatically create a position or manually publish the position to each recruitment channel. Use AI to help write JD (job responsibilities) to optimize the job description.

[0109] Step 3), candidates submit resumes through various channels. AI resume evaluation, automatically screen resumes, and provide a matching score.

[0110] Step 4), AI interview for preliminary screening. The recruiter conducts an interview invitation. The candidate signs in for the interview and has the interview.

[0111] Step 5), after the interview, conduct an interview evaluation. According to the evaluation results, decide whether to conduct a second interview or send a notice for a re-examination.

[0112] After determining the candidate, send an employment notice. The candidate accepts the employment invitation.

[0113] Step 6), handle the employment procedures and enter the talent pool. The new employee starts work.

[0114] Step 7), according to the job requirements, determine whether training is needed. If so, send a training notice, participate in the training and take a final exam.

[0115] After passing the exam, send an employment notice. If not passed, re-training may be required or other candidates may be considered.

[0116] Step 8), after the new employee starts work, conduct regular performance evaluations. Use AI technology for data analysis to evaluate the employee's performance.

[0117] Step 9), according to the performance appraisal results, provide career development paths for employees.

[0118] For the above method provided by the embodiments of the present invention, in all aspects of recruitment, training, and appraisal, the AI consulting assistant provides 7x24-hour automatic response services to answer the questions of employees and candidates. It can also continuously use the AI large model to update and maintain the talent pool. Therefore, in the entire recruitment process, through data analysis using AI technology, the recruitment strategy is optimized, and the recruitment quality is improved.

[0119] For the AI-based full-process talent management method provided by the foregoing embodiments, the embodiments of the present invention also provide an AI-based full-process talent management device. Refer to Figure 3 the structural schematic diagram of an AI-based full-process talent management device shown in

[0120] The job creation module 301 is used to obtain the recruitment application submitted by the recruiter, and after the recruitment application is approved, create a recruitment position, and generate the job responsibilities information of the recruitment position based on the first sub-module of the AI model;

[0121] The resume evaluation module 302 is used to obtain the resume information of the applicant, and perform the resume information evaluation task based on the second sub-module of the AI model, and calculate the matching degree score between the resume information and the recruitment position;

[0122] The interview evaluation module 303 is used to send an interview invitation to the applicant based on the matching degree score, and perform the interview evaluation task through the third sub-module of the AI model to interview the applicant who has received the interview invitation, and obtain the evaluation result of the interviewed applicant;

[0123] The notification module 304 is used to determine the candidate based on the evaluation result, and send a second-round interview notice or an employment invitation notice to the candidate.

[0124] For the above AI-based full-process talent management device provided by the embodiments of the present invention, in the recruitment process, it can generate the job responsibilities information of the recruitment position through the first sub-module of the AI model to optimize the job description; perform resume evaluation through the second sub-module of the AI model, automatically screen resumes, and provide the matching degree score between the resume and the position, thereby improving the resume screening efficiency, reducing the influence of subjective factors, and being able to accurately evaluate the matching degree between the candidate and the position according to the matching degree score; in addition, the above device can also interview the applicant who has passed the resume screening through the third sub-module of the AI model and perform interview evaluation to obtain the evaluation result, and decide whether to conduct a second-round interview or send an employment invitation notice according to the evaluation result, thereby optimizing the recruitment process and improving the recruitment quality.

[0125] The device provided in the embodiments of the present invention has the same implementation principle and technical effects as those in the foregoing method embodiments. For the sake of brief description, for the parts not mentioned in the device embodiments, reference may be made to the corresponding content in the foregoing method embodiments.

[0126] It should be noted that the specific values provided in the embodiments of the present invention are only exemplary and are not limited herein.

[0127] The embodiments of the present invention also provide an electronic device. Specifically, the electronic device includes a processor and a storage device; a computer program is stored on the storage device, and when the computer program is run by the processor, it executes the method described in any one of the above embodiments.

[0128] Figure 4 FIG. is a schematic structural diagram of an electronic device provided in an embodiment of the present invention. The electronic device 100 includes: a processor 40, a memory 41, a bus 42, and a communication interface 43. The processor 40, the communication interface 43, and the memory 41 are connected through the bus 42; the processor 40 is configured to execute an executable module stored in the memory 41, such as a computer program.

[0129] Among them, the memory 41 may include a high-speed random access memory (RAM, Random Access Memory), and may also include a non-volatile memory, such as at least one disk memory. Through at least one communication interface 43 (which may be wired or wireless), a communication connection is established between the system network element and at least one other network element, and the Internet, wide area network, local area network, metropolitan area network, etc. can be used.

[0130] The bus 42 may be an ISA bus, a PCI bus, an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 4 only a bidirectional arrow is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0131] Among them, the memory 41 is used to store a program. After receiving an execution instruction, the processor 40 executes the program. The method executed by the device defined by the flow process disclosed in any one of the foregoing embodiments of the present invention can be applied to the processor 40 or implemented by the processor 40.

[0132] The processor 40 may be an integrated circuit chip with the ability to process signals. In the implementation process, the steps of the above method can be completed by the integrated logic circuit of the hardware in the processor 40 or the instructions in the form of software. The above-mentioned processor 40 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP for short), an application specific integrated circuit (ASIC for short), a field-programmable gate array (FPGA for short), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present invention can be directly embodied as being executed and completed by the hardware decoding processor, or executed and completed by a combination of the hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory 41, and the processor 40 reads the information in the memory 41 and combines its hardware to complete the steps of the above method.

[0133] The computer program product of the readable storage medium provided by the embodiments of the present invention includes a computer-readable storage medium storing program code, and the instructions included in the program code can be used to execute the method described in the foregoing method embodiments. For the specific implementation, reference can be made to the foregoing method embodiments, which will not be elaborated here.

[0134] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0135] Finally, it should be noted that the above-mentioned embodiments are only specific embodiments of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed by the present invention can still modify the technical solutions described in the foregoing embodiments or can easily conceive of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. An AI-based full-process talent management method, characterized in that, Including: Obtain the recruitment application submitted by the recruiter, and after the recruitment application is approved, create a recruitment position, and generate the job responsibilities information of the recruitment position based on the first sub-module of the AI model; Obtain the resume information of the applicant, and perform a resume information evaluation task based on the second sub-module of the AI model, and calculate the matching score between the resume information and the recruitment position; Send an interview invitation to the applicant based on the matching score, and perform an interview evaluation task through the third sub-module of the AI model to interview the applicant who has received the interview invitation, and obtain the evaluation result of the interviewed applicant; Determine the candidate based on the evaluation result, and send a second-round interview notice or an employment invitation notice to the candidate.

2. The method according to claim 1, wherein After sending the employment invitation notice to the candidate, it further includes: After the candidate accepts the employment invitation, obtain the identity information of the candidate and add it to the preset talent pool; Based on the job requirements of the recruitment position, determine whether the candidate needs to participate in training; If so, send a training notice to the candidate so that the candidate can participate in the training and take a final exam; Obtain the final exam results of the candidates participating in the training, and based on the final exam results, send an employment notice to the candidates who have passed the final exam.

3. The method according to claim 1, wherein After sending the employment notice to the candidates who have passed the final exam, it further includes: After the new employee joins the company, conduct a performance appraisal on the new employee according to a preset cycle, and perform a performance appraisal evaluation task through the fourth sub-module of the AI model to analyze the performance appraisal results of the new employee; Generate the career development path of the new employee based on the analysis result of the performance appraisal result.

4. The method according to any one of claims 1 to 3, characterized in that The AI model includes multiple sub-models; the method further includes: Calculate the adaptation degree of the sub-model to the task, and determine the main model and the alternative model based on the adaptation degree; wherein, the task at least includes: resume information evaluation task, interview evaluation task, performance appraisal task, talent pool maintenance and update task, job recommendation task; Execute the task based on the main model and the alternative model, and verify the task execution result of the main model based on the task execution result of the alternative model to obtain the final task execution result.

5. The method according to claim 4, wherein Verifying the task execution result of the main model based on the task execution result of the alternative model to obtain the final task execution result, including: Based on the task execution result of the alternative model, verify the task execution result of the main model according to a preset verification function; If the result of the verification function is the task execution result of the main model, then determine the task execution result of the main model as the final task execution result; otherwise, re-determine the main model; wherein, the verification function is:

6. The method according to claim 4, wherein It further includes: Calculate the health scores of the main model and the alternative model; wherein, the health score is obtained by weighted calculation of multiple model evaluation indicators; If the health score of the main model is lower than a preset first critical value, update the main model to the alternative model; If the health score of the alternative model is higher than a preset second critical value within a preset time period, update the alternative model to the main model; wherein, the second critical value is greater than the first critical value.

7. The method according to claim 4, wherein It further includes: Split the task into multiple subtasks, and determine the target submodel for executing the subtask based on the task requirements of the subtask and the model performance of the submodel.

8. An AI-based full-process talent management device, characterized in that, It includes: A job creation module, configured to obtain a recruitment application submitted by a recruiter, and create a recruitment position after the recruitment application is approved, and generate job responsibilities information for the recruitment position based on a first submodule of an AI model; A resume evaluation module, configured to obtain resume information of an applicant, and perform a resume information evaluation task based on a second submodule of the AI model, and calculate a matching degree score between the resume information and the recruitment position; An interview evaluation module, configured to send an interview invitation to the applicant based on the matching degree score, and perform an interview evaluation task through a third submodule of the AI model to interview the applicant who has received the interview invitation, and obtain an evaluation result of the interviewed applicant; A notification module, configured to determine candidates based on the evaluation result, and send a second-round interview notice or an employment invitation notice to the candidates.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it executes the steps of the method according to any one of claims 1 to 7 above.

Images