Markov decision-based career full-period personalized cultivation method and device

Abstract

The invention discloses a career full-cycle personalized training method and device based on Markov decision, and relates to the technical field of artificial intelligence. The method comprises the steps of obtaining a personalized training result according to employee portrait data, project portrait data, training portrait data, enterprise complete post portrait data and a multi-dimensional evaluation result; an optimal training path meeting organization targets and employee personalized requirements is planned based on the Markov decision process, the optimal training path not only considers the current ability and post requirements of the employees, but also considers the potential and direction of future development of the employees, and it is ensured that the employees smoothly adapt to the post requirements based on the optimal training path, so that the employees can be trained more accurately. According to the invention, synchronous improvement of individual occupational development and organization targets is realized, more intelligent and personalized training and development schemes are provided for enterprises and employees, and the employee occupational career management efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method and device for personalized career development throughout the entire cycle based on Markov decision-making. Background Art

[0002] In today's highly competitive business environment, employees face ever-changing market and technological trends, requiring them to continuously upgrade their skills and adapt to new business demands. At the same time, to remain competitive, companies need to ensure their employees possess the most appropriate knowledge and skills to navigate an increasingly complex business environment. Therefore, providing personalized, full-career development for employees has become a critical requirement for innovative enterprises. Existing technologies typically manage employee careers manually. However, this approach is often inefficient and fails to effectively tailor development and management to the individual needs of each employee. Furthermore, the lack of systematic planning and intervention across the entire employee career cycle leads to low talent growth and limited career advancement. Summary of the Invention

[0003] The purpose of this application is to provide a full-cycle career personalized training method and device based on Markov decision-making, which can realize intelligent personalized management and training of employees' full-cycle careers, provide enterprises and employees with more intelligent and personalized training and development programs, and improve the efficiency of employee career management.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a career cycle personalized training method based on Markov decision making, comprising:

[0006] Obtaining portrait data, wherein the portrait data includes employee portrait data, project portrait data, training portrait data, and complete enterprise position portrait data, wherein the project portrait data includes a first actual impact degree of different projects on employee capability improvement, the training portrait data includes a second actual impact degree of different training on employee capability improvement, and the complete enterprise position portrait data includes position responsibilities and position requirements data;

[0007] Conducting a multi-dimensional assessment of the employee's current capabilities based on the employee portrait data to obtain an assessment result, wherein the multi-dimensional assessment includes a performance dimension, an ability dimension, a behavior dimension, a potential dimension, and a professional quality dimension;

[0008] Based on the portrait data and the evaluation results, constructing a fully connected career development graph using a graph embedding algorithm, wherein the fully connected career development graph includes employee nodes, project nodes, training nodes, target position nodes, and edges, and the weights of the edges are variables determined based on the first actual impact degree and the second actual impact degree;

[0009] Taking any employee node as the starting point, the path with the highest cumulative reward is searched in the career development fully connected graph based on the Markov decision process to obtain the optimal employee training path.

[0010] Secondly, this application provides a career cycle personalized training system based on Markov decision making, including:

[0011] A data acquisition module for acquiring portrait data, wherein the portrait data includes employee portrait data, project portrait data, training portrait data, and complete enterprise position portrait data. The project portrait data includes the first actual impact of different projects on employee capability improvement, the training portrait data includes the second actual impact of different trainings on employee capability improvement, and the complete enterprise position portrait data includes job responsibilities and job requirements data.

[0012] An employee evaluation module is used to perform a multi-dimensional evaluation of an employee's current capabilities based on the employee portrait data to obtain an evaluation result, wherein the multi-dimensional evaluation includes a performance dimension, an ability dimension, a behavior dimension, a potential dimension, and a professional quality dimension;

[0013] A path planning module is used to construct a fully connected career development graph based on the portrait data and the evaluation results using a graph embedding algorithm, and take any of the employee nodes as a starting point, and search for the path with the highest cumulative reward in the fully connected career development graph based on a Markov decision process to obtain the optimal employee training path, wherein the fully connected career development graph includes employee nodes, project nodes, training nodes, target position nodes and edges, and the weights of the edges are variables determined based on the first actual impact level and the second actual impact level.

[0014] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the personalized career development method based on Markov decision-making throughout the entire career cycle as described in the first aspect above.

[0015] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the career life-cycle personalized training method based on Markov decision-making as described in the first aspect above.

[0016] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0017] The present application provides a career full cycle personalized training method and device based on Markov decision, which plans the optimal training path that meets organizational goals and employee personalized requirements based on employee portrait data, project portrait data, training portrait data, enterprise complete position portrait data and multi-dimensional evaluation results, based on the Markov decision process and career development full connection graph. The portrait data can more comprehensively understand and analyze the multi-dimensional information of employees, projects and training, and the multi-dimensional evaluation results can provide a basis for subsequent training and management decisions. The optimal training path for employees can ensure that employees can meet the requirements of their positions by participating in specific projects and receiving specific training. The training path not only takes into account the current capabilities and job requirements of employees, but also the potential and direction of their future development. A series of projects and training that help employees improve their abilities and skills will be arranged on the training path to ensure that employees can smoothly adapt to job requirements and achieve the simultaneous improvement of personal career development and organizational goals, providing enterprises and employees with more intelligent and personalized training and development programs, and improving the efficiency of employee career management. This method can provide each employee with a tailored training program, thereby effectively improving their work performance and professional quality, ensuring that employees can acquire the necessary skills and knowledge in the process of career development, and meeting the organization's demand for talent. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 This is a schematic diagram of the structure of a career cycle personalized training system based on Markov decision-making in Example 1 of this application;

[0020] Figure 2 A flow chart of a method for personalized career development throughout the entire cycle based on Markov decision-making is provided in Example 2 of the present application. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0022] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0023] Example 1

[0024] In the modern enterprise environment, employees are constantly faced with changing technologies, work methods, and market trends, requiring them to possess the ability to continuously learn and adapt. Traditional training methods, however, are often one-off and static, failing to meet the learning needs of employees at different stages of their careers. Furthermore, traditional training methods often lack personalization, making it difficult to meet the needs of diverse employees. Furthermore, companies face challenges managing and evaluating employee training effectiveness.

[0025] Against this backdrop, the application of artificial intelligence (AI) technology has become a crucial approach to addressing these issues. AI, a field that integrates a variety of advanced technologies, including machine learning, natural language processing, and recommendation systems, aims to create machines or software systems that can mimic human intelligence, encompassing everything from basic data processing and pattern recognition to complex decision-making and problem-solving. In employee training and career development management, AI is often used to analyze employee capabilities and goals, providing personalized feedback and recommendations accordingly. The core of this technology lies in enabling machines to analyze employee capabilities and goals, and then recommend relevant training content based on their current capabilities and training knowledge base.

[0026] Traditional employee training and career development management usually lacks depth and accuracy in employee assessment and analysis and development path generation, and often cannot provide sufficiently personalized support and advice. In addition, traditional systems are usually not accurate and personalized enough in development path generation and training content recommendations, and cannot meet the diverse development goals of employees. The limitations of traditional training models need to be broken, and intelligent management and training of employees' entire career cycle must be achieved to provide companies and employees with more intelligent and personalized training and development plans.

[0027] Based on this, this embodiment makes full use of artificial intelligence technology and builds a comprehensive full-cycle career training system for corporate employees to provide employees with personalized, efficient and scientific training paths, while providing companies with accurate training management and evaluation methods.

[0028] See Figure 1 ,This embodiment provides a career full-cycle personalized training system based on Markov decision, which includes a data acquisition module (i.e., a job portrait unit in a portrait module and a training path recommendation module), an employee evaluation module, an employee evaluation module, a path planning module (i.e., a training path generation unit in a training path recommendation module), an early warning module, and a memory module.

[0029] A data acquisition module is used to acquire portrait data, wherein the portrait data includes employee portrait data, project portrait data, training portrait data and complete enterprise job portrait data. The project portrait data includes the first actual impact degree of different projects on employee capacity improvement, the training portrait data includes the second actual impact degree of different trainings on employee capacity improvement, and the complete enterprise job portrait data includes job responsibilities and job requirements data.

[0030] The employee evaluation module is used to conduct a multi-dimensional evaluation of the employee's current capabilities based on the employee portrait data to obtain an evaluation result, wherein the multi-dimensionality includes performance dimension, ability dimension, behavior dimension, potential dimension and professional quality dimension.

[0031] A path planning module is used to construct a fully connected career development graph based on the portrait data and the evaluation results using a graph embedding algorithm, and take any of the employee nodes as a starting point, and search for the path with the highest cumulative reward in the fully connected career development graph based on a Markov decision process to obtain the optimal employee training path, wherein the fully connected career development graph includes employee nodes, project nodes, training nodes, target position nodes and edges, and the weights of the edges are variables determined based on the first actual impact level and the second actual impact level.

[0032] The portrait module includes an employee portrait unit, a project portrait unit, a training project unit and an enterprise complete position portrait unit.

[0033] The employee portrait unit is used to analyze text data based on employee portrait demand data using a natural language processing model (BERT model), extract keywords and topics, and generate employee portrait data.

[0034] The employee portrait demand data includes the employee's personal personnel file data, project participation data, learning and training participation data, performance appraisal data, and evaluation scale assessment data.

[0035] The employee portrait data includes basic information, occupation information, ability information, performance information, behavior information and interest preference information.

[0036] The basic information includes personal information, educational background, and work experience, where personal information includes name, gender, and age; educational background includes academic qualifications, graduation school, and major studied; and work experience includes date of employment, date of leaving the company, years of work experience, and previous positions.

[0037] The occupational information includes position information, promotion records and training records; the position information includes the current position, position level and scope of responsibilities; the promotion records include the number of promotions, promotion time and the position after promotion; the training records include the number of trainings, training content and training results.

[0038] The capability information includes professional skills, soft skills and certification qualifications, wherein professional skills include skill name, proficiency and acquisition time; soft skills include communication skills, leadership skills, and teamwork problem-solving skills; and certification qualifications include certificate name, issuing agency and acquisition time.

[0039] The performance information includes assessment results, target completion status, rewards and penalties, among which assessment results include annual assessment, quarterly assessment and monthly assessment; target completion status includes key performance indicators (KPIs) and target achievement rate; rewards and penalties include the number of rewards, reasons for rewards, number of penalties and reasons for penalties.

[0040] The behavioral information includes attendance records, project participation, work habits and learning behavior, among which attendance records include the number of attendance days, number of latenesses and number of early departures; project participation includes the number of projects participated in, project roles, project completion status, number of team project discussions participated in, teamwork scores, communication ability scores and frequency of use of collaboration tools (such as Slack, Trello, etc.); work habits include work enthusiasm, overtime and work efficiency; learning behavior includes learning frequency (daily, weekly, monthly), learning time (hours) and learning methods (online courses, offline training, self-study).

[0041] The interest preference information includes career interests, learning preferences and extracurricular interests, wherein career interests include career planning and interest areas; learning preferences include preferred learning methods and preferred learning content; and extracurricular interests include personal hobbies and participation in extracurricular activities.

[0042] The project portrait unit is used to analyze project text data based on company project data using the BERT model, extract keywords and topics, and generate a project portrait.

[0043] Projects are categorized based on project profiles. By analyzing historical data on performance evaluations and competency assessments after completing specific projects, the actual impact of different projects on employee competency development (i.e., the primary degree of actual impact) is assessed. Project summaries also collect feedback from employees who have participated in different projects to understand their growth and gains from these projects.

[0044] The project portrait includes the project name, project type, project goals, project cycle, project employee data, project feedback information (employee feedback information), project evaluation information (leadership evaluation information) and the project's five-dimensional standards. The project's five-dimensional standards include performance dimension project impact standards, capability dimension project impact standards, behavior dimension project impact standards, potential dimension project impact standards and professional quality dimension project impact standards.

[0045] The project impact standard in the performance dimension refers to the impact of the project on employees' work performance, goal achievement rate and task completion quality; the project impact standard in the ability dimension refers to the impact of the project on employees' professional skills, communication skills, leadership skills, teamwork, problem-solving skills and innovation capabilities; the project impact standard in the behavior dimension refers to the impact of the project on employees' work attitude, teamwork and sense of responsibility; the project impact standard in the potential dimension refers to the impact of the project on employees' learning ability, development potential and adaptability; the project impact standard in the professional quality dimension refers to the impact of the project on employees' professional ethics, professional accomplishment and career stability.

[0046] The training portrait unit is used to analyze training text data based on the company's training data using the BERT model, extract keywords and topics, and generate training portraits.

[0047] Training is categorized based on training profiles. By analyzing historical data on performance evaluations and competency assessments after specific training sessions, the actual impact of different training sessions on employee competency development (i.e., the second level of actual impact) is assessed. Training summaries also collect feedback from employees who have participated in different training sessions to understand their growth and gains from these sessions.

[0048] The training portrait includes the training name, training type, training objectives, training cycle, training participant employee data, training feedback information (employee feedback information), training evaluation information (leadership evaluation information) and the five-dimensional training standards. The five-dimensional training standards include the performance dimension training impact standards, the ability dimension training impact standards, the behavior dimension training impact standards, the potential dimension training impact standards and the professional quality dimension training impact standards.

[0049] The training impact standard in the performance dimension refers to the impact of training on employees' work performance, goal achievement rate and task completion quality; the training impact standard in the ability dimension refers to the impact of training on employees' professional skills, communication skills, leadership skills, teamwork, problem-solving skills and innovation capabilities; the training impact standard in the behavior dimension refers to the impact of training on employees' work attitude, teamwork and sense of responsibility; the training impact standard in the potential dimension refers to the impact of training on employees' learning ability, development potential and adaptability; the training impact standard in the professional quality dimension refers to the impact of training on employees' professional ethics, professional accomplishment and career stability.

[0050] This embodiment generates portrait data of employees, projects, and training, which can more comprehensively understand and analyze the multi-dimensional information of employees, projects, and training, so as to facilitate more accurate human resource management and decision support, and achieve: 1) Optimize employee allocation: Based on the ability information and performance information in the employee portrait, reasonably arrange employees' jobs and project participation, improve work efficiency and team collaboration. 2) Personalized training: Use training portraits to analyze employees' training needs, design training plans that meet their personal career development and interest preferences, and enhance training effectiveness. 3) Project management: Through project portraits, evaluate the impact of projects on employee capacity building, optimize project management strategies, and improve project success rates. 4) Performance evaluation: Combine the performance dimension standards of employees and projects to conduct more objective performance evaluations, providing a basis for employee promotion and rewards. 5) Career planning: Employees can plan their personal career development paths based on their career interests and learning preferences in combination with portrait information. 6) Data-driven decision-making: The detailed data support provided by the portrait module helps companies make more scientific human resource decisions based on data.

[0051] The employee evaluation module is used to evaluate from five dimensions: performance dimension, ability dimension, behavior dimension, potential dimension and professional quality dimension based on the portrait data generated by the employee portrait unit.

[0052] The employee evaluation module includes a performance evaluation unit, an ability evaluation unit, a behavior evaluation unit, a potential evaluation unit and a professional quality evaluation unit.

[0053] The performance evaluation unit is used to utilize the annual assessment, quarterly assessment, monthly assessment, key performance indicators (KPIs), target achievement rate, number of rewards, reasons for rewards, number of penalties, and reasons for penalties in the employee portrait data, and use a multivariate regression model to construct an employee performance evaluation unit model to calculate the employee's performance score.

[0054] The competency assessment unit utilizes the skill name, proficiency level, acquisition time, communication skills, leadership skills, teamwork skills, problem-solving skills, certification name, issuing institution, and acquisition time in the employee profile data to construct a competency hierarchy model using the Analytic Hierarchy Process (AHP) and determine the weight of each indicator. A fuzzy comprehensive evaluation method is used to score each competency indicator and calculate a comprehensive competency score.

[0055] The behavior assessment unit is used to utilize the employee portrait data, including attendance days, number of latenesses, number of early departures, number of projects participated in, project roles, project completion status, number of team projects participated in, teamwork scores, communication skills scores, frequency of collaboration tool use, work enthusiasm, overtime, and work efficiency data, to use a behavior scoring method to define and quantify the scoring criteria for each behavior indicator and calculate the behavior score.

[0056] The potential assessment unit is used to construct a potential assessment unit matrix using the employee profile data, including years of service, previous positions, current position, job level, scope of responsibilities, number of promotions, promotion time, position after promotion, number of training sessions, training content, training effectiveness, learning frequency, learning duration, learning methods, and career planning data. Using the potential matrix method, the unit weights each indicator. The potential value assessment method is then used to calculate the employee's potential score.

[0057] The professional quality assessment unit is used to use the career planning, interest areas, preferred learning methods, preferred learning content, personal hobbies, and extracurricular activities data in the employee portrait data to score the employee's professional quality using a professional quality assessment scale; and to assess the employee's professional interests and preferences through a career interest test.

[0058] The training path recommendation module includes a job profile unit and a training path generation unit.

[0059] The job portrait unit uses the BERT model to analyze text data based on the company's job data and enterprise demand data, extract keywords and topics, and generate complete job portrait data containing job responsibilities and job requirements data (including basic requirements, skill requirements, performance indicator requirements and behavioral requirements).

[0060] The job data includes basic job information, job responsibilities, basic requirements, skill requirements, performance indicator requirements, ability requirements and training and development data, among which basic job information includes job title, job level, department and direct superior; job responsibilities include main work content, key responsibilities and daily tasks; basic requirements include educational requirements, professional background and work experience; skill requirements include essential skills, bonus skills and proficiency requirements; performance indicator requirements include key performance indicators (KPIs) and target achievement rate; ability requirements include teamwork, communication skills, leadership skills and problem-solving skills; training and development includes training plans and promotion paths.

[0061] Use K-Means clustering or hierarchical clustering algorithms to classify similar positions and form job categories. Utilize rule-based generation algorithms and template matching algorithms to create complete job portraits containing job responsibilities and job requirements data (including basic requirements, skill requirements, performance indicator requirements, and behavioral requirements), providing data support for developing specific training paths.

[0062] The development path generation unit initializes an employee node based on the employee portrait data in the portrait module, the evaluation results of the employee evaluation module, and the position portrait data, and assigns the evaluation results of the employee evaluation module to the initialized employee node. Using the project portrait data and training project portrait data in the portrait module, each project or training is treated as a node to form a fully connected graph. Using the graph embedding algorithm Node2Vec, the initialized nodes are randomly moved around the graph, assigning node attributes to each employee node during the movement to generate a node sequence that satisfies the employee and position match. A Markov decision process (MDP) algorithm is then used to generate a personalized career development path for each employee.

[0063] The state of the Markov decision process represents the current career development status of the employee. The corresponding data includes job title, core skills, performance, recent training experience, and development potential information. Action refers to the path that employees can choose in the process of career growth, including participation in training, projects, promotions, or job changes. Rewards reflect the benefits of employees after performing a certain action. Project portrait data, training portrait data, and the company's complete job portrait data contain the actual impact of employee participation in project training and job rotation on employee capacity improvement. The points for participating in the project, the assessment results of the training, and the results of the rotation assessment are converted into reward points for each action. The measurement indicators include performance improvement, capacity improvement, and changes in promotion probability. The ultimate goal is to search for the path with the highest cumulative reward in the fully connected career development graph to obtain the optimal personalized training plan.

[0064] By using projects and training as nodes, a fully connected network for employee career development is formed, comprehensively reflecting the various career development possibilities. Using the MDP algorithm for path planning within this network, an optimal development path is generated for each employee. This path not only considers the employee's current capabilities and job requirements, but also their future potential and direction. This approach ensures that employee development paths are scientifically based and personalized, helping employees continuously advance in their careers while also cultivating more talent that meets the organization's future needs.

[0065] The early warning module uses collaborative filtering and cosine similarity to calculate the employee-position match based on employee profiles, assessment results, and position profile data. If the match falls below α, the system automatically triggers an alert based on preset thresholds and rules, when certain employee indicators show abnormalities or meet alert conditions. Once an alert is triggered, the development path generation unit will readjust the development path.

[0066] The memory module generates the learning history of employees based on employee portrait data and training portrait data, and establishes a personalized knowledge file (i.e., employee personalized knowledge file). BERT is used to model the knowledge learned by employees to form a deeper level of abstract representation; by establishing a knowledge graph, knowledge points in different fields are linked to form a complete learning system. Important concepts and key knowledge nodes are identified in the knowledge graph; the training path generation unit adjusts the training path based on the employee's learning history, knowledge level, and position to achieve personalized training path recommendations. According to the employee's learning history and knowledge graph, the training path is continuously adjusted to adapt to the employee's learning progress and changing knowledge needs. Regularly evaluate the learning effect of employees, adjust the training path, and regularly evaluate employees in five dimensions through the employee evaluation module.

[0067] This embodiment constructs a comprehensive full-cycle career training system for enterprise employees, aiming to provide employees with personalized, efficient and scientific training paths through effective artificial intelligence technology, realize personalized and systematic training for employee career development, and provide enterprises with accurate training management and evaluation methods.

[0068] Example 2

[0069] like Figure 2 As shown, the career life cycle personalized training method based on Markov decision-making provided in the embodiment of the present application includes the following steps 201 to 204. Among them:

[0070] Step 201, obtain portrait data, wherein the portrait data includes employee portrait data, project portrait data, training portrait data and enterprise complete position portrait data, the project portrait data includes the first actual impact degree of different projects on employee capability improvement, the training portrait data includes the second actual impact degree of different trainings on employee capability improvement, and the enterprise complete position portrait data includes job responsibilities and job requirements data.

[0071] Step 202: Conduct a multi-dimensional assessment of the employee's current capabilities based on the employee portrait data to obtain an assessment result, wherein the multi-dimensional assessment includes a performance dimension, an ability dimension, a behavior dimension, a potential dimension, and a professional quality dimension.

[0072] Step 203: Based on the portrait data and the evaluation results, a graph embedding algorithm is used to construct a fully connected graph of career development, wherein the fully connected graph of career development includes employee nodes, project nodes, training nodes, target position nodes and edges, and the weight of the edge represents the probability of ability improvement or the extent of ability improvement, and the weight of the edge is a variable determined based on the first actual impact degree and the second actual impact degree.

[0073] Step 204 , taking any employee node as a starting point, searching for the path with the highest cumulative reward in the career development fully connected graph based on a Markov decision process, and obtaining the optimal employee training path.

[0074] Step 204-1 models the Markov decision process to obtain a modeled Markov decision process. The state of the Markov decision process represents the employee's current career development status, with corresponding data including job grade, core skills, performance, recent training experience, and development potential. Actions refer to the paths an employee can take in their career development, including participation in training, projects, promotion, or job change. Rewards reflect the benefits of an employee performing a specific action. Project profile data, training profile data, and the company's complete job profile data contain the actual impact of employee participation in project training and job rotation on employee competency improvement. Project participation points, training assessment results, and job rotation assessment results are converted into reward points for each action. Metrics include performance improvement, competency improvement, and change in promotion probability. The ultimate goal is to search the fully connected career development graph for the path with the highest cumulative reward and obtain the optimal personalized development plan.

[0075] Step 204 - 2 : searching for the optimal employee development path in the career development fully connected graph according to the modeled Markov decision process.

[0076] Example 3

[0077] This embodiment provides a computer device, which can be a server or a terminal. The computer device includes a processor, a memory, an input / output interface (I / O), and a communication interface. The processor, memory, and I / O interface are connected via a system bus, and the communication interface is connected to the system bus via the I / O interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store various data in the career development method based on artificial intelligence. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements the career development method based on Markov decision making in Example 2.

[0078] Example 4

[0079] This embodiment provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the method for personalized career development throughout the entire cycle based on Markov decision-making in embodiment 2 is implemented.

[0080] Example 5

[0081] This embodiment provides a computer program product, including a computer program, which, when executed by a processor, implements the Markov decision-based personalized career development method of embodiment 2.

[0082] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0083] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Any reference to memory, database or other media used in the embodiments provided in this application is omitted.

[0084] The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. The non-relational database may include, but is not limited to, a distributed database based on blockchain.

[0085] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0086] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A career cycle personalized training method based on Markov decision making, characterized by: The Markov decision-making-based personalized career development method includes: Obtaining portrait data, wherein the portrait data includes employee portrait data, project portrait data, training portrait data, and complete enterprise position portrait data, wherein the project portrait data includes a first actual impact degree of different projects on employee capability improvement, the training portrait data includes a second actual impact degree of different training on employee capability improvement, and the complete enterprise position portrait data includes position responsibilities and position requirements data; Conducting a multi-dimensional assessment of the employee's current capabilities based on the employee portrait data to obtain an assessment result, wherein the multi-dimensional assessment includes a performance dimension, an ability dimension, a behavior dimension, a potential dimension, and a professional quality dimension; Based on the portrait data and the evaluation results, constructing a fully connected career development graph using a graph embedding algorithm, wherein the fully connected career development graph includes employee nodes, project nodes, training nodes, target position nodes, and edges, and the weights of the edges are variables determined based on the first actual impact degree and the second actual impact degree; Taking any employee node as the starting point, the path with the highest cumulative reward is searched in the career development fully connected graph based on the Markov decision process to obtain the optimal employee training path.

2. The career life cycle personalized training method based on Markov decision making according to claim 1 is characterized in that: Based on the Markov decision process, the path with the highest cumulative reward is searched in the fully connected career development graph to obtain the optimal employee training path, which specifically includes: Modeling the Markov decision process to obtain a post-modeled Markov decision process, wherein the state in the post-modeled Markov decision process represents the employee's current career development status, including job grade, core skills, performance, recent training experience, and development potential information; the action in the post-modeled Markov decision process represents the employee's next action, including participation in training, projects, promotion, or job change; and the reward in the post-modeled Markov decision process represents the benefit of the employee after performing a certain action; The optimal employee training path is searched in the career development fully connected graph according to the modeled Markov decision process.

3. The career life cycle personalized training method based on Markov decision making according to claim 1 is characterized in that: After executing the step of "searching for the path with the highest cumulative reward in the career development fully connected graph based on a Markov decision process to obtain the optimal employee training path", the career life cycle personalized training method based on Markov decision further includes: Calculate the matching degree between employees’ current capabilities and target positions; If the matching degree is less than the matching degree threshold, the optimal employee training path is replanned in the career development fully connected graph based on the Markov decision process.

4. The career life cycle personalized training method based on Markov decision making according to claim 1, characterized in that: After executing the step of "searching for the path with the highest cumulative reward in the career development fully connected graph based on a Markov decision process to obtain the optimal employee training path", the career life cycle personalized training method based on Markov decision further includes: The optimal employee training path is adjusted according to the knowledge graph and the employee's personalized knowledge profile, wherein the employee's personalized knowledge profile is used to record the employee's learning history, and the employee's personalized knowledge profile is data generated based on the employee portrait data and the training portrait data.

5. The career life cycle personalized training method based on Markov decision making according to claim 1 is characterized in that: The process of obtaining the portrait data specifically includes: Obtain employee profile demand data, project data, training data, enterprise position data, and enterprise demand data, wherein the employee profile demand data includes employee personnel file data, project data participated in, training data participated in, performance appraisal data, and pre-training assessment scale evaluation data; Generating the employee portrait data using a natural language processing model based on the employee portrait demand data; Generate the project portrait data using a natural language processing model based on the project data; Generating the training profile data using a natural language processing model based on the training data; Based on the enterprise job data and enterprise demand data, the natural language processing model is used to generate the complete job portrait data of the enterprise.

6. The career life cycle personalized training method based on Markov decision making according to claim 5, characterized in that: The employee profile data includes employee basic information, occupation information, ability information, performance information, behavior information and interest preference information; The project profile data also includes project name, project type, project goal, project cycle, project employee data, project feedback information, project evaluation information and project five-dimensional standards, which include performance dimension project impact standards, ability dimension project impact standards, behavior dimension project impact standards, potential dimension project impact standards and professional quality dimension project impact standards; The training portrait data also includes training name, training type, training objectives, training cycle, training participant employee data, training feedback information, training evaluation information and five-dimensional training standards. The five-dimensional training standards include performance dimension training impact standards, ability dimension training impact standards, behavior dimension training impact standards, potential dimension training impact standards and professional quality dimension training impact standards.

7. The career life cycle personalized training method based on Markov decision making according to claim 6 is characterized in that: Conduct a multi-dimensional assessment of employees based on the employee portrait data to obtain assessment results, including: Processing the performance information according to an employee performance evaluation model to obtain a performance score, wherein the employee performance evaluation model is a multiple regression model constructed with the performance information as an independent variable and the performance score as a dependent variable; Using fuzzy comprehensive evaluation method to process the ability information and calculate the ability score; The behavioral information is processed using a behavioral anchor rating method to calculate a behavioral score; calculating a potential score based on the occupational information and the learning behavior data in the behavior information; Calculating a professional quality score based on the interest preference information; The performance score, the ability score, the behavior score, the potential score and the professional quality score are used as the evaluation results.

8. A career cycle personalized training system based on Markov decision making, characterized by: The career cycle personalized training system based on Markov decision-making includes: A data acquisition module for acquiring portrait data, wherein the portrait data includes employee portrait data, project portrait data, training portrait data, and complete enterprise position portrait data. The project portrait data includes the first actual impact of different projects on employee capability improvement, the training portrait data includes the second actual impact of different trainings on employee capability improvement, and the complete enterprise position portrait data includes job responsibilities and job requirements data. An employee evaluation module is used to perform a multi-dimensional evaluation of an employee's current capabilities based on the employee portrait data to obtain an evaluation result, wherein the multi-dimensional evaluation includes a performance dimension, an ability dimension, a behavior dimension, a potential dimension, and a professional quality dimension; A path planning module is used to construct a fully connected career development graph based on the portrait data and the evaluation results using a graph embedding algorithm, and take any of the employee nodes as a starting point, and search for the path with the highest cumulative reward in the fully connected career development graph based on a Markov decision process to obtain the optimal employee training path, wherein the fully connected career development graph includes employee nodes, project nodes, training nodes, target position nodes and edges, and the weights of the edges are variables determined based on the first actual impact level and the second actual impact level.

9. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the career life-cycle personalized training method based on Markov decision-making as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the Markov decision-making-based personalized career development method according to any one of claims 1 to 7 is implemented.

Images