CN-121998603-A - Method and device for predicting succession of trunk and recommending person decision, electronic equipment and storage medium

Abstract

The invention provides a method, a device, electronic equipment and a storage medium for stem succession prediction and personnel decision recommendation, which belong to the technical field of human resource management, in the method, multi-source data are considered, and the station vacancy probability, personnel loss risk and succession matching degree can be quantitatively predicted, the method has the advantages that the early warning is further realized, the relay recommended scheme corresponding to the early warning level is provided, the early warning can be performed in advance in the process, the relay recommended scheme is obtained through data driving and is independent of subjective experience of HR, the human intervention is reduced, and the method is more scientific.

Inventors

• You chaoyang
• SUN JIANG
• JI WEIGUO

Assignees

• 北森云计算有限公司

Dates

Publication Date

20260508

Application Date

20260123

Claims (10)

1. 1. The method for predicting the succession of the trunk and recommending the decision of the person is characterized by comprising the following steps: Acquiring multi-source data, wherein the sources of the multi-source data comprise HR system data, 360 evaluation results and organization structure relations; performing post vacancy probability prediction on post feature vectors in the multi-source data by adopting a post vacancy probability model to obtain the vacancy probability of each post; talent loss probability prediction is carried out on employee feature vectors in the multi-source data by adopting a talent loss probability model, so that loss probability of each employee is obtained; performing succession matching calculation on staff post maps in the multi-source data by adopting a succession matching degree model to obtain succession matching scores of staff and posts; Determining an early warning level according to the vacancy probability of each post, the loss probability of each employee and the succession matching score of the employee and the post; and determining a succession recommendation scheme corresponding to the early warning level.
2. 2. The method of claim 1, wherein the post feature vector comprises post responsible person age, post responsible person span, post history flow frequency, post importance, organizational hierarchy; The post vacancy probability model comprises: , wherein, Representation post At the time of The probability of a subsequent occurrence of a vacancy event, Representing a reference accumulated risk of occurrence, Representing the weight vector obtained by the training, Representing the post feature vector.
3. 3. The method of claim 1, wherein the employee feature vector comprises a last 12 month performance average, a competency match score, a training frequency, a superior variability frequency, and an attendance anomaly count; The talent loss probability model comprises: , wherein, Representing staff At the time of The probability of the inner departure from the job, Representing the output accumulated values of all decision trees, , The weights of the XGBoost models are represented, The offset is indicated as being a function of the offset, Representing the employee feature vector.
4. 4. The method of claim 1, wherein the staff position map comprises staff nodes, position nodes and edges among the nodes, wherein the edges represent the relationships among the nodes; the succession matching degree model comprises the following steps: , wherein, Representing staff And post with Is used to determine the successor match score of (c), Representing staff By passing through The characteristic representation after the layer GNN has propagated, Representation post By passing through The characteristics of the layer GNN after propagation are represented.
5. 5. The method of claim 1, wherein determining the pre-warning level based on the void probability for each post, the churn probability for each employee, and the employee and post successor match score comprises: if the vacancy probability of the target post is greater than a first preset threshold, or the loss probability of the target staff is greater than a second preset threshold and the succession matching score of the target staff and the post is less than a third preset threshold, determining that the early warning level is a high risk level; if the vacancy probability of the target post is greater than a fourth preset threshold and not greater than the first preset threshold, or the loss probability of the staff is greater than a fifth preset threshold and not greater than the second preset threshold, determining that the early warning level is a risk level.
6. 6. The method of claim 1, wherein determining a succession recommendation corresponding to the pre-warning level comprises: and if the early warning level is a high risk level, outputting a list of the highest pre-set staff corresponding to the target post in the succession matching score of the staff and the post to the target post corresponding to the high risk level.
7. 7. The method of claim 1, wherein determining a succession recommendation corresponding to the pre-warning level further comprises: If the early warning grade is the risk grade, a culture plan with preset duration is formulated.
8. 8. A trunk succession prediction and human decision recommendation device, comprising: The acquisition unit is used for acquiring multi-source data, wherein the sources of the multi-source data comprise HR system data, 360 evaluation results and organization structure relations; The post vacancy probability prediction unit is used for predicting post vacancy probability of the post feature vectors in the multi-source data by adopting a post vacancy probability model to obtain the vacancy probability of each post; the talent loss probability prediction unit is used for predicting talent loss probability of employee feature vectors in the multi-source data by adopting a talent loss probability model to obtain loss probability of each employee; The succession matching calculation unit is used for carrying out succession matching calculation on staff post maps in the multi-source data by adopting a succession matching degree model to obtain succession matching scores of staff and posts; The first determining unit is used for determining an early warning level according to the vacancy probability of each post, the loss probability of each employee and the succession matching score of the employee and the post; and the second determining unit is used for determining a succession recommended scheme corresponding to the early warning level.
9. 9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of any of the preceding claims 1 to 7 when the computer program is executed.
10. 10. A computer readable storage medium storing machine executable instructions which, when invoked and executed by a processor, cause the processor to perform the method of any one of the preceding claims 1 to 7.

Description

Method and device for predicting succession of trunk and recommending person decision, electronic equipment and storage medium Technical Field The invention relates to the technical field of human resource management, in particular to a method and a device for predicting succession of a trunk and recommending decision-making of people, electronic equipment and a storage medium. Background In the current organizational management scenario, enterprises increasingly rely on systematic, data-driven talent management approaches. The following problems are common in the traditional talent portrait construction and risk identification schemes: Most systems ignore the behavior and judgment of HRBP, business upper-level and other multi-roles in talent management process only based on employee own data (such as performance score and attendance record); the prediction capability is insufficient, and the quantized prediction of the station vacancy probability, the risk of loss of people and the succession matching degree is lacking. In conclusion, the traditional talent portrait construction and risk identification scheme has the technical problems of single role visual angle and insufficient prediction capability. Disclosure of Invention In view of the above, the present invention aims to provide a method, a device, an electronic device and a storage medium for stem succession prediction and personnel decision recommendation, so as to alleviate the technical problems of single role view and insufficient prediction capability of the conventional talent portrait construction and risk identification scheme. In a first aspect, an embodiment of the present invention provides a method for predicting a succession of a trunk and recommending a person decision, including: Acquiring multi-source data, wherein the sources of the multi-source data comprise HR system data, 360 evaluation results and organization structure relations; performing post vacancy probability prediction on post feature vectors in the multi-source data by adopting a post vacancy probability model to obtain the vacancy probability of each post; talent loss probability prediction is carried out on employee feature vectors in the multi-source data by adopting a talent loss probability model, so that loss probability of each employee is obtained; performing succession matching calculation on staff post maps in the multi-source data by adopting a succession matching degree model to obtain succession matching scores of staff and posts; Determining an early warning level according to the vacancy probability of each post, the loss probability of each employee and the succession matching score of the employee and the post; and determining a succession recommendation scheme corresponding to the early warning level. Further, the post feature vector comprises post responsible person age, post responsible person span, post history flow frequency, post importance and organization level; The post vacancy probability model comprises: , wherein, Representation postAt the time ofThe probability of a subsequent occurrence of a vacancy event,Representing a reference accumulated risk of occurrence,Representing the weight vector obtained by the training,Representing the post feature vector. Further, the employee feature vector comprises a latest 12-month performance average value, a competence matching degree total score, training frequency, a superior change frequency and attendance abnormal times; The talent loss probability model comprises: , wherein, Representing staffAt the time ofThe probability of the inner departure from the job,Representing the output accumulated values of all decision trees,,The weights of the XGBoost models are represented,The offset is indicated as being a function of the offset,Representing the employee feature vector. Further, the staff post map comprises staff nodes, post nodes and edges among the nodes, wherein the edges represent the relationship among the nodes; the succession matching degree model comprises the following steps: , wherein, Representing staffAnd post withIs used to determine the successor match score of (c),Representing staffBy passing throughThe characteristic representation after the layer GNN has propagated,Representation postBy passing throughThe characteristics of the layer GNN after propagation are represented. Further, determining an early warning level according to the vacancy probability of each post, the loss probability of each employee and the succession matching score of the employee and the post includes: if the vacancy probability of the target post is greater than a first preset threshold, or the loss probability of the target staff is greater than a second preset threshold and the succession matching score of the target staff and the post is less than a third preset threshold, determining that the early warning level is a high risk level; if the vacancy probability of the target post is greater than a fourth