A performance management method based on data analysis

Abstract

The present invention discloses a performance management method based on data analysis, which relates to the technical field of performance management and includes the following steps: S1, obtaining a performance report, extracting key fields of the performance report, and performing annotation to obtain the latest performance report; S2, generating task parameters; S3, generating a parsing sequence; S4, generating a management key for the user ID; S5, when the user receives the latest performance report, inputting the management key to obtain the latest performance report. The present invention realizes secure access control of performance data, effectively prevents data leakage and abuse, and the user only needs to input the management key to obtain the latest performance report, improving the work efficiency of enterprise management.

Description

Technical Field

[0001] The present invention relates to the technical field of performance management, and particularly relates to a performance management method based on data analysis. Background Art

[0002] In today's highly competitive business environment, enterprises' demand for efficient and accurate performance management is increasing. In the traditional mode, the storage and access of performance data usually lack strict security mechanisms and permission controls. Employee performance data may be accessed, tampered with, or leaked randomly, bringing potential risks to both enterprises and employees. For example, once sensitive information such as employees' salaries and promotions is leaked, it may trigger internal conflicts and trust crises.

[0003] Existing management methods often adopt fixed rules and algorithms and lack adaptability to different business scenarios and user needs. For example, for employees in different departments and positions, the key points and methods of performance evaluation may vary, but existing encryption management methods are difficult to adjust flexibly according to these differences.

[0004] Based on the above background, in order to overcome the limitations of traditional performance management methods and improve the security of performance management, the present invention proposes a performance management method based on data analysis. Summary of the Invention

[0005] The present invention proposes a performance management method based on data analysis to solve the above problems.

[0006] The technical solution of the present invention is: A performance management method based on data analysis includes the following steps:

[0007] S1. Obtain a performance report, extract the key fields of the performance report, and perform annotation to obtain the latest performance report;

[0008] S2. Generate task parameters according to the graph data of the latest performance report and several task lists;

[0009] S3. Determine the query vector representation according to the parsing instruction of the latest performance report and generate a parsing sequence;

[0010] S4. Obtain the user ID that receives the latest performance report, and generate a management key for the user ID according to the task parameters and parsing sequence of the latest performance report;

[0011] S5. When the user receives the latest performance report, input the management key to obtain the latest performance report.

[0012] Further, S2 includes the following sub-steps:

[0013] S21. Construct graph data for the latest performance report, and extract several entity nodes included in each task list in the graph data; wherein, the entity nodes include task name, person in charge name, and department name.

[0014] S22. Construct relationship objects for each entity node of the task list.

[0015] S23. Generate task parameters for the latest performance report according to the relationship objects corresponding to each entity node of the task list.

[0016] The beneficial effects of the above further solution are: In the present invention, by constructing graph data, the information in the latest performance report is structurally represented in the form of nodes and edges, which can comprehensively cover various elements in performance management, such as tasks, persons in charge, and departments. Accurately extracting the entity nodes included in each task list in the graph data, including task name, person in charge name, and department name, ensures the accuracy and integrity of the data. These entity nodes are the core elements of performance management, and their accurate extraction is crucial for the subsequent generation of task parameters.

[0017] Further, in S22, the relationship object of the task name has the following expression:

[0018] ;

[0019] In the formula, represents the task completion rate corresponding to the task name, represents the closeness centrality of the entity node corresponding to the task name, represents a random constant between 0 and 1;

[0020] The task completion rate is used to count the completion situation of business tasks and reflect the progress of achieving business goals. The relationship object expression of the task name comprehensively considers the closeness centrality of the task in the graph data and the task completion rate, and introduces a random constant to increase the flexibility of the model.

[0021] In S22, the relationship object of the person in charge name has the following expression:

[0022] ;

[0023] In the formula, represents the ability dimension parameter, represents the management dimension parameter, represents the technical dimension parameter, represents the compliance dimension parameter, represents the closeness centrality of the entity node corresponding to the person in charge name;

[0024] The parameters of the person in charge can be scored according to their capabilities. Among them, the parameter dimensions of capabilities include goal alignment ability (such as whether cloud resource usage can be aligned with business goals), technical decision-making ability (such as whether reasonable technical decisions can be made based on the characteristics of the cloud platform), and technical decision-making ability; the parameter dimensions of management include cross-team collaboration, risk control, and risk control; the parameter dimensions of technology include automation ability and monitoring and alerting ability; the parameter dimensions of compliance include permission management and auditing, etc.

[0025] In S22, the relationship object of the department name The expression is:

[0026] ;

[0027] In the formula, represents the closeness centrality of the entity node corresponding to the department name, represents the responsibility weight of the department.

[0028] The department responsibility weight is a quantitative indicator used in organizational management to measure and allocate the importance of the responsibilities assumed by each department in the process of achieving organizational goals, and can be set artificially.

[0029] Furthermore, S23 includes the following sub-steps:

[0030] S231. Combine the relationship object of the task name, the relationship object of the person in charge name, and the relationship object of the department name to obtain the relationship column vector of the task list;

[0031] S232. Calculate the Manhattan distance between the relationship column vectors of every two task lists in the latest performance report to obtain a number of distance parameters;

[0032] S233. Take the average value of all distance parameters as the task parameter of the latest performance report.

[0033] The beneficial effect of the above further solution is: In the present invention, by combining the relationship object of the task name, the relationship object of the person in charge name, and the relationship object of the department name to obtain the relationship column vector of the task list, the comprehensive integration of information in multiple aspects such as tasks, person in charge, and departments is realized. The generated task parameters can be used for analysis between different task lists.

[0034] Furthermore, S3 includes the following sub-steps:

[0035] S31. Obtain the parsing instruction of the latest performance report and determine the SQL query statement corresponding to the parsing instruction;

[0036] S32. Perform encoding processing on the SQL query statement to obtain the query vector representation;

[0037] S33. Generate a parsing sequence based on the query vector representation.

[0038] The beneficial effect of the above further solution is as follows: In the present invention, by obtaining the parsing instruction of the latest performance report and determining the corresponding SQL query statement, the accurate capture of the user's intention and the conversion into an executable data query operation are realized. Different parsing instructions can correspond to different SQL query statements, and the vector representation can effectively represent the key features of the SQL query. Extract the L2 norm of the query vector representation, divide it by the immediate number of the parsing instruction, and then randomly insert the operation result into the query vector representation, which further enriches the information content of the vector, can introduce additional semantic information, enables the parsing sequence to better reflect the complex features of the SQL query statement and the user's intention, and improves the quality of the parsing sequence.

[0039] In S32, the SQL query is regarded as a sequence and encoded using a recurrent neural network (RNN) or its variants (LSTM, GRU), and the last hidden state is output as the vector representation of the entire query. The vector representation is generally in the form of a row vector and contains several elements.

[0040] Further, S33 includes the following sub-steps:

[0041] S331. Extract the L2 norm of the query vector representation;

[0042] S332. Divide the immediate number of the parsing instruction by the L2 norm, randomly insert the operation result into the query vector representation, and convert it into a parsing sequence.

[0043] The query vector representation is in the form of a row vector, which is a set of elements. After inserting the operation result, it is converted into a common sequence form.

[0044] Further, S4 includes the following sub-steps:

[0045] S41. Obtain the user ID of the user who receives the latest performance report;

[0046] S42. Obtain the management parameter according to the task parameter and the parsing sequence;

[0047] S44. Combine the user ID and the management parameter as the management key.

[0048] The beneficial effect of the above further solution is as follows: In the present invention, obtaining the user ID of the user who receives the latest performance report can clarify the identity of the data recipient. By performing a summation operation on the hash value of the task parameter and the hash value of the sequence parameter, etc., the multi-dimensional comprehensive evaluation of the performance data is realized. Combining the user ID and the management parameter as the password.

[0049] Further, in S42, an averaging operation is performed on the parsing sequence of the latest performance report to obtain a sequence parameter. The hash values of the task parameter, the sequence parameter, and the user ID are summed to obtain a management parameter.

[0050] The beneficial effects of the present invention are as follows: The present invention automatically extracts and labels the key fields of the performance report, generates task parameters based on the graph data and task list of the latest performance report, and generates a query vector representation and a parsing sequence according to the parsing instruction, enabling the present invention to be flexibly adjusted according to different business scenarios and user requirements, improving the pertinence and effectiveness of performance management; The present invention also combines the user ID for obtaining the performance report to generate a management key and verifies it when the user receives the latest performance report, realizing secure access control of performance data, effectively preventing data leakage and abuse. The user only needs to input the management key to obtain the latest performance report, improving the work efficiency of enterprise management. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a flowchart of a performance management method based on data analysis. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0053] As Figure 1 shown, the present invention provides a performance management method based on data analysis, including the following steps:

[0054] S1. Obtain a performance report, extract the key fields of the performance report, and perform annotation to obtain the latest performance report;

[0055] S2. Generate task parameters according to the graph data of the latest performance report and several task lists;

[0056] S3. Determine the query vector representation according to the parsing instruction of the latest performance report and generate a parsing sequence;

[0057] S4. Obtain the user ID for receiving the latest performance report, and generate a management key for the user ID according to the task parameter and parsing sequence of the latest performance report;

[0058] S5. When the user receives the latest performance report, input the management key to obtain the latest performance report.

[0059] In the embodiment of the present invention, S2 includes the following sub-steps:

[0060] S21. Construct graph data for the latest performance report, and extract several entity nodes included in each task list in the graph data; wherein, the entity nodes include task name, person in charge name, and department name;

[0061] S22. Construct a relationship object for each entity node in the task list;

[0062] S23. Generate task parameters for the latest performance report according to the relationship objects corresponding to the entity nodes in the task list.

[0063] In the present invention, by constructing graph data, the information in the latest performance report is structurally represented in the form of nodes and edges, which can comprehensively cover various elements in performance management, such as tasks, responsible persons, and departments. Accurately extracting the entity nodes included in each task list in the graph data, including task names, responsible person names, and department names, ensures the accuracy and integrity of the data. These entity nodes are the core elements of performance management, and their accurate extraction is crucial for subsequent task parameter generation.

[0064] In the embodiment of the present invention, in S22, the relationship object of the task name has the following expression:

[0065] ;

[0066] In the formula, represents the task completion rate corresponding to the task name, represents the closeness centrality of the entity node corresponding to the task name, represents a random constant between 0 and 1;

[0067] The task completion rate is used to count the completion of business tasks and reflects the progress of achieving business goals. The expression of the relationship object of the task name comprehensively considers the closeness centrality of the task in the graph data and the task completion rate, and introduces a random constant to increase the flexibility of the model.

[0068] In S22, the relationship object of the responsible person name has the following expression:

[0069] ;

[0070] In the formula, represents the ability dimension parameter, represents the management dimension parameter, represents the technology dimension parameter, represents the compliance dimension parameter, represents the closeness centrality of the entity node corresponding to the responsible person name;

[0071] The parameters of the person in charge can be scored and evaluated according to their capabilities. Among them, the parameter of the ability dimension includes the ability to align with the goal (such as whether the use of cloud resources can be aligned with the business goal), the technical decision-making ability (such as whether reasonable technical decisions can be made based on the characteristics of the cloud platform), and the technical decision-making ability; the parameter of the management dimension includes the ability of cross-team collaboration, risk control, and risk control; the parameter of the technical dimension includes the automation ability and the monitoring and alerting ability; the parameter of the compliance dimension includes the permission management and the audit.

[0072] In S22, the relationship object of the department name The expression is:

[0073] ;

[0074] In the formula, represents the closeness centrality of the entity node corresponding to the department name, represents the responsibility weight of the department.

[0075] The department responsibility weight is a quantitative index used in organizational management to measure and allocate the importance of the responsibilities borne by each department in the process of achieving organizational goals, and can be set artificially.

[0076] In the embodiment of the present invention, S23 includes the following sub-steps:

[0077] S231. Combine the relationship object of the task name, the relationship object of the person in charge name, and the relationship object of the department name to obtain the relationship column vector of the task list;

[0078] S232. Calculate the Manhattan distance between the relationship column vectors of every two task lists in the latest performance report to obtain a number of distance parameters;

[0079] S233. Take the average value of all distance parameters as the task parameter of the latest performance report.

[0080] In the present invention, by combining the relationship object of the task name, the relationship object of the person in charge name, and the relationship object of the department name to obtain the relationship column vector of the task list, the comprehensive integration of information in multiple aspects such as tasks, persons in charge, and departments is realized. The generated task parameters can be used for the analysis between different task lists.

[0081] In the embodiment of the present invention, S3 includes the following sub-steps:

[0082] S31. Obtain the parsing instruction of the latest performance report and determine the SQL query statement corresponding to the parsing instruction;

[0083] S32. Perform encoding processing on the SQL query statement to obtain the query vector representation;

[0084] S33. Generate an analysis sequence based on the query vector representation.

[0085] In the present invention, by obtaining the parsing instruction of the latest performance report and determining the corresponding SQL query statement, the accurate capture of the user's intention and the conversion into an executable data query operation are realized. Different parsing instructions can correspond to different SQL query statements, and the vector representation can effectively represent the key features of the SQL query. Extract the L2 norm of the query vector representation, divide it by the immediate number of the parsing instruction, and then randomly insert the operation result into the query vector representation, which further enriches the information content of the vector, can introduce additional semantic information, enables the analysis sequence to better reflect the complex features of the SQL query statement and the user's intention, and improves the quality of the analysis sequence.

[0086] In S32, the SQL query is regarded as a sequence and encoded using a recurrent neural network (RNN) or its variants (LSTM, GRU), and the last hidden state is output as the vector representation of the entire query. The vector representation is generally in the form of a row vector and contains several elements.

[0087] In the embodiment of the present invention, S33 includes the following sub-steps:

[0088] S331. Extract the L2 norm of the query vector representation;

[0089] S332. Divide the immediate number of the parsing instruction by the L2 norm, randomly insert the operation result into the query vector representation, and convert it into an analysis sequence.

[0090] The form of the query vector representation is a row vector, which is a set of elements. After inserting the operation result, it is then converted into a normal sequence form.

[0091] In the embodiment of the present invention, S4 includes the following sub-steps:

[0092] [[ID=;24]]S41. Obtain the user ID that receives the latest performance report;

[0093] S42. Obtain management parameters according to the task parameters and the analysis sequence;

[0094] S44. Combine the user ID and the management parameters as the management key.

[0095] In the present invention, obtaining the user ID that receives the latest performance report can clarify the identity of the data recipient. By performing a summation operation on the hash value of the task parameters and the hash value of the sequence parameters, etc., management parameters are obtained, realizing the multi-dimensional comprehensive evaluation of the performance data. Combining the user ID and the management parameters as the password.

[0096] In an embodiment of the present invention, in S42, an averaging operation is performed on the parsing sequence of the latest performance report to obtain a sequence parameter, and a summation operation is performed on the hash value of the task parameter, the hash value of the sequence parameter, and the hash value of the user ID to obtain a management parameter.

[0097] Those of ordinary skill in the art will realize that the embodiments described herein are to assist the reader in understanding the principles of the present invention, and it should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not depart from the essence of the present invention based on the technical revelations disclosed in the present invention, and these deformations and combinations are still within the scope of protection of the present invention.

Claims

1. A performance management method based on data analysis, characterized in that It includes the following steps: S1. Obtain a performance report, extract the key fields of the performance report, and perform annotation to obtain the latest performance report; S2. Generate task parameters according to the graph data of the latest performance report and several task lists; S3. Determine the query vector representation according to the parsing instruction of the latest performance report and generate a parsing sequence; S4. Obtain the user ID that receives the latest performance report, and generate a management key for the user ID according to the task parameters and parsing sequence of the latest performance report; S5. When the user receives the latest performance report, input the management key to obtain the latest performance report; The S2 includes the following sub-steps: S21. Construct graph data for the latest performance report, and extract several entity nodes included in each task list in the graph data; among them, the entity nodes include task names, person-in-charge names, and department names; S22. Construct relationship objects for each entity node of the task list; S23. Generate task parameters for the latest performance report according to the relationship objects corresponding to each entity node of the task list; The S3 includes the following sub-steps: S31. Obtain the parsing instruction of the latest performance report and determine the SQL query statement corresponding to the parsing instruction; S32. Perform encoding processing on the SQL query statement to obtain the query vector representation; S33. Generate a parsing sequence according to the query vector representation; The S4 includes the following sub-steps: S41. Obtain the user ID that receives the latest performance report; S42. Obtain management parameters according to the task parameters and parsing sequence; S44. Combine the user ID and management parameters as the management key.

2. The performance management method based on data analysis according to claim 1, wherein In S22, the relationship object of the task name has the following expression: ； In the formula, represents the task completion rate corresponding to the task name, represents the closeness centrality of the entity node corresponding to the task name, represents a random constant between 0 and 1; In S22, the relational object of the responsible person's name The expression is: ； In the formula, represents the ability dimension parameter, represents the management dimension parameter, represents the technology dimension parameter, represents the compliance dimension parameter, represents the closeness centrality of the entity node corresponding to the name of the person in charge; In S22, the relational object of the department name has the following expression: ； In the formula, represents the closeness centrality of the entity node corresponding to the department name, represents the responsibility weight of the department.

3. The performance management method based on data analysis according to claim 1, characterized in that The S23 includes the following sub-steps: S231. Combine the relationship object of the task name, the relationship object of the person-in-charge name, and the relationship object of the department name to obtain the relationship column vector of the task list; S232. Calculate the Manhattan distance between the relationship column vectors of every two task lists in the latest performance report to obtain several distance parameters; S233. Take the average value of all distance parameters as the task parameters of the latest performance report.

4. The performance management method based on data analysis according to claim 1, wherein The S33 includes the following sub-steps: S331. Extract the L2 norm of the query vector representation; S332. Divide the immediate number of the parsing instruction by the L2 norm, randomly insert the operation result into the query vector representation, and convert it into a parsing sequence.

5. The performance management method based on data analysis according to claim 1, wherein In the S42, perform an averaging operation on the parsing sequence of the latest performance report to obtain a sequence parameter, and perform a summation operation on the hash value of the task parameter, the hash value of the sequence parameter, and the hash value of the user ID to obtain the management parameter.

Images