Talent management method and system based on objective data evaluation

Abstract

The invention discloses a talent management method and system based on objective data evaluation, and relates to the technical field of data processing, and the method comprises the steps: analyzing the entropy change condition of each class of reflection indexes, integrating all classes of reflection indexes, determining the comprehensive quality level, describing the overall ability level of different persons through the comprehensive quality level, and carrying out the evaluation of the overall ability level. The objectivity of talent evaluation is improved, and the influence of subjective factors on talent ability evaluation is avoided, thereby helping to optimize talent management. The post information is analyzed to determine the post difference coefficient of each group post, the hyperbola attenuation model of each group post is established, the time change content of multiple factors in the post information is considered, the weight of each piece of post information is determined, the historical value and timeliness are balanced through hyperbola characteristics, and the accuracy of the hyperbola attenuation model is improved. The attenuation speed of the capability value of the post is described, and the capability portrait of each enterprise staff is constructed, so that capability comparison between different posts can be realized, and enterprise talent management and post distribution are optimized.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a talent management method and system based on objective data evaluation. Background Art

[0002] Using advanced technologies such as big data and artificial intelligence, we can comprehensively and deeply collect and analyze the human resources data of enterprises. This solution builds an information database to collect employees' identity information, performance data, training records, etc., forming a comprehensive talent data foundation. At the same time, we can enrich the dimensions of talent information with the help of multi-channel data such as social media and market research. Using data analysis models and data mining technology, we can deeply explore the laws and trends behind the data and provide a scientific basis for the talent management of enterprises.

[0003] The program aims to optimize human resource allocation, improve talent management, develop personalized training plans, and reasonable incentive strategies through objective data evaluation. Its advantage is that it can overcome the shortcomings of traditional talent management methods, which are highly subjective and lack scientific basis, and make talent management more scientific, accurate, and efficient. Through this program, companies can better understand the capabilities and potential of employees, achieve job matching, improve overall performance, and enhance the competitiveness of the company.

[0004] In the prior art, the evaluation of enterprise personnel is often based on the subjective evaluation and judgment of superiors, which leads to poor objectivity and low accuracy of talent evaluation, and is not conducive to the effective and adaptive management of talents by enterprises.

[0005] Therefore, how to improve the objectivity and accuracy of talent evaluation is a technical problem that needs to be solved. Summary of the invention

[0006] The purpose of the present invention is to solve the problem of poor objectivity and low accuracy of talent evaluation in the prior art, and to propose a talent management method based on objective data evaluation, which includes: Conduct multi-dimensional and comprehensive digital monitoring of enterprise personnel and generate enterprise personnel activity logs, which contain structured and unstructured data. Indicators are extracted from structured and unstructured data. Analyze the entropy changes of each type of reflection index, and integrate all types of reflection indexes to obtain the comprehensive quality level of enterprise personnel; Divide the enterprise personnel into groups according to business positions, collect the position information of the enterprise personnel in each group position, analyze the position information to determine the position difference coefficient of each group position, and establish a hyperbolic attenuation model for each group position; Through the comprehensive quality level of enterprise personnel and the hyperbolic attenuation model of group positions, we construct the ability profile of each enterprise personnel, so as to achieve the management of enterprise talents.

[0007] In some embodiments of the present application, reflection indicators are extracted from structured data and unstructured data, including: For structured data, determine the value information set in the structured data, expand the value information set through the relevance of the value information set, extract the expanded value information set in the structured data, and extract the reflection indicator from the expanded value information set; For unstructured data, valuable information sets are extracted from the unstructured data through multi-source technology. Multi-source technology includes one or more of text mining, natural language processing, and image and audio analysis to extract reflection indicators from the valuable information sets.

[0008] In some embodiments of the present application, the entropy change of each type of reflection indicator is analyzed, including: Preset multiple typical time periods for each type of reflection indicator, determine the degree of dispersion of each type of reflection indicator, determine the number of lags in the autocorrelation function of each type of reflection indicator according to the degree of dispersion, thereby calculating the autocorrelation value of each type of reflection indicator, and draw an autocorrelation graph of the reflection indicator, and select a typical time period from multiple typical time periods as a standard time period of the reflection indicator according to the autocorrelation graph of the reflection indicator; The mean and change rate of information entropy are calculated according to the standard time period of each type of reflecting indicator, and the mean and change rate of information entropy reflecting the indicator are taken as the entropy change.

[0009] In some embodiments of the present application, a typical time period is selected from a plurality of typical time periods as a standard time period reflecting the indicator according to the autocorrelation diagram reflecting the indicator, including: Identify multiple significant peaks on the autocorrelation graph reflecting the index, determine the peak value, peak position and peak shape of each significant peak, the peak shape includes the peak width and the rate of change of the autocorrelation value near the peak; The peak width and the rate of change of the autocorrelation value near the peak are combined to determine the sharpness and smoothness of the peak. The peak description is determined by the sharpness and smoothness and the peak value of the peak. Multiple significant peaks are ranked according to the peak description. According to the peak position of the first-ranked significant peak, a similar typical time period is selected from multiple typical time periods as the standard time period reflecting the indicator.

[0010] In some embodiments of the present application, all categories of reflection indicators are integrated to obtain the comprehensive quality level of enterprise personnel, including: Standardize all categories of reflection indicators and determine the comprehensive quality level of enterprise personnel based on the mean and change rate of information entropy of reflection indicators; ; in, For the comprehensive quality level of enterprise personnel, To reflect the number of indicator categories, For the The mean information entropy of the indicator, represents the sum of the normalized information entropy means of all indicators, Indicates The weight of the indicator is reflected. For the The size of the indicator is reflected. For the The rate of change of the information entropy of the indicator is For the The first constant reflecting the indicator.

[0011] In some embodiments of the present application, analyzing the position information to determine the position difference coefficient of each group position includes: Position information includes information on technology changes, market dynamics, and risk fluctuations; Analyze technology change information, market dynamic change information and risk fluctuation information respectively, and score technology change information, market dynamic change information and risk fluctuation information to define job stability level; ; in, For job stability level, To stabilize the conversion coefficient, It includes information on technology changes, market dynamics and risk fluctuations. For the The stable weight of each job information, For the Rating of job information, for The maximum value in For the The second constant of job information; The job difference coefficient of each group of jobs is determined by the job stability level.

[0012] In some embodiments of the present application, a hyperbolic decay model for each group position is established, including: ; in, For the Group positions at time The weight under For the The initial weight of each group position, For the The job difference coefficient of each group of jobs, For the time interval.

[0013] In some embodiments of the present application, the capability profile of each enterprise personnel is constructed through the comprehensive quality level of enterprise personnel and the hyperbolic attenuation model of group positions, including: The capability profile of enterprise personnel includes all categories of reflection indicators, identity information, comprehensive quality level of enterprise personnel, hyperbolic attenuation model of group positions and capability scores; The ability score is calculated based on the hyperbolic attenuation model of the overall quality level of enterprise personnel and group positions.

[0014] Correspondingly, the present application also provides a talent management system based on objective data evaluation, including: The first module is used to conduct multi-dimensional and comprehensive digital monitoring of enterprise personnel and generate enterprise personnel activity logs. The enterprise personnel activity logs contain structured data and unstructured data, and reflect indicators are extracted from structured data and unstructured data; The second module is used to analyze the entropy changes of each type of reflection indicators and integrate all types of reflection indicators to obtain the comprehensive quality level of enterprise personnel; The third module is used to divide enterprise personnel into groups according to business positions, collect the position information of enterprise personnel in each group position, analyze the position information to determine the position difference coefficient of each group position, and establish a hyperbolic attenuation model for each group position; The fourth module is used to construct the capability profile of each enterprise employee through the comprehensive quality level of enterprise employees and the hyperbolic attenuation model of group positions, so as to achieve the management of enterprise talents.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. Extract reflection indicators from structured data and unstructured data, and process the two types of data reflecting business personnel in a targeted manner, so as to accurately extract reflection indicators and provide a reliable basis for subsequent talent evaluation and management. Analyze the entropy changes of each type of reflection indicator, and integrate all types of reflection indicators to determine the comprehensive quality level. The comprehensive quality level is used to describe the overall ability level of different personnel, improve the objectivity of talent evaluation, avoid the influence of subjective factors on talent ability evaluation, and help optimize talent management.

[0016] 2. Analyze the job information to determine the job difference coefficient of each group of jobs, establish a hyperbolic attenuation model for each group of jobs, consider the time changes of multiple factors in the job information, and determine the weight of each job information. The hyperbolic characteristic balances the historical value and timeliness, and describes the decay rate of the job's ability value. Construct a capability portrait of each enterprise employee, which can realize the ability comparison between different positions and optimize the enterprise's talent management and job allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flow chart of a talent management method based on objective data evaluation proposed by the present invention; Figure 2 This is a structural diagram of a talent management system based on objective data evaluation proposed by the present invention. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0019] Reference Figure 1 , a talent management method based on objective data evaluation, comprising the following steps, Step S101, conduct multi-dimensional and comprehensive digital monitoring of enterprise personnel, generate enterprise personnel activity logs, and the enterprise personnel activity logs contain structured data and unstructured data, and extract reflection indicators from the structured data and unstructured data.

[0020] In this embodiment, regular expressions, JSON Path and other methods are used to extract indicators reflecting the work ability or business ability of enterprise personnel from structured data, such as working hours, number of tasks completed, sales, etc. Regular expressions, XPath, CSS selectors and other methods are used to extract valuable information such as text and images from unstructured data. For example, keywords can be extracted from work reports and project summaries.

[0021] In some embodiments of the present application, reflection indicators are extracted from structured data and unstructured data, including: For structured data, determine the value information set in the structured data, expand the value information set through the relevance of the value information set, extract the expanded value information set in the structured data, and extract the reflection indicator from the expanded value information set; For unstructured data, valuable information sets are extracted from the unstructured data through multi-source technology. Multi-source technology includes one or more of text mining, natural language processing, and image and audio analysis to extract reflection indicators from the valuable information sets.

[0022] In this embodiment, structured data often has a fixed format, which can more obviously extract the reflection indicators and determine the initial value information set in the structured data, such as basic work performance data. The value information set is expanded by analyzing the correlation between the value information sets, such as the relationship between the working hours and the number of tasks completed. Unstructured data includes various forms such as text, images, and audio, such as work reports, project summaries, employee evaluations, etc. These data are usually stored in document management systems, social media, or internal company databases. Use regular expressions, XPath, CSS selectors, and other methods to extract text information from unstructured data. Regular expressions are also applicable to the matching and extraction of unstructured text data; XPath and CSS selectors are often used for parsing HTML or XML documents. For non-text data such as images and audio, image recognition and audio analysis technologies are needed to extract valuable information. Multi-source technologies include text mining, natural language processing (NLP), image and audio analysis, etc. These technologies can work together to extract deeper information from unstructured data. Text mining can help extract keywords, topics, and sentiment tendencies; NLP technology can understand and generate human language and further analyze text content; image and audio analysis can extract features and information from image and audio data. The information extracted from unstructured data through multi-source technology can be used to construct a valuable information set. Indicators that reflect the work ability or business ability of corporate personnel, such as innovation ability, communication ability, and teamwork ability, can be extracted from the valuable information set.

[0023] Step S102, analyzing the entropy change of each type of reflection index, and integrating all types of reflection indexes to obtain the comprehensive quality level of enterprise personnel.

[0024] In this embodiment, the entropy value of each indicator is calculated to measure its information content and uncertainty, and the weight of each reflecting indicator is determined according to the entropy value to obtain the comprehensive quality level of enterprise personnel.

[0025] In some embodiments of the present application, the entropy change of each type of reflection indicator is analyzed, including: Preset multiple typical time periods for each type of reflection indicator, determine the degree of dispersion of each type of reflection indicator, determine the number of lags in the autocorrelation function of each type of reflection indicator according to the degree of dispersion, thereby calculating the autocorrelation value of each type of reflection indicator, and draw an autocorrelation graph of the reflection indicator, and select a typical time period from multiple typical time periods as a standard time period of the reflection indicator according to the autocorrelation graph of the reflection indicator; The mean and change rate of information entropy are calculated according to the standard time period of each type of reflecting indicator, and the mean and change rate of information entropy reflecting the indicator are taken as the entropy change.

[0026] In this embodiment, each type of reflection indicator has a different time period, which has an impact on the calculation of the subsequent entropy value. Therefore, the time period that best reflects the characteristics and laws of each type of reflection indicator is first determined. Use statistical software or programming language (such as Python's statsmodels library) to calculate the autocorrelation function (ACF) under different time periods. The autocorrelation function can reveal the correlation of data at different lag times. The lag number can be set according to the data dispersion degree of the reflection indicator. The data dispersion degree refers to the degree of dispersion between each data value in a set of data, which reflects the degree of aggregation or dispersion of the data value. The greater the dispersion degree, the more dispersed the data value; the smaller the dispersion degree, the more aggregated the data value. The lag number refers to the number of time delays between the data and itself in the calculation of the autocorrelation function. It determines the time range for observing the autocorrelation of the data. The degree of data dispersion may affect the selection of the lag number. For example, when the data dispersion degree is large, a larger lag number may be required to capture the periodicity or long-term trend in the data. Because a larger degree of dispersion may mean that there are more fluctuations and changes in the data, a longer time delay is required to observe the correlation between these changes.

[0027] In some embodiments of the present application, a typical time period is selected from a plurality of typical time periods as a standard time period reflecting the indicator according to the autocorrelation diagram reflecting the indicator, including: Identify multiple significant peaks on the autocorrelation graph reflecting the index, determine the peak value, peak position and peak shape of each significant peak, the peak shape includes the peak width and the rate of change of the autocorrelation value near the peak; The peak width and the rate of change of the autocorrelation value near the peak are combined to determine the sharpness and smoothness of the peak. The peak description is determined by the sharpness and smoothness and the peak value of the peak. Multiple significant peaks are ranked according to the peak description. According to the peak position of the first-ranked significant peak, a similar typical time period is selected from multiple typical time periods as the standard time period reflecting the indicator.

[0028] In this embodiment, multiple significant peaks are peaks with relatively large peak sizes, and the peaks reflect the strength of the autocorrelation, and the peak position (time) directly indicates the length of the period. For example, if the autocorrelation graph has a significant peak at lag 7, this may mean that the data has a periodic pattern with a period of 7. The shape of the peak (such as sharp or gentle) may also provide information about the periodic pattern. The peak width can measure the width of the peak on the correlation coefficient axis. A narrower peak usually indicates a sharper shape, while a wider peak indicates a gentler shape. The rate at which the correlation coefficient changes with the lag value near the peak is calculated. A faster rate of change usually indicates a sharper peak, while a slower rate of change indicates a gentler peak. The sharpness and gentleness of the peak are determined by combining the peak width and the rate of change of the autocorrelation value near the peak. The greater the sharpness and gentleness, the gentler it is. The peak description is determined by combining the sharpness and gentleness and the peak (weighted average or summation). The greater the peak description, the stronger the autocorrelation and the more regular the data.

[0029] In some embodiments of the present application, all categories of reflection indicators are integrated to obtain the comprehensive quality level of enterprise personnel, including: Standardize all categories of reflection indicators and determine the comprehensive quality level of enterprise personnel based on the mean and change rate of information entropy of reflection indicators; ; in, For the comprehensive quality level of enterprise personnel, To reflect the number of indicator categories, For the The mean information entropy of the indicator, represents the sum of the normalized information entropy means of all indicators, Indicates The weight of the indicator is reflected. For the The size of the indicator is reflected. For the The rate of change of the information entropy of the indicator is For the The first constant reflecting the indicator.

[0030] In this embodiment, the weight of the reflection index is determined by the mean value of the information entropy. The rate of change of information entropy is The correction, In order to balance the size of the correction function.

[0031] Step S103, divide the enterprise personnel into groups according to business positions, collect the position information of the enterprise personnel in each group position, analyze the position information to determine the position difference coefficient of each group position, and establish a hyperbolic attenuation model for each group position.

[0032] In this embodiment, the relative value of the positions of enterprise personnel in different positions has different degrees of attenuation, matching the attenuation speed of the ability of different positions (position difference coefficient). The greater the value, the faster the attenuation rate of the ability value. For example, the position difference coefficient of the technical position = 0.1: reflecting the long half-life of technical knowledge (weight is halved in about 10 months) and the position difference coefficient of the marketing position = 0.3: reflecting the rapid iteration characteristics of market dynamics (weight is halved in about 3.3 months).

[0033] In some embodiments of the present application, analyzing the position information to determine the position difference coefficient of each group position includes: Position information includes information on technology changes, market dynamics, and risk fluctuations; Analyze technology change information, market dynamic change information and risk fluctuation information respectively, and score technology change information, market dynamic change information and risk fluctuation information to define job stability level; ; in, For job stability level, To stabilize the conversion coefficient, It includes information on technology changes, market dynamics and risk fluctuations. For the The stable weight of each job information, For the Rating of job information, for The maximum value in For the The second constant of job information; The job difference coefficient of each group of jobs is determined by the job stability level.

[0034] In this embodiment, different job stability levels correspond to different job difference coefficients, technology replacement information, market dynamic change information and risk fluctuation information (technical risk, market risk, quality risk, etc.). For example, technology trends: by studying the industry's scientific research results, technical papers, patent information, science and technology media and industry associations, understand the latest developments and future trends of current technologies. Technological impact: Assess the impact of new technologies on job content, skill requirements, workflows, etc. The hyperbolic characteristic balances historical value and timeliness, and is more in line with the long-term growth law of technical talents than the exponential model. In actual applications, the job difference coefficient needs to be dynamically adjusted in conjunction with the job analysis model.

[0035] In this embodiment, It represents information on technology changes, market dynamics, and risk fluctuations. The maximum value of the correction to the sum of the three, This is to balance the size of the correction function.

[0036] In some embodiments of the present application, a hyperbolic decay model for each group position is established, including: ; in, For the Group positions at time The weight under For the The initial weight of each group position, For the The job difference coefficient of each group of jobs, For the time interval.

[0037] In this embodiment, this It is a weight and also a time decay factor, which can help reduce the impact of historical data on the judgment in time and can also help with subsequent predictions.

[0038] Step S104, constructing a capability profile of each enterprise employee through the comprehensive quality level of enterprise employees and the hyperbolic attenuation model of group positions, so as to achieve the management of enterprise talents.

[0039] In some embodiments of the present application, the capability profile of each enterprise personnel is constructed through the comprehensive quality level of enterprise personnel and the hyperbolic attenuation model of group positions, including: The capability profile of enterprise personnel includes all categories of reflection indicators, identity information, comprehensive quality level of enterprise personnel, hyperbolic attenuation model of group positions and capability scores; The ability score is calculated based on the hyperbolic attenuation model of the overall quality level of enterprise personnel and group positions.

[0040] In this embodiment, the product of the comprehensive quality level of enterprise personnel and the hyperbolic attenuation model of group positions is the ability score, which can achieve relative comparison of different people in different positions and facilitate subsequent talent management and standard formulation.

[0041] Correspondingly, this application also provides a talent management system based on objective data evaluation, such as Figure 2 As shown, including, The first module is used to conduct multi-dimensional and comprehensive digital monitoring of enterprise personnel and generate enterprise personnel activity logs. The enterprise personnel activity logs contain structured data and unstructured data, and reflect indicators are extracted from structured data and unstructured data; The second module is used to analyze the entropy changes of each type of reflection indicators and integrate all types of reflection indicators to obtain the comprehensive quality level of enterprise personnel; The third module is used to divide enterprise personnel into groups according to business positions, collect the position information of enterprise personnel in each group position, analyze the position information to determine the position difference coefficient of each group position, and establish a hyperbolic attenuation model for each group position; The fourth module is used to construct the capability profile of each enterprise employee through the comprehensive quality level of enterprise employees and the hyperbolic attenuation model of group positions, so as to achieve the management of enterprise talents.

[0042] Compared with the prior art, the present invention has the following beneficial effects: 1. Extract reflection indicators from structured data and unstructured data, and process the two types of data reflecting business personnel in a targeted manner, so as to accurately extract reflection indicators and provide a reliable basis for subsequent talent evaluation and management. Analyze the entropy changes of each type of reflection indicator, and integrate all types of reflection indicators to determine the comprehensive quality level. The comprehensive quality level is used to describe the overall ability level of different personnel, improve the objectivity of talent evaluation, avoid the influence of subjective factors on talent ability evaluation, and help optimize talent management.

[0043] 2. Analyze the job information to determine the job difference coefficient of each group of jobs, establish a hyperbolic attenuation model for each group of jobs, consider the time changes of multiple factors in the job information, and determine the weight of each job information. The hyperbolic characteristic balances the historical value and timeliness, and describes the decay rate of the job's ability value. Construct a capability portrait of each enterprise employee, which can realize the ability comparison between different positions and optimize the enterprise's talent management and job allocation.

[0044] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.

[0045] Those skilled in the art will appreciate that the accompanying drawings are merely schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily required for implementing the present invention.

[0046] Those skilled in the art will appreciate that the modules in the system in the implementation scenario can be distributed in the system of the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more systems different from the implementation scenario. The modules in the above implementation scenario can be combined into one module, or can be further split into multiple submodules.

[0047] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A talent management method based on objective data evaluation, characterized in that: include, Conduct multi-dimensional and comprehensive digital monitoring of enterprise personnel and generate enterprise personnel activity logs, which contain structured and unstructured data. Indicators are extracted from structured and unstructured data. Analyze the entropy changes of each type of reflection index, and integrate all types of reflection indexes to obtain the comprehensive quality level of enterprise personnel; Divide the enterprise personnel into groups according to business positions, collect the position information of the enterprise personnel in each group position, analyze the position information to determine the position difference coefficient of each group position, and establish a hyperbolic attenuation model for each group position; Through the comprehensive quality level of enterprise personnel and the hyperbolic attenuation model of group positions, we construct the ability profile of each enterprise personnel, so as to achieve the management of enterprise talents.

2. The talent management method based on objective data evaluation according to claim 1, characterized in that: Extract reflection indicators for structured and unstructured data, including: For structured data, determine the value information set in the structured data, expand the value information set through the relevance of the value information set, extract the expanded value information set in the structured data, and extract the reflection indicator from the expanded value information set; For unstructured data, valuable information sets are extracted from the unstructured data through multi-source technology. Multi-source technology includes one or more of text mining, natural language processing, and image and audio analysis to extract reflection indicators from the valuable information sets.

3. The talent management method based on objective data evaluation according to claim 1, characterized in that: Analyze the entropy changes of each type of reflection indicator, including: Preset multiple typical time periods for each type of reflection indicator, determine the degree of dispersion of each type of reflection indicator, determine the number of lags in the autocorrelation function of each type of reflection indicator according to the degree of dispersion, thereby calculating the autocorrelation value of each type of reflection indicator, and draw an autocorrelation graph of the reflection indicator, and select a typical time period from multiple typical time periods as a standard time period of the reflection indicator according to the autocorrelation graph of the reflection indicator; The mean and change rate of information entropy are calculated according to the standard time period of each type of reflecting indicator, and the mean and change rate of information entropy reflecting the indicator are taken as the entropy change.

4. The talent management method based on objective data evaluation according to claim 3, characterized in that: According to the autocorrelation diagram of the reflecting indicator, a typical time period is selected from multiple typical time periods as the standard time period reflecting the indicator, including: Identify multiple significant peaks on the autocorrelation graph reflecting the index, determine the peak value, peak position and peak shape of each significant peak, the peak shape includes the peak width and the rate of change of the autocorrelation value near the peak; The peak width and the rate of change of the autocorrelation value near the peak are combined to determine the sharpness and smoothness of the peak. The peak description is determined by the sharpness and smoothness and the peak value of the peak. Multiple significant peaks are ranked according to the peak description. According to the peak position of the first-ranked significant peak, a similar typical time period is selected from multiple typical time periods as the standard time period reflecting the indicator.

5. The talent management method based on objective data evaluation according to claim 3, characterized in that: And integrate all categories of reflection indicators to obtain the comprehensive quality level of enterprise personnel, including: Standardize all categories of reflection indicators and determine the comprehensive quality level of enterprise personnel based on the mean and change rate of information entropy of reflection indicators; ； in, For the comprehensive quality level of enterprise personnel, To reflect the number of indicator categories, For the The mean information entropy of the reflecting index, represents the sum of the normalized information entropy means of all indicators, Indicates The weight of the indicator is reflected. For the The size of the indicator is reflected. For the The rate of change of the information entropy of the indicator is For the The first constant reflecting the indicator.

6. The talent management method based on objective data evaluation according to claim 1, characterized in that: Analyze job information to determine the job difference coefficient for each group of jobs, including: Position information includes information on technology changes, market dynamics, and risk fluctuations; Analyze the information on technology changes, market dynamics changes and risk fluctuations respectively, and score them to define the job stability level; ； in, For job stability level, To stabilize the conversion coefficient, It includes information on technology changes, market dynamics and risk fluctuations. For the The stable weight of each job information, For the Rating of job information, for The maximum value in For the The second constant of job information; The job difference coefficient of each group of jobs is determined by the job stability level.

7. The talent management method based on objective data evaluation according to claim 1, characterized in that: Build a hyperbolic decay model for each group position, including, ； in, For the Group positions at time The weight under For the The initial weight of each group position, For the The job difference coefficient of each group of jobs, is the time interval.

8. The talent management method based on objective data evaluation according to claim 3, characterized in that: Through the comprehensive quality level of enterprise personnel and the hyperbolic attenuation model of group positions, we can build the ability portrait of each enterprise personnel. include, The capability profile of enterprise personnel includes all categories of reflection indicators, identity information, comprehensive quality level of enterprise personnel, hyperbolic attenuation model of group positions and capability scores; The ability score is calculated based on the hyperbolic attenuation model of the overall quality level of enterprise personnel and group positions.

9. A talent management system based on objective data evaluation, characterized in that: include, The first module is used to conduct multi-dimensional and comprehensive digital monitoring of enterprise personnel and generate enterprise personnel activity logs. The enterprise personnel activity logs contain structured data and unstructured data, and reflect indicators are extracted from structured data and unstructured data; The second module is used to analyze the entropy changes of each type of reflection indicators and integrate all types of reflection indicators to obtain the comprehensive quality level of enterprise personnel; The third module is used to divide enterprise personnel into groups according to business positions, collect the position information of enterprise personnel in each group position, analyze the position information to determine the position difference coefficient of each group position, and establish a hyperbolic attenuation model for each group position; The fourth module is used to construct the capability profile of each enterprise employee through the comprehensive quality level of enterprise employees and the hyperbolic attenuation model of group positions, so as to achieve the management of enterprise talents.

Images