Search

CN-121981587-A - Talent ability and potential evaluation method and system based on nonlinear weighting model

CN121981587ACN 121981587 ACN121981587 ACN 121981587ACN-121981587-A

Abstract

The invention discloses a talent ability and potential evaluation method and a system based on a nonlinear weighting model, wherein the method comprises the following steps: and determining the human ability and potential core evaluation dimension, grading according to the importance of fruits, and collecting corresponding performance data and supporting evidence information to form a standardized evaluation data set. Based on the data set, a piecewise nonlinear weighted regression model is adopted, and the original score of each dimension is calculated in two sections according to the range of 20 years from the current year. And normalizing the original score, combining a dimension weight optimization mapping logic, unifying the original score to a preset score interval, and dividing the competitive level to form a multi-dimensional grading portrait. And collecting continuous period evaluation data, constructing a capacity increment trend model by combining support main support force, generating culture potential sequences through transverse and longitudinal analysis, and outputting a visual chart, a good and bad potential report and culture suggestions. The accurate assessment of talent ability and the quantitative prediction of potential are realized.

Inventors

  • LI WENJIE
  • SONG YUANYIN

Assignees

  • 中国交通建设集团有限公司
  • 中交公路长大桥建设国家工程研究中心有限公司

Dates

Publication Date
20260505
Application Date
20251208

Claims (10)

  1. 1. A method for evaluating talent ability and potential based on a nonlinear weighting model, comprising: S1, determining core evaluation dimensions of talent ability and potential, dividing each dimension into multiple levels according to fruit importance levels, synchronously collecting performance data corresponding to each dimension and related support evidence information provided by related support main bodies, and forming a standardized evaluation data set; S2, calculating the original score of each dimension by adopting a piecewise nonlinear weighted regression model based on the standardized evaluation data set, dividing two sections of calculation logic according to the current year of the result, wherein one section is within 20 years of the current year, carrying out weighted calculation by combining the basic score of the result, the secondary weight and the time effect weight decreasing with time, and the other section is 20 years or more of the current year, and finally obtaining the original score result of each dimension by adopting a weighted calculation mode of adapting to the long-term result value; s3, carrying out standardized treatment on original scores of all dimensions by adopting a normalization method, and uniformly mapping the scores to a preset score interval by combining weight proportion optimization score mapping logic of all dimensions in a talent assessment system; S4, collecting multiple rounds of evaluation data in a talent continuous period, constructing a capacity increment trend model by combining the support force indexes of the related support main body, generating a talent culture potential sequencing result through transverse comparison and longitudinal analysis, and outputting a talent competitiveness visualization chart, a priority analysis report and a culture suggestion.
  2. 2. The method for evaluating the talent ability and the potential based on the nonlinear weighted model according to claim 1, wherein the core evaluation dimension in S1 at least comprises five dimensions including academic achievements, academic reputation, leading edge cooperation, scientific rewards and important honor.
  3. 3. The method of claim 1, wherein the performance data in S1 includes a ranking of achievements, a ranking of earnings, and a number of years from this.
  4. 4. The method for evaluating the talent ability and potential based on the nonlinear weighted model according to claim 1, wherein the construction process of the standardized evaluation data set in S1 is specifically as follows: Let the evaluation dimension set be , wherein, Corresponding to the academic results, the method has the advantages of, Corresponding to the academic reputation, the method comprises the steps of, The corresponding front-edge co-operation is performed, Corresponding to the scientific and technological rewards, Corresponding important honor, and the importance level set of the achievements of each dimension is , Assigning decreasing basic weight from high to low according to the influence of fruits, defining single-dimension data tuples as , wherein, The corresponding dimension number of the assessment is, Is the first one in the dimension Item performance sequence number; for the importance level weight of the performance, Rank the coefficients for the item's performance, For this performance to be several years apart, Screening effective data for support evidence intensity coefficients of the relevant support subjects for the performance by the following data cleaning rules: , Wherein, the To support the lowest threshold of evidence intensity, null indicates deletion of invalid data; represent the first Dimension down The number of years from the performance of the item No less than 0, i.e., the performance is generated, non-future data; represent the first Dimension down Importance level weighting of item performance Belonging to a preset class set I.e. the performance level meets the classification standard of the assessment system; represent the first Dimension down Supporting main body supporting evidence intensity coefficient corresponding to item performance Not lower than a preset minimum threshold I.e. the support material that performs meets the evaluation requirements; Final construction of normalized evaluation dataset Wherein, support the evidence of intensity coefficient The method is obtained by quantitatively converting the resource inclination degree, the project support degree and the organization management efficiency of the support main body.
  5. 5. The method for evaluating the talent ability and the potential based on the nonlinear weighted model according to claim 4, wherein the evaluating method for the importance level in S1 is as follows: let the industry application coverage quantization value corresponding to a certain performance in a single dimension be The value of the method is the ratio of the number of main bodies of the application of the performance technical scheme in the industry to the total number of main bodies in the industry, and the range The quantification value of the public evaluation frequency of the same row is The value of the method is the ratio of the concordance, the positive evaluation times and the average times of the concordance results in the same field, and the range Then construct the impact comprehensive value of the performance : , Meanwhile, based on the project level corresponding to the performance, a project level correction factor is defined Finally obtaining the importance level associated value of the performance : , Will be Dividing the intervals in descending order, and mapping to the result importance level set And finishing the differentiated division of the importance level.
  6. 6. The method for evaluating the talent ability and potential based on the nonlinear weighted model according to claim 1, wherein the nonlinear weighted regression model in S2 is specifically: For the first in single dimension Item performance in terms of a result base score Rank sub-weight The achievements are several years apart For input variables, constructing a piecewise nonlinear weighted regression model, and obtaining a single-dimension total score Calculated by the following formula: , Wherein, the Aggregate effective performance in that dimension; by logarithmically transforming the differentiated weights of rank ranks, And (3) with The time attenuation characteristic of fruits within 20 years and the value precipitation characteristic of fruits of 20 years and more are respectively adapted to the time effect function of time intervals.
  7. 7. The talent ability and potential assessment method based on a nonlinear weighting model according to claim 1, wherein the specific process of combining the weight ratio optimization score mapping logic of each dimension in the talent assessment system in S3 is as follows: set the preset weight set of each evaluation dimension as First, the The original score set of all evaluation objects under the dimension is as follows , Simultaneously introducing dimension discrimination coefficient for evaluating the number of objects in the dimension The calculation formula is as follows: , Wherein, the Is the first Standard deviation of scores of all evaluation objects in the dimension, Is the first Average value of scores of all evaluation objects in dimension, and the higher the degree of distinction is The larger; firstly, carrying out nonlinear normalization in the dimension on the original scores of all dimensions, compressing fluctuation of the extreme scores through logarithmic transformation, and improving the distinguishing degree of the intermediate scores: , Wherein, the Is the first The nonlinear normalized score of the dimension is calculated, Denoted as the first Dimension down Original scores of individual assessment objects; And dynamically adjusting the dimension calibration weight by combining the dimension weight and the discrimination coefficient: , Wherein, the Is the first Dynamic calibration weight of dimension, denominator The sum of the preset weight and the distinguishing degree coefficient is used for all dimensions; the higher the discrimination degree dimension, the higher the actual calibration weight ratio, and the enhanced evaluation contribution of the high discrimination degree dimension: , Wherein, the And (3) finishing double optimization score mapping of 'discrimination dynamic weighting + nonlinear normalization' for adapting to the nonlinear mapping function of the preset score interval.
  8. 8. The method for evaluating the human ability and potential based on the nonlinear weighted model according to claim 1, wherein the ability increment trend model in S4 is specifically: Let talent continuous evaluation period set be , wherein, , Represent the first The round evaluation period corresponds to the multi-dimensional calibration score matrix of each period as : , Wherein, the Is the first Cycle number Dimension calibration scores; The support main body support force quantization index is The value range [0,1] is obtained by comprehensively quantifying the resource input intensity, the policy support strength and the platform support level; Building a capability increment core function: , Wherein, the Is the first In the period of The capacity increment value of the dimension, To support the force attenuation factor, the nonlinear energization of the support force to the capacity increment is completed; building a trend fitting model based on the increment value of each dimension: , Wherein, the Is the first Dimension down The ability potential of the cycle predicts the score, Is 1 st to 1 st Within the period of The capability increment of the dimension is accumulated into a value, The real-time effect coefficient, the accumulated effect coefficient and the support force adaptation coefficient of the dimension increment are obtained through least square fitting; by transverse comparison of talents of the same batch Matrix similarity and longitudinal analysis of single person And integrating the incremental potential, outputting talent culture potential sequencing results, and providing a quantification basis for culture suggestions.
  9. 9. A human ability and potential assessment system based on a nonlinear weighted model, comprising: The index system and data set construction unit is used for determining core evaluation dimensions of talent ability and potential, dividing each dimension into multiple levels according to fruit importance levels, synchronously collecting performance data corresponding to each dimension and related support evidence information provided by related support main bodies, and forming a standardized evaluation data set; Dividing two sections of calculation logic according to the number of years of the result, wherein one section is within 20 years of the current year, weighting calculation is carried out by combining the basic result value, the rank weight and the time effect weight which is reduced along with time, and the other section is 20 years or more from the current year, and finally obtaining the original score result of each dimension by adopting a weighting calculation mode which is adaptive to the long-term result value; The normalization and grading portrait generation unit is used for carrying out normalization processing on the original scores of all the dimensions by adopting a normalization method, and combining weight proportion optimization score mapping logic of all the dimensions in a talent evaluation system to uniformly map the scores to a preset score interval; The multi-period potential evaluation output unit is used for collecting multi-round evaluation data in a talent continuous period, constructing a capacity increment trend model by combining support main body support force indexes, generating a talent culture potential sequencing result through transverse comparison and longitudinal analysis, and outputting a talent competitiveness visualization chart, a priority and inferiority analysis report and a culture suggestion.
  10. 10. A computer readable storage medium having stored thereon a computer program, characterized in that the computer program is executed by a processor for a method of human ability and potential assessment based on a nonlinear weighted model according to any one of claims 1-8.

Description

Talent ability and potential evaluation method and system based on nonlinear weighting model Technical Field The invention belongs to the technical field of talent resource assessment, and particularly relates to a talent capability and potential assessment method and system based on a nonlinear weighting model. Background In the current talent assessment field, the traditional method mostly adopts a linear weighting or single-dimension scoring mode, and is difficult to adapt to the multiple value attributes of talent performance. In most evaluation processes, the data screening lacks unified standards, and performance data of different types and different timelines are often directly combined and calculated, so that the influence of core factors such as importance of achievements, contribution rank, time attenuation and the like is weakened, and the actual value ratio of talents in performance cannot be accurately distinguished. Meanwhile, the traditional evaluation of static scoring of multi-focus current capability lacks of differentiation processing on time characteristics of performance, the value weight distribution of recent results and long-term results is unreasonable, and the situation that the evaluation result is disjointed with the actual contribution of talents easily occurs. In addition, the weight of each dimension in the existing evaluation system is mostly fixed, the distinguishing capability of the uncombined dimension is dynamically adjusted, the evaluation value of part of the high-distinguishing dimension is covered, and the redundancy duty ratio of the low-distinguishing dimension may interfere with the result. Meanwhile, the potential prediction link often depends on subjective experience judgment, and the quantitative integration of continuous periodic capacity change and external support resources is lacking, so that the growth situation of talent capacity is difficult to objectively reflect, and the evaluation result only can reflect the current state and cannot provide prospective decision basis for talent cultivation. These problems make traditional talent assessment deficient in accuracy, objectivity and foresight, which is not beneficial to accurately identifying the actual capability level of talents, and is also difficult to support scientific talent echelon construction and culture planning. Under the background, a set of methods which can adapt to multiple performance attributes, dynamically adjust evaluation weights and give consideration to capability evaluation and potential prediction is constructed, so that the method becomes a key requirement for optimizing talent evaluation systems and improving talent resource allocation efficiency, and has important practical significance for promoting scientization and precision of talent evaluation. Disclosure of Invention The method aims to solve the problems of non-uniform data standard, linear weighting misalignment, dimensional weight solidification and subjective potential prediction in the traditional talent assessment, and realizes accurate assessment of talent capability and quantitative prediction of potential by constructing a standardized data screening, piecewise nonlinear weighting model, dynamic dimensional weight adjustment and capability increment trend model, so as to support scientific talent echelon construction. In response to the above-mentioned shortcomings or improvements of the prior art, the present invention provides a method for human ability and potential assessment based on a nonlinear weighted model, comprising: S1, determining core evaluation dimensions of talent ability and potential, dividing each dimension into multiple levels according to fruit importance levels, synchronously collecting performance data corresponding to each dimension and related support evidence information provided by a support main body, and forming a standardized evaluation data set; S2, calculating the original score of each dimension by adopting a piecewise nonlinear weighted regression model based on the standardized evaluation data set, dividing two sections of calculation logic according to the current year of the result, wherein one section is within 20 years of the current year, carrying out weighted calculation by combining the basic score of the result, the secondary weight and the time effect weight decreasing with time, and the other section is 20 years or more of the current year, and finally obtaining the original score result of each dimension by adopting a weighted calculation mode of adapting to the long-term result value; s3, carrying out standardized treatment on original scores of all dimensions by adopting a normalization method, and uniformly mapping the scores to a preset score interval by combining weight proportion optimization score mapping logic of all dimensions in a talent assessment system; S4, collecting multiple rounds of evaluation data in a talent continuous period, constructing a capacity