Urban talent development evaluation method and device and storage medium

Abstract

The invention relates to an urban talent development evaluation method and device and a storage medium, and the method comprises the steps: collecting and obtaining the data of three-level indexes in each index system through calling a constructed urban talent development evaluation index system; according to the data distribution of the third-level indexes in each index system, performing standardization processing on the data to obtain standardized numerical values of the third-level indexes in each index system; determining weights of three-level indexes in each index system by adopting a factor analysis method; and determining a city talent development evaluation score according to the standardized numerical values and weights of the three-level indexes in each index system. According to the invention, by constructing the urban talent development evaluation system, multi-level, multi-dimensional, quantitative and systematic comprehensive evaluation of each innovation subject and innovation development environment of the city is realized. The factor analysis method is adopted to determine the weight of each level of index, the analysis and calculation process is simplified, the interference of subjective judgment of experts is reduced, and the evaluation result is more objective and accurate.

Description

Technical Field

[0001] This application belongs to the field of urban development technology, and specifically relates to an urban talent development evaluation method, equipment and storage medium. Background Technology

[0003] Existing talent evaluation systems employ methods including: single-indicator methods, such as using the number of published papers as a standard for evaluating talent, and using talent scale as a standard for evaluating a city's talent competitiveness; constructing evaluation indicators based on the Delphi method to determine indicator weights, which is a method that reaches consensus through multiple rounds of expert opinion exchange and feedback, thereby determining indicator weights; and constructing evaluation indicators based on the entropy method to determine the weight of each indicator, which determines the weight of each indicator based on the amount of information provided by each indicator.

[0004] However, existing technical methods have the following drawbacks:

[0005] Single indicator method: Using only a single indicator to evaluate the quality of talent or the development of talent in a city lacks comprehensiveness and the evaluation results are too one-sided; the evaluation results obtained by using different indicators are not comparable and are easily manipulated; it lacks flexibility, as different fields have different requirements for talent or cities, making it impossible to make cross-comparisons of the evaluation results.

[0006] Delphi method: The Delphi method usually requires multiple rounds of anonymous expert surveys and feedback, which is time-consuming; the selection of experts and the determination of their weights are subject to subjective influences; the results of the Delphi method are based on expert consensus and have poor interpretability for non-expert groups.

[0007] Entropy method: It relies entirely on data, which leads to the weighting results not matching the actual importance of the indicators; it assumes that the indicators are independent of each other and ignores the correlation between the indicators, resulting in unreasonable weight allocation.

[0008] Therefore, how to conduct a comprehensive evaluation of urban talent development in a multi-dimensional, quantitative, and systematic way is a topic worthy of study in the current urban planning and development. Summary of the Invention

[0009] Based on the above analysis, the embodiments of the present invention aim to provide a method for evaluating urban talent development, which aims to solve the shortcomings of existing talent evaluation methods that adopt a single dimension, are highly subjective, and have poor evaluation results.

[0010] The first aspect of this application provides a method for evaluating urban talent development, including:

[0011] The constructed urban talent development evaluation index system is invoked. The urban talent development evaluation index system includes a talent quality index system, an institutional platform index system, and an urban development index system. Each level of the index system includes primary indicators, secondary indicators, and tertiary indicators. Each primary indicator measures the comprehensive score of talent, institution, or city. Each secondary indicator measures the influence of the evaluated subject at that level in different dimensions. Each tertiary indicator is a quantifiable indicator corresponding to the secondary indicator.

[0012] Collect and acquire data on the tertiary indicators in each indicator system;

[0013] Based on the data distribution of the tertiary indicators in each indicator system, the data is standardized to obtain the standardized values ​​of the tertiary indicators in each indicator system.

[0014] Factor analysis was used to determine the weights of the tertiary indicators in each indicator system. In the factor analysis, variables that are highly correlated with specific factors were used to calculate factor scores.

[0015] Based on the standardized values ​​and weights of the three-level indicators in each indicator system, the city's talent development evaluation score is determined:

[0016] Where π represents the city's talent development evaluation score, γ f M represents the weight of the third-level indicators in each indicator system. f This represents the standardized value of the third-level indicators in each indicator system.

[0017] Optionally, the determination of the weights of the tertiary indicators in each indicator system using factor analysis includes:

[0018] Establish factor analysis models for three-level indicators of each city and each level of the system:

[0019]

[0020] Among them, Y i X represents the i-th third-level indicator in each hierarchical system. j ε represents the factor whose j-th eigenvalue is greater than 1. i Let α represent the error vector. ij This represents the loading of the i-th variable on the j-th factor;

[0021] The factor analysis model is represented in matrix form as: Y = AX + ε

[0022] in, A three-level indicator vector, For the factor loading matrix, For factor vectors, Error vector;

[0023] Common factors were extracted using the principal component factorization method, and the correlation matrix was solved to calculate the eigenvalues ​​and contribution rates.

[0024] in, The vector is formed by arranging the extracted principal components in descending order of their eigenvalues. Let λ be the eigenvalue of the correlation matrix composed of the original index Y. i (λ1…λ i The components of the eigenvector corresponding to >0);

[0025] The number of common factors w(λ) is determined by selecting the eigenvalues ​​of the correlation coefficient matrix. w ≥1), classify and name the variables according to the factor loadings on each variable, that is:

[0026] Y1…Y i ∈M1

[0027] Where M1 represents factor 1, Y1…Y i The variable with the highest factor loading coefficient on factor 1;

[0028] Calculate the variance contribution rate and cumulative variance explained by the principal components. The variance contribution rate of the w-th principal component is: The cumulative variance explained is:

[0029] The variables are rotated using the maximum variance orthogonal rotation method to obtain the interpretable common factors and the rotated factor loading coefficient matrix A′=A*T, where T is an orthogonal matrix;

[0030] Based on the rotated factor loading coefficient matrix A′, the index weights and factor scores are calculated using regression estimation:

[0031]

[0032] in, The factor score vector obtained through regression estimation. This is a three-level indicator weight matrix. This is a vector of third-level indicators belonging to the corresponding factor category;

[0033] Calculate the weight of each factor based on the variance contribution rate and cumulative variance contribution rate of each common factor after rotation:

[0034] Optionally, it also includes: determining the scores of each city in various dimensions such as talent attraction, development platform, talent cultivation, career advancement, living security and rational allocation based on the standardized values ​​and weights of the tertiary indicators in each indicator system;

[0035] The scores of each city in the dimensions of talent attraction, development platform, talent cultivation, career advancement, living security, and rational allocation are calculated using the following formula:

[0036]

[0037] Where, π d γ represents the city's score in dimension d. f M represents the weight of the factor corresponding to dimension d. f β represents the score of the factor corresponding to dimension d; id Y represents the weights of the three-level indicators included in dimension d. id These are the standardized values ​​of the three-level indicators included in dimension d.

[0038] Optionally, the collection and acquisition of data for the third-level indicators in each indicator system includes:

[0039] Web crawlers based on natural language processing technology and large text analysis models extract structured data from web pages to obtain data on three levels of indicators.

[0040] Optionally, a web crawler based on natural language processing technology and a text segmentation model performs structured extraction of web page data to obtain data for the three-level indicators, including:

[0041] Natural language processing technology is used to identify abnormal data in web page data and to process the abnormal data.

[0042] Convert data from different sources and formats into a standard format suitable for analysis and computation;

[0043] By leveraging the contextual understanding capabilities of large deep learning models, key features are extracted from the corresponding data of each city, and data augmentation is performed.

[0044] Optionally, the step of standardizing the data based on the data distribution of the tertiary indicators in each indicator system to obtain the standardized values ​​of the tertiary indicators in each indicator system includes:

[0045] The data were standardized using the Max-Min dimensionless method to obtain the standardized values ​​of the third-level indicators in each indicator system:

[0046] Where Z represents the standardized value of each tertiary indicator; x is the initial value of the corresponding tertiary indicator; min and max correspond to the lower and upper limits of the indicator distribution range, respectively.

[0047] Optionally, in the talent quality indicator system, the secondary indicators include the talent achievement index, talent discipline index, talent interdisciplinary index, talent social index, talent cooperation index, and talent international index;

[0048] The institutional platform indicator system includes secondary indicators such as institutional platform talent scale index, institutional platform scientific research index, institutional platform social index, institutional platform cooperation index, and institutional platform international index.

[0049] The urban development indicator system includes secondary indicators such as the urban talent attraction index, urban development platform index, urban talent training index, urban career advancement index, urban livelihood security index, and urban rational allocation index.

[0050] Optionally, the collection and acquisition of data for the third-level indicators in each indicator system includes:

[0051] In the talent quality indicator system, data are collected on H-index, activity level, number of patents, percentage of ESI highly cited papers, patent valuation, average percentile, percentage of citations by top researchers, percentage of first authors, collaboration efficiency index, percentage of industry-university-research collaboration, impact of standardized citations in journals, disciplinary diversity, impact of standardized citations in disciplines, number of authoritative awards in the field, web search volume, high-level talent index, and percentage of international collaboration.

[0052] The platform's indicator system collects data on the scale of academic talent, the scale of commercial talent, the number of scientific and technological publications, the number of researchers, the number of patents, the academic influence of the discipline relative to the global level, the percentage of first authors, the collaboration index, the percentage of industry-university-research collaboration, the number of award winners, the number of awards, the percentage of international collaboration, and the number of countries with which the institution collaborates.

[0053] The urban development indicator system collects data on the following metrics per 10,000 employed persons: number of R&D personnel, growth rate of R&D personnel, number of high-level talents, number of outstanding young talents, number of reserve talents, number of scientific and technological publications, percentage of ESI highly cited papers, patent valuation, number of research positions in research institutions, number of Fortune Global 500 companies, number of China's top 500 private enterprises, number of national key laboratories, per capita R&D expenditure, proportion of basic research expenditure, number of national-level technology transfer platforms, number of patents per 10,000 people, time for talent promotion, per capita salary, GDP growth rate, number of beds per 10,000 people, per capita commuting time, per capita residential area, population density, per capita park green space area, annual air quality index good / excellent rate, per capita number of cultural activity venues, student-teacher ratio in primary and secondary schools; student-teacher ratio in higher education institutions, average years of education for the resident population, proportion of education expenditure to GDP, talent retention rate, talent attraction index, and proportion of floating population.

[0054] A second aspect of this application provides an urban talent development evaluation device, including a memory and a processor, wherein the memory stores a computer program, and the computer program, when executed by the processor, implements the urban talent development evaluation method according to any of the above.

[0055] A third aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the urban talent development evaluation method according to any of the above.

[0056] The urban talent development evaluation method provided in this application collects and obtains data on the tertiary indicators of each indicator system by calling upon a constructed urban talent development evaluation indicator system. Based on the data distribution of the tertiary indicators in each indicator system, the data is standardized to obtain standardized values ​​for the tertiary indicators. Factor analysis is used to determine the weights of the tertiary indicators in each indicator system. Based on the standardized values ​​and weights of the tertiary indicators in each indicator system, the urban talent development evaluation score is determined. This application constructs an urban talent development evaluation system encompassing talent quality, institutional platforms, and urban development, achieving a multi-level, multi-dimensional, quantitative, and systematic comprehensive evaluation of various innovation entities and the innovation development environment in a city. Furthermore, by using factor analysis to determine the weights of indicators at each level, and by determining common factors and indicator weights based on the variability and correlation of the data itself, the potential structure and relationships between indicators are revealed. Dimensionality reduction of the indicators simplifies the analysis and calculation process while retaining important information, reduces the interference of expert subjective judgment, and makes the evaluation results more objective and accurate.

[0057] In addition, this application also provides urban talent development evaluation equipment and storage medium with the above-mentioned technical effects. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings.

[0059] Figure 1 A flowchart illustrating a specific implementation method of the urban talent development evaluation method provided in this application;

[0060] Figure 2 A schematic diagram illustrating the process of collecting and acquiring data for the tertiary indicators in various indicator systems;

[0061] Figure 3 The structural diagram of the urban talent development evaluation equipment provided in this application. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. It should be noted that, unless otherwise specified, the implementation methods and features in the implementation methods in this disclosure can be combined, separated, interchanged, and / or rearranged. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0063] The terminology used herein is for the purpose of describing particular embodiments and is not intended to be limiting. As used herein, unless the context clearly indicates otherwise, the singular forms “a” and “the” are intended to include the plural forms as well. Furthermore, when the terms “comprising” and / or “including” and variations thereof are used in this specification, it indicates the presence of the stated features, integrals, steps, operations, parts, components, and / or groups thereof, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, parts, components, and / or groups thereof. It should also be noted that, as used herein, the terms “substantially,” “about,” and other similar terms are used as approximate terms rather than as terms of degree, thus explaining the inherent biases in measurements, calculated values, and / or provided values ​​that would be recognized by one of ordinary skill in the art.

[0064] A flowchart of a specific implementation method of the urban talent development evaluation method provided in this application is shown below. Figure 1 As shown, the method specifically includes:

[0065] S101: Invoke the constructed urban talent development evaluation index system.

[0066] The city talent development evaluation index system includes a talent quality index system, an institutional platform index system, and a city development index system. Each index system includes primary, secondary, and tertiary indicators. Each primary indicator measures the comprehensive score of talent, institution, or city. Each secondary indicator measures the influence of the evaluated entity in different dimensions. Each tertiary indicator is a quantifiable indicator corresponding to the secondary indicator.

[0067] The talent quality indicator system includes secondary indicators such as talent achievement index, talent discipline index, talent interdisciplinary index, talent social index, talent cooperation index, and talent international index, which respectively measure the quantity and quality of talent achievements, talent influence in their discipline, talent influence in the entire academic or industrial community, social attention, contribution rate in cooperation with others, and international exchanges.

[0068] The institutional platform indicator system includes secondary indicators such as the institutional platform talent scale index, institutional platform scientific research index, institutional platform social index, institutional platform cooperation index, and institutional platform international index. These indicators measure the number, scale, and types of talent cultivated by the institution, the institution's influence in the scientific research field, social attention, contribution rate in cooperation with other institutions, and internationalization, respectively.

[0069] The urban development indicator system includes secondary indicators such as the city talent attraction index, the city development platform index, the city talent training index, the city career advancement index, the city livelihood security index, and the city rational allocation index. These indicators respectively measure the city's talent competitiveness, the platforms provided for talent development, the financial support provided for talent advancement, the social conditions provided for talent career advancement, the infrastructure and service guarantees provided for talent settlement, and the assistance provided for talent mobility.

[0070] S102: Collect and obtain data on the tertiary indicators in each indicator system.

[0071] The tertiary indicators are the quantifiable indicators corresponding to the secondary indicators, and the quantification method for each tertiary indicator is determined.

[0072] Data on the quantity and type of scientific and technological achievements, the type, publication year, journal, and citation of scientific and technological papers and patents, the number and type of talents, scientific and technological papers, patents, talent policies, infrastructure, economic level, and innovation bases of each city are collected. Based on the definition and quantification method of each tertiary indicator, the specific values ​​of each tertiary indicator are calculated.

[0073] Due to the frequent updates to talent data, as a specific implementation method, web crawlers based on natural language processing (NLP) and large-scale text analysis models can be used to perform structured extraction of web page data to obtain data for third-level indicators. By performing rule-based structured extraction from publicly available websites such as city economic census yearbooks, the China Top 500 Enterprises list, and the Ministry of Science and Technology's official website, a process for indicator data collection, information extraction, and data merging is constructed. Web crawler technology based on NLP and large-scale text analysis models is then used for targeted data collection. For example... Figure 2 The process of collecting and acquiring data for the three-level indicators in various indicator systems is illustrated in the diagram. This process specifically includes:

[0074] S201: Use web crawler technology to collect data in a targeted manner to obtain web page data.

[0075] The process specifically includes:

[0076] a) Initiate an HTTP request to the target server

[0077] b) Processing HTTP responses

[0078] c) Parse the returned response body.

[0079] d) Data storage and inbound

[0080] S202: Using natural language processing technology, identify abnormal data in web page data and process the abnormal data, including but not limited to cleaning or correcting the abnormal data.

[0081] S203: Convert data from different sources and formats into a standard format suitable for analysis and calculation.

[0082] S204: Utilize the contextual understanding capabilities of large deep learning models to extract key features from the corresponding data of each city and perform data augmentation.

[0083] By leveraging the contextual understanding capabilities of large models, key features can be extracted from the economic census data of various cities, including industry classification, education level, age, etc., thereby enhancing the analytical value of the data.

[0084] S103: Based on the data distribution of the tertiary indicators in each indicator system, the data is standardized to obtain the standardized values ​​of the tertiary indicators in each indicator system.

[0085] The data were standardized using the Max-Min dimensionless method.

[0086] Where Z represents the standardized value of each tertiary indicator; x is the initial value of the corresponding tertiary indicator; min and max correspond to the lower and upper limits of the indicator distribution range, respectively.

[0087] S104: Use factor analysis to determine the weights of the tertiary indicators in each indicator system.

[0088] Understandably, factor analysis, as a statistical method, reveals the underlying structure and relationships behind indicators while determining factors and the weights of each indicator based on the information provided by the data, avoiding subjective interference and helping to understand the intrinsic connections between indicators; it also reduces the dimensionality of multiple related indicators, simplifying the analysis process while retaining important information; it is applicable to various types of data and different research fields, and has strong versatility.

[0089] This application further optimizes traditional factor analysis. Traditionally, factor analysis involves calculating the score for each factor using all relevant variables, which can capture common variations among variables to some extent. However, this method may introduce noise by including variables that are not strongly correlated with a particular factor, thus reducing the accuracy and explanatory power of factor scores. This application uses only variables highly correlated with a specific factor to calculate factor scores; such scores more accurately reflect the actual meaning and characteristics of the factor. By excluding variables with low correlation to the factor, the calculated factor scores are more accurate, reducing the random error that may be caused by uncorrelated variables. When a variable is correlated with multiple factors, selectively using the variable most correlated with a specific factor can reduce cross-interference between factor scores, enhancing the consistency and stability of specific factor scores. Therefore, this application, by precisely selecting input variables, can enhance the overall explanatory power, accuracy, and stability of factor scores.

[0090] S105: Determine the city's talent development evaluation score based on the standardized values ​​and weights of the three-level indicators in each indicator system.

[0091] The formula used to determine the evaluation score for urban talent development is:

[0092] Where π represents the city's talent development evaluation score, γ f M represents the weight of the third-level indicators in each indicator system. f This represents the standardized value of the third-level indicators in each indicator system.

[0093] As a specific implementation method, in this embodiment of the application, the weights of the third-level indicators in each indicator system are determined using factor analysis, including:

[0094] Step 1: Establish factor analysis models for the three-level indicators of each city and each level of the system:

[0095]

[0096] Among them, Y i X represents the i-th third-level indicator in each hierarchical system. j ε represents the factor whose j-th eigenvalue is greater than 1. i Let α represent the error vector. ij This represents the loading of the i-th variable on the j-th factor;

[0097] The factor analysis model is represented in matrix form as: Y = AX + ε

[0098] in, A three-level indicator vector, For the factor loading matrix, For factor vectors, Error vector.

[0099] Step 2: Perform the KMO test and Bartlett's test of sphericity to determine whether the indicators are suitable for factor analysis.

[0100] A KMO value greater than 0.5 indicates a strong correlation between variables, suitable for factor analysis. A p-value less than 0.05 in Bartlett's test of sphericity indicates a significant correlation between variables, also suitable for factor analysis. Both the KMO and Bartlett tests must be satisfied.

[0101] Step 3: Extract common factors using the principal component factorization method and solve the correlation matrix to calculate eigenvalues ​​and contribution rates.

[0102] in, The vector is formed by arranging the extracted principal components in descending order of their eigenvalues. Let λ be the eigenvalue of the correlation matrix composed of the original index Y. i (λ1…λ i The components of the eigenvector corresponding to >0).

[0103] Step 4: Determine the number of common factors w(λ) based on the eigenvalues ​​of the correlation coefficient matrix. w ≥1), classify and name the variables according to the factor loadings on each variable, that is:

[0104] Y1…Y i ∈M1

[0105] Where M1 represents factor 1, Y1…Y i The variable that has the highest factor loading coefficient on factor 1.

[0106] Step 5: Calculate the variance contribution rate and cumulative variance explained by the principal components. The variance contribution rate of the w-th principal component is: The cumulative variance explained is:

[0107] Step 6: Perform factor rotation on the variables using the maximum variance orthogonal rotation method to obtain the interpretable common factors and the rotated factor loading coefficient matrix A′=A*T, where T is an orthogonal matrix;

[0108] Step 7: Based on the rotated factor loading coefficient matrix A′, calculate the index weights and factor scores using regression estimation:

[0109]

[0110] in, The factor score vector obtained through regression estimation. This is a three-level indicator weight matrix. This represents the three-level index vector belonging to the corresponding factor category. For example, if M1 is factor 1, then Y1…Y i Only variables under this factor category are included.

[0111] Step 8: Calculate the factor weights based on the variance contribution rate and cumulative variance contribution rate of each common factor after rotation.

[0112] It is worth noting that factor scores are typically calculated using all variables (not just those belonging to the factor category). However, the factor analysis method used in this application calculates the factor score using only the variables belonging to the factor category, for the following reasons:

[0113] (1) Since each factor is often highly correlated with only a few variables, the factor scores calculated in this way can better reflect the actual significance of the factor, thereby increasing the explanatory power of the factor scores. For example, when studying the ranking of urban talent, if a factor represents the talent training support that a city can provide, using only variables related to urban talent training (such as scientific research funding provided by the city, the rate of achievement transformation, etc.) to calculate the factor score can more accurately reflect the city's performance in talent training. Other variables, such as per capita park green space area, are not related to talent training.

[0114] (2) Excluding variables that are not closely related to specific factors can reduce the random errors caused by these variables, thereby improving the accuracy of factor scores. Similarly, in the above example, removing variables such as the average park green space area can eliminate the influence of non-talent development factors, thus more accurately assessing the city's talent development status.

[0115] (3) If some variables are correlated with multiple factors, but only variables that are highly correlated with a particular factor are used when calculating the score of a specific factor, the consistency and stability of factor scores can be improved.

[0116] As one specific implementation method, the three-level indicators in this application may specifically include:

[0117] In the talent quality index system, the talent achievement index includes the H-index, activity level, number of patents, percentage of ESI highly cited papers, and patent valuation; the talent discipline index includes the average percentile and the percentage of citations by top researchers; the talent collaboration index includes the percentage of first authors, collaboration efficiency index, and percentage of industry-university-research collaboration; the talent interdisciplinary index includes the influence of standardized citations in journals, discipline diversity, and the influence of standardized citations in disciplines; the talent social index includes the number of authoritative awards in the field, online search volume, and the high-level talent index; and the talent international exchange index includes the percentage of international collaboration.

[0118] In the institutional platform indicator system, the institutional platform talent scale index includes the scale of academic talent and the scale of business talent; the institutional platform scientific research index includes the number of scientific and technological publications, the number of researchers, the number of patents, and the disciplinary influence relative to the global level; the institutional platform cooperation index includes the percentage of first authors, the cooperation index, and the percentage of industry-university-research cooperation; the institutional platform social index includes the number of award winners and the number of awards; and the institutional platform international index includes the percentage of international cooperation and the number of countries cooperating.

[0119] In the urban development indicator system, the city's talent attraction index includes the number of R&D personnel per 10,000 employed persons, the growth rate of R&D personnel, the number of high-level talents, the number of outstanding young talents, and the number of reserve talents; the city's development platform index includes the number of scientific and technological publications, the percentage of ESI highly cited papers, patent valuation, the number of research positions in research institutions, the number of Fortune Global 500 companies, the number of China's top 500 private enterprises, and the number of national key laboratories; the city's talent cultivation index includes per capita R&D expenditure, the proportion of basic research expenditure in total social R&D expenditure, and the transformation of national-level achievements. The index includes: number of platforms, number of patents per 10,000 people; urban career advancement index, including: time for talent to be promoted to a higher professional title; urban living security index, including: per capita wage, GDP growth rate, number of beds per 10,000 people, per capita commuting time, per capita residential area, population density, per capita park green space area, annual good / excellent rate of air quality index (AQI), number of cultural activity venues per capita; teacher-student ratio in primary and secondary schools; teacher-student ratio in higher education institutions, average years of education for the resident population, and education expenditure as a percentage of GDP; urban rational allocation index, including: talent retention rate, talent attraction index, and proportion of floating population.

[0120] The methods for obtaining each tertiary indicator are as follows: (The definitions and quantification methods used are detailed below.)

[0121] 1. Regarding the talent quality evaluation system

[0122] (1) Talent Achievement Index

[0123] H-index: H papers published by the talent institute are cited at least H times, and the remaining papers are cited less than or equal to H times.

[0124]

[0125] H i c represents the H-index of author i; ik This represents the number of citations for the author's k-th paper; 1(c ik The expression ≥k) determines whether the citation count of the k-th paper is greater than k, and sets it to 1 otherwise.

[0126] Activity level: Number of papers published by the talent in the past 5 years, in units of papers.

[0127] Number of patents: Total number of patents published by the talent, unit: item.

[0128] ESI Highly Cited Papers Percentage: The percentage of papers published by this talent that are ESI highly cited papers, %.

[0129] Patent valuation: The total value of the patents published by the talent, estimated with reference to the market transaction price of similar patents, in yuan.

[0130] (2) Talent Discipline Index

[0131] Mean percentile: The relative percentile ranking of talents in their respective disciplines, %

[0132]

[0133] AP i This represents the mean percentile of author i; This indicates the sequence number of author i's xth paper, which is arranged in ascending order of citations among other papers published in the same type (d), subject (f), and year (t). p represents the highest citation count among all articles published in the same year and of the same subject as x; i This indicates the total number of papers published by the author.

[0134] Citation percentage by leading experts: The percentage of papers published by this individual that are cited by leading experts in the field, in percentages.

[0135] (3) Talent Cooperation Index:

[0136] Percentage of first authors: The percentage of papers published by this author who is the first author, in %.

[0137] Cooperation Index: The performance and contribution of individuals in teamwork.

[0138]

[0139] The authors are divided into m groups according to their contributions, where l represents the l-th group in which the author belongs. Let A be the number of people in group q of the x-th collaborative paper. x This is the author's xth paper.

[0140] Percentage of industry-academia-research collaboration: The percentage of papers published by the talent pool that are co-authored with enterprises, %.

[0141] (4) Talent Interdisciplinary Index:

[0142] Journal citation impact: The influence of the journal in which a researcher publishes their paper among journals in the same or different fields.

[0143]

[0144] This represents the number of citations of author i's document x published in journal j of type d in year t; p represents the expected (average) number of citations for all documents in the same journal (j), of the same type (d), and published in the same year (t) as paper x; i This refers to the total number of papers published by the author.

[0145] Disciplinary diversity: the participation of talent in multiple fields and disciplines:

[0146]

[0147]

[0148] P a (t) represents the proportion of topics in author a's papers.

[0149] Subject-specific citation impact: The influence of a talent's published papers within their own subject area or across different subject areas:

[0150]

[0151] This represents the number of citations of author i's document x, published in year t, in subject f and of type d; p represents the expected (average) number of citations for all documents of the same subject (f), type (d), and publication year (t) as paper x; i This refers to the total number of papers published by the author.

[0152] (5) Talent Social Index:

[0153] Number of authoritative awards in the field: The number of authoritative awards received by talents in the field, in units of: awards.

[0154] Web search volume: The number of times a publication or patent is searched on an internet search engine, measured in times.

[0155] National High-Level Talent: Whether or not one is a national high-level talent; 1, yes; 0, no.

[0156] (6) International Talent Index:

[0157] International collaboration percentage: The proportion of papers published by the talent that were co-authored with foreign institutions or individuals, in percentages.

[0158] 2. Institutional Platform Evaluation System

[0159] (1) Talent Scale Index of Institutional Platforms

[0160] Academic talent scale: The number of people in the institution who have obtained titles such as "Academician", "Outstanding Young Scientist", and "Changjiang Scholar", unit: person.

[0161] Business talent scale: The number of senior executives of China's top 500 companies trained by the institution, unit: person.

[0162] (2) Institutional Platform Research Index

[0163] Number of scientific and technological publications: The number of SCI Q1 papers published by talents in the institution, in units of papers.

[0164] Number of research personnel: The number of research personnel in the institution, in person.

[0165] Number of patents: The total number of patents published by talents in the institution, in units of: patents.

[0166] Academic influence relative to global levels: the institution's relative position in the global field.

[0167]

[0168] The number of citations of document x published by the institution in year t, which is of subject f and type d; p represents the expected (average) number of citations for all publications worldwide from all institutions in the same discipline (f), type (d), and publication year (t) as paper x; i The total number of papers published by the institution.

[0169] (3) Institutional Platform Cooperation Index

[0170] First author percentage: The percentage of papers published by this institution whose authors are the first authors.

[0171] Cooperation Index: The organization's performance and contribution in teamwork.

[0172]

[0173] Collaborators are divided into m groups based on their contributions, where l represents the group in which the author from that institution belongs, and l ≤ m. Let A be the number of people in group q of the x-th collaborative paper. x This is the author's xth paper.

[0174] Percentage of industry-academia-research collaboration: The percentage of papers published by the institution that are co-authored with enterprises, in %.

[0175] (4) Social Index of Institutional Platforms

[0176] Number of award winners: The number of people in the organization who have won national-level awards, unit: person.

[0177] Number of awards: The number of national-level awards received by talents in the organization, in units of: awards.

[0178] (5) International Institutional Platform Index

[0179] International collaboration percentage: The proportion of papers published by this institution that were co-authored with foreign institutions or individuals.

[0180] Number of collaborating countries: The average number of collaborators from different countries in papers where the institution is the first or second author.

[0181] 3. Urban Development Evaluation System

[0182] (1) City Talent Attraction Index

[0183] R&D personnel per 10,000 employed persons: The number of people engaged in R&D-related positions per 10,000 employed persons in a city, unit: person.

[0184] Average annual growth rate of R&D personnel over five years: The average annual growth rate of R&D personnel in the city over the past five years.

[0185] Number of high-level talents: The number of academicians of the Chinese Academy of Sciences and the Chinese Academy of Engineering, strategic scientists and outstanding engineers in the city, in person.

[0186] Number of outstanding young talents: The number of recipients of talent programs such as the 10,000 Talents Program, the Yangtze River Talent Program, the Outstanding Young Scientists Program, and the Excellent Young Scientists Program in the city, in units of individuals.

[0187] Number of reserve talents: The number of people with doctoral degrees or above in the city, in person.

[0188] (2) Urban Development Platform Index

[0189] Number of scientific and technological publications: The total number of SCI papers published by talents in the city as the first author, in units of papers.

[0190] ESI Highly Cited Papers Percentage: The percentage of SCI papers published by talents in the city as the first author that were selected as ESI highly cited papers.

[0191] Patent valuation: The total market value of invention patents owned by the city, in RMB 10,000.

[0192] Number of research positions in scientific research institutions: The total number of scientific research positions in the city, in units of individual positions.

[0193] Number of Fortune Global 500 Companies: The number of Fortune Global 500 companies a city has, in units of companies.

[0194] Number of China's Top 500 Private Enterprises: The number of China's Top 500 Private Enterprises owned by a city, unit: enterprises.

[0195] Number of National Key Laboratories: The total number of national laboratories approved and established by the state, unit: number.

[0196] (3) Urban talent cultivation index

[0197] R&D expenditure per capita: The ratio of total urban R&D expenditure to the total number of R&D personnel, in ten thousand yuan.

[0198] The proportion of basic research funding in total social R&D expenditure: the ratio of total urban basic research funding to total R&D expenditure.

[0199] Number of national-level technology transfer platforms: The total number of national-level technology transfer platforms owned by the city, in units of: platforms.

[0200] Patent grants per 10,000 people: The average number of granted patents per 10,000 people in a city, in units of patents.

[0201] (4) Urban Career Advancement Index

[0202] Average time for professional title promotion in surveyed talents: Average promotion time for talents in cities, unit: months.

[0203] (5) Urban living security index

[0204] Average wage: The average wage of on-the-job employees in urban non-private units.

[0205] Regional GDP growth rate: The growth rate of urban GDP compared with the same period of the previous year, %.

[0206] Hospital beds per 10,000 people: The ratio of the number of hospital beds in a city to the number of permanent residents in the city (10,000 people), unit: beds.

[0207] Average commute time per person: The average one-way commute time for urban employees, in minutes.

[0208] Per capita residential area: The ratio of total urban residential area to the number of permanent residents, in square meters.

[0209] Population density: The ratio of urban area to resident population, measured in people per square kilometer.

[0210] Per capita park green space area: The ratio of urban park green space area to the number of permanent residents, expressed in square meters.

[0211] Air Quality Index (AQI) Annual Good / Good Rate: The percentage of a city with good or excellent air quality throughout the year, expressed as a percentage.

[0212] Number of cultural activity venues per capita: The ratio of the total number of cultural activity venues in the city to the number of permanent residents, in units of venues.

[0213] Student-teacher ratio in primary and secondary schools: The ratio of the total number of students enrolled in urban primary and secondary schools to the total number of full-time teachers in primary and secondary schools.

[0214] Student-to-faculty ratio in higher education institutions: The ratio of the total number of students enrolled in urban regular higher education institutions to the total number of full-time teachers in regular higher education institutions.

[0215] Average years of schooling for permanent residents: The average years of schooling for permanent residents aged 15 and above in urban areas, in years.

[0216] Education expenditure as a percentage of GDP: The proportion of education expenditure in urban general public budget expenditure to total expenditure, %.

[0217] (6) Urban rational allocation index

[0218] Talent retention rate: The ratio of the increase in talent at the end of the year to the total number of talent at the beginning of the year, %.

[0219] Talent Attraction Index: Talent Demand in Urban Areas

[0220]

[0221] FI wf P represents the talent attraction index of region w in field f. w Let P be the total population of region w. s T represents the total population of the country. wf For the number of talents in the field of f in the region, T sf This refers to the number of talents in the field of f nationwide. wf >1. Strong demand for talent; FI wf <1, the demand for talent is relatively small.

[0222] Proportion of migrant population: (resident population - registered population) / resident population

[0223] Based on any of the above embodiments, this application may further include: summing up the corresponding relationships between the third-level indicators and the second-level indicators, and between the second-level indicators and the first-level indicators, to obtain the weight of each second-level indicator and the score of each first-level indicator.

[0224] This allows us to determine the scores of each city in various dimensions, including talent attraction, development platforms, talent cultivation, career advancement, living security, and rational allocation. Based on these scores, we can rank the cities and see how different cities fare across these dimensions.

[0225] The scores of each city in the dimensions of talent attraction, development platform, talent cultivation, career advancement, living security, and rational allocation are calculated using the following formula:

[0226]

[0227] or π d =β 1d Y 1d +β 2d Y 2d +…+β id Y id

[0228] Where, π d γ represents the city's score in dimension d. f M represents the weight of the factor corresponding to dimension d. f β represents the factor score corresponding to dimension d; id Y represents the weights of the three-level indicators included in dimension d. id These are the standardized values ​​of the three-level indicators included in dimension d.

[0229] This application constructs an urban talent development evaluation system that integrates talent quality, institutional platforms, and urban development. This system considers multiple components and dimensions of talent quality and urban development, providing a more comprehensive reflection of the actual situation of urban talent development. The various components within the system are intrinsically linked; for example, scientific and technological achievements can attract and cultivate talent, while talent quality can promote the transformation of scientific and technological achievements. Institutional platforms provide development space for talent, and urban development provides a good living and working environment for talent. This comprehensive and systematic system helps to reveal the internal driving forces and mechanisms of urban talent development.

[0230] This application achieves a comprehensive, multi-level, multi-dimensional, quantitative, and systematic evaluation of various innovation entities and the innovation development environment in a city. Furthermore, factor analysis is used to determine the weights of indicators at each level. Common factors and indicator weights are determined based on the variability and correlation of the data itself, revealing the potential structure and relationships between indicators. Dimensionality reduction of indicators simplifies the analysis and calculation process while retaining important information, reducing interference from expert subjective judgment, and resulting in more objective and accurate evaluation results.

[0231] In addition, this application also provides an urban talent development evaluation device, such as... Figure 3 The structural block diagram of the urban talent development evaluation device provided in this application shows that the device specifically includes a memory 31 and a processor 32. The memory 31 stores a computer program, and when the computer program is executed by the processor 32, it implements the urban talent development evaluation method according to any of the above.

[0232] In addition, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the urban talent development evaluation method according to any of the above.

[0233] Computer-readable storage media include both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0234] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0235] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0236] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above description is only a specific embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for evaluating urban talent development, characterized in that, include: The constructed urban talent development evaluation index system is invoked. The urban talent development evaluation index system includes a talent quality index system, an institutional platform index system, and an urban development index system. Each level of the index system includes primary indicators, secondary indicators, and tertiary indicators. Each primary indicator measures the comprehensive score of talent, institution, or city. Each secondary indicator measures the influence of the evaluated subject at that level in different dimensions. Each tertiary indicator is a quantifiable indicator corresponding to the secondary indicator. Collect and acquire data on the tertiary indicators in each indicator system; Based on the data distribution of the tertiary indicators in each indicator system, the data is standardized to obtain the standardized values ​​of the tertiary indicators in each indicator system. Factor analysis was used to determine the weights of the tertiary indicators in each indicator system. In the factor analysis, variables that are highly correlated with specific factors were used to calculate factor scores. Based on the standardized values ​​and weights of the three-level indicators in each indicator system, the city's talent development evaluation score is determined: Where π represents the city's talent development evaluation score, γ f M represents the weight of the third-level indicators in each indicator system. f This represents the standardized value of the third-level indicators in each indicator system.

2. The urban talent development evaluation method according to claim 1, characterized in that, The method of determining the weights of the tertiary indicators in each indicator system using factor analysis includes: Establish factor analysis models for three-level indicators of each city and each level of the system: Among them, Y i X represents the i-th third-level indicator in each hierarchical system. j ε represents the factor whose j-th eigenvalue is greater than 1. i Let α represent the error vector. ij This represents the loading of the i-th variable on the j-th factor; The factor analysis model can be represented in matrix form as: Y = AX + ε in, A three-level indicator vector, For the factor loading matrix, For factor vectors, Error vector; Common factors were extracted using the principal component factorization method, and the correlation matrix was solved to calculate the eigenvalues ​​and contribution rates. in, The vector is formed by arranging the extracted principal components in descending order of their eigenvalues. Let λ be the eigenvalue of the correlation matrix composed of the original index Y. i (λ1…λ i The components of the eigenvector corresponding to >0); The number of common factors w(λ) is determined by selecting the eigenvalues ​​of the correlation coefficient matrix. w ≥1), classify and name the variables according to the factor loadings on each variable, that is: Y1…Y i ∈M1 Where M1 represents factor 1, Y1…Y i The variable with the highest factor loading coefficient on factor 1; Calculate the variance contribution rate and cumulative variance explained by the principal components. The variance contribution rate of the w-th principal component is: The cumulative variance explained is: The variables are rotated using the maximum variance orthogonal rotation method to obtain the interpretable common factors and the rotated factor loading coefficient matrix A′=A*T, where T is an orthogonal matrix; Based on the rotated factor loading coefficient matrix A′, regression estimation is used to calculate the index weights and factor scores: in, The factor score vector obtained through regression estimation. This is a three-level indicator weight matrix. This is a vector of third-level indicators belonging to the corresponding factor category; Calculate the weight of each factor based on the variance contribution rate and cumulative variance contribution rate of each common factor after rotation:

3. The urban talent development evaluation method according to claim 1, characterized in that, Also includes: Based on the standardized values ​​and weights of the tertiary indicators in each indicator system, the scores of each city in the dimensions of talent attraction, development platform, talent cultivation, career advancement, living security and rational allocation are determined. The scores of each city in the dimensions of talent attraction, development platform, talent cultivation, career advancement, living security, and rational allocation are calculated using the following formula: or d =b 1d Y 1d +b 2d Y 2d +…+b id Y id Where, π d γ represents the city's score in dimension d. f M represents the weight of the factor corresponding to dimension d. f β represents the score of the factor corresponding to dimension d; id Y represents the weights of the three-level indicators included in dimension d. id The standardized values ​​of the three-level indicators included in dimension d.

4. The urban talent development evaluation method according to claim 1, characterized in that, The collection and acquisition of data for the third-level indicators in each indicator system includes: Web crawlers based on natural language processing technology and large text analysis models extract structured data from web pages to obtain data on three levels of indicators.

5. The urban talent development evaluation method according to claim 4, characterized in that, Web crawlers based on natural language processing technology and text segmentation models perform structured extraction of web page data to obtain data for three-level indicators, including: Natural language processing technology is used to identify abnormal data in web page data and to process the abnormal data. Convert data from different sources and formats into a standard format suitable for analysis and computation; By leveraging the contextual understanding capabilities of large deep learning models, key features are extracted from the corresponding data of each city, and data augmentation is performed.

6. The urban talent development evaluation method according to any one of claims 1 to 5, characterized in that, The process of standardizing the data based on the data distribution of the tertiary indicators in each indicator system to obtain standardized values ​​for the tertiary indicators in each indicator system includes: The data was standardized using the Max-Min dimensionless method to obtain the standardized values ​​of the third-level indicators in each indicator system: Where Z represents the standardized value of each tertiary indicator; x is the initial value of the corresponding tertiary indicator; min and max correspond to the lower and upper limits of the indicator distribution range, respectively.

7. The urban talent development evaluation method according to any one of claims 1 to 5, characterized in that, The talent quality indicator system includes secondary indicators such as talent achievement index, talent discipline index, talent interdisciplinary index, talent social index, talent cooperation index, and talent international index. The institutional platform indicator system includes secondary indicators such as institutional platform talent scale index, institutional platform scientific research index, institutional platform social index, institutional platform cooperation index, and institutional platform international index. The urban development indicator system includes secondary indicators such as the urban talent attraction index, urban development platform index, urban talent training index, urban career advancement index, urban livelihood security index, and urban rational allocation index.

8. The urban talent development evaluation method according to claim 7, characterized in that, The collection and acquisition of data for the third-level indicators in each indicator system includes: In the talent quality indicator system, data are collected on H-index, activity level, number of patents, percentage of ESI highly cited papers, patent valuation, average percentile, percentage of citations by top researchers, percentage of first authors, collaboration efficiency index, percentage of industry-university-research collaboration, impact of standardized citations in journals, disciplinary diversity, impact of standardized citations in disciplines, number of authoritative awards in the field, web search volume, high-level talent index, and percentage of international collaboration. The platform's indicator system collects data on the scale of academic talent, the scale of commercial talent, the number of scientific and technological publications, the number of researchers, the number of patents, the academic influence of the discipline relative to the global level, the percentage of first authors, the collaboration index, the percentage of industry-university-research collaboration, the number of award winners, the number of awards, the percentage of international collaboration, and the number of countries with which the institution collaborates. The urban development indicator system collects data on the following metrics per 10,000 employed persons: number of R&D personnel, growth rate of R&D personnel, number of high-level talents, number of outstanding young talents, number of reserve talents, number of scientific and technological publications, percentage of ESI highly cited papers, patent valuation, number of research positions in research institutions, number of Fortune Global 500 companies, number of China's top 500 private enterprises, number of national key laboratories, per capita R&D expenditure, proportion of basic research expenditure, number of national-level technology transfer platforms, number of patents per 10,000 people, time for talent promotion, per capita salary, GDP growth rate, number of beds per 10,000 people, per capita commuting time, per capita residential area, population density, per capita park green space area, annual air quality index good / excellent rate, per capita number of cultural activity venues, student-teacher ratio in primary and secondary schools; student-teacher ratio in higher education institutions, average years of education for the resident population, proportion of education expenditure to GDP, talent retention rate, talent attraction index, and proportion of floating population.

9. A device for evaluating urban talent development, characterized in that, It includes a memory and a processor, the memory storing a computer program, which, when executed by the processor, implements the urban talent development evaluation method according to any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, It stores a computer program, which, when executed by a processor, implements the urban talent development evaluation method according to any one of claims 1-8.

Images