CN118411069A - Scientific and technological innovation quantitative evaluation method - Google Patents

Abstract

The invention discloses a quantitative evaluation method for technological innovation, and belongs to the field of technological innovation capability evaluation. It comprises the following steps: establishing a scientific and technological innovation quantitative evaluation index element information base of a scientific and research institution, and constructing an innovation capability evaluation index system; calculating importance of each index in the index layer, calculating a judgment matrix by using a analytic hierarchy process, and objectively assigning scientific and technological innovation quantitative evaluation indexes; selecting a laboratory to which the personal information is input, checking the integrity of the data, and determining the validity of the data; according to the effective evaluation data and the evaluation index weight, calculating innovation ability values of all evaluation objects by using linear regression, and comparing and analyzing; and visualizing the final evaluation result of the evaluation object and analysis of the result so as to analyze the evaluation result by the object. The method can provide a standard for the evaluation of each innovation platform, and provides references and guidance for the determination of the important research direction, the allocation of resources and the decision making and management.

Description

Scientific and technological innovation quantitative evaluation method

Technical Field

The invention relates to the field of scientific innovation capability assessment of scientific research institutions, in particular to a scientific innovation quantitative assessment method.

Background

The quantitative evaluation of the technological innovation plays a vital role in promoting the technological innovation capability of scientific research institutions in China, scientific research teams can be effectively organized through the quantitative evaluation of the technological innovation capability, the main direction and key content of research are clear, various technological and material resources are better configured, and guidance is provided for scientific and reasonable daily management. In addition, the good innovation evaluation system can fully reflect the direct demands of the superior management departments, so that scientific research activities can be matched with national development strategies, and scientific research results can be better served for the country.

In summary, the quantitative evaluation of technological innovation is a cow nose leading the technological innovation, and has the functions of wind vane, baton and booster. In the concrete practice of technological innovation, a scientific and reasonable quantitative evaluation system is established, and an analysis model of each evaluation element is established, so that the research direction of each scientific research institution can be better guided, the research weight can be more effectively determined, and the development and establishment of the national technological innovation system can be better promoted.

Disclosure of Invention

In view of the above, the present invention aims to provide a quantitative evaluation method for technological innovation, which constructs an elastically extensible evaluation index system and an evaluation element system of an evaluation object according to the grade, personnel composition and study direction difference of each innovation platform (laboratory), designs an evaluation model, highlights data visualization, operation facilitation and interaction affinity, develops a quantitative evaluation prototype system for technological innovation based on a B/S architecture, provides a standard for evaluation of each innovation platform, and provides references and guides for determination of key study direction, allocation of resources, decision making and management.

The invention aims at realizing the following technical scheme:

a technological innovation quantitative evaluation method comprises the following steps:

step S1: establishing a scientific and technological innovation quantitative evaluation index element information base of a scientific and research institution, and constructing an innovation capability evaluation index system;

Step S2: calculating the importance degree of each index in the index layer by using a principal component analysis method, reducing unimportant evaluation indexes, obtaining a judgment matrix by using a analytic hierarchy process, and objectively assigning the scientific and technological innovation quantitative evaluation indexes by using the analytic hierarchy process;

Step S3: the evaluation system collects data by taking a laboratory as a unit through data reported by individuals and stores the data to a local place, selects a corresponding affiliated laboratory according to the input personal information, and checks the integrity of the data by referring to a system evaluation index system to determine the validity of the collected data;

Step S4: according to the effective evaluation data and the evaluation index weight, calculating innovation ability values of all evaluation objects by using linear regression, and comparing and analyzing;

Step S5: and visualizing the final evaluation result of each evaluation object and analyzing the result.

Further, the step S1 specifically includes:

step S101: dividing the indexes collected by the index library into a plurality of different types by referring to the existing laboratory evaluation schemes at all levels and the evaluation schemes formulated by research institutions;

Step S102: the system decomposes the evaluation targets into three levels of index systems with reference to the current national level, provincial level, and each research institution's own set of evaluation program frameworks.

Further, the step S2 specifically includes:

Step S201: calculating the contribution rate of each index element based on a principal component analysis method, simplifying the evaluation index and constructing an evaluation index system;

step S202: and setting weights of the index elements by using an analytic hierarchy process, and checking and evaluating whether the weights of the indexes are reasonable.

Further, the step S3 specifically includes:

step S301: the system collects data in units of laboratories, including collecting data reported by individuals;

Step S302: storing the input information by taking departments as units;

step S303: selecting a corresponding laboratory evaluation rule scheme;

Step S304: and carrying out data verification on each item of entered information according to an evaluation index library.

Further, the step S4 specifically includes:

step S401: according to the obtained effective evaluation data, referring to the numerical values of all indexes in an evaluation index system, analyzing evaluation index elements;

Step S402: according to the obtained effective evaluation data, referring to the numerical values of all indexes in an evaluation index system, carrying out innovation capability analysis on an evaluation object;

step S403: calculating the score of each innovation body, and sequencing the innovation capacity of each evaluation object;

step S404: according to the obtained evaluation value, carrying out statistical analysis according to time;

Step S405: constructing linear regression predicts innovation ability of the evaluation object.

Further, the evaluation element index library in step S101 includes all innovative evaluation index elements, and the indexes collected by the index library are classified into different types by referring to the existing laboratory evaluation schemes and the evaluation schemes established by the research institutions, including: the personnel information module is divided into two types of special personnel talent layer constitution and external expert talents constitution, and further classified by the information indexes such as academic, title, academic honour and the like, the paper module is further classified according to the paper type, whether the paper module is a Top periodical, periodical grade, signature and the like, the patent module is further classified according to the information indexes such as the patent type, the patent state, the ranking and the like, the copyright module is further classified according to the information indexes such as the copyright type, the patent state, the ranking and the like, the standard module is further classified according to the information indexes such as the standard grade and the ranking, the integrated circuit design module is further classified according to the ranking, the technical resource module is further classified according to the information indexes such as the priority type, the ranking and the like, the information indexes such as the priority and the ranking and the like, the paper module is further classified according to the information indexes such as the paper type, whether the paper module is a Top periodical grade, the signature and the like, the patent module is further classified according to the information indexes such as the patent type, the patent state, the ranking and the like, the copyright is further classified according to the information indexes such as the copyright grade and the ranking, the standard module is further classified according to the information indexes such as the copyright type and the ranking, the standard module is further classified according to the standard grade, the information indexes such as the ranking and the ranking, the information indexes such as the priority, the technical item is further classified according to the priority, the priority and the information index, the priority and the priority, the information module is further classified according to the priority grade and the priority grade, the open communication module is further classified according to the communication type and the participation type, the department platform information module is further classified according to the platform type, and the major matters are further classified according to the major matters names and the grades;

Further, in the step S102, the evaluation system decomposes the evaluation target into three levels of index systems as follows: the index information is the basis of calculating the technological innovation capability, the above-mentioned technological innovation capability evaluation index is referred to, meanwhile, the partial first-level and second-level index systems in the "China university technological innovation capability index report (2017) are referred to, the partial indexes related to technological human resources, scientific research material conditions and technological activity investment in the" China regional technological innovation evaluation report 2021 "are referred to, the partial patent information indexes in the" national technological innovation 100 tops index report 2022-enterprise, university and research institution "are selected, the index system of the project is designed by combining the positioning and current related indexes of the research institution on the basis of the technological innovation capability evaluation index, the scientific research platform technological innovation capability evaluation index system is divided into 4 first-level indexes, 12 second-level indexes and 37 third-level indexes, wherein the 37 third-level indexes are extracted and calculated according to the 15 data type information, partial data can be multiplied by different weights according to further distinguishing grades, and the relative weights of the indexes are determined artificially.

Further, in step S201, the evaluation system calculates the contribution rate of each index element by combining with the principal component analysis method, simplifies the evaluation index and constructs an evaluation index system, and avoids the redundancy of the questionnaire of the evaluation object, so as to improve the system operation rate as follows: extracting important k (k < q) indexes from q evaluation indexes of P ₁,P₂,P₃,…,P_n by using the effect of reducing the dimension of a principal component analysis method, eliminating the correlation among indexes, synthesizing a plurality of index information into a comprehensive index value for evaluation, and converting statistical indexes of different sides, layers and different dimensions of a target object into relative evaluation values instead of simple combination of indexes, wherein the specific steps of main index element analysis are as follows:

(1) Constructing a sample matrix P _n×q:

P＝(p_ij)＝(p1,p2,…,p_q)

Wherein p _ij represents the original data of the j-th evaluation index in the evaluation object i, p _j represents the original dataset of the j-th evaluation index, i=1, 2, …, n, j=1, 2, …, q;

(2) The sample matrix is normalized to obtain a normalized matrix X _n×q:

X＝(X_ij)＝(X₁,X₂,…,X_p)

Wherein, Represents the average value of the j-th evaluation index, R _j represents the standard deviation of the j-th evaluation index,

(3) Calculate the correlation coefficient matrix CovX _q×q＝(c_ij):

Wherein c _ij denotes a sample correlation coefficient calculated based on the normalized data X _ij;

(4) Calculating a characteristic value alpha _i and a characteristic vector beta _i of the correlation coefficient matrix CovX:

|α_i·E-CovX|＝0

Wherein E is an identity matrix, α _i is a feature value, and the feature values are sorted by α ₁>α₂>…>α_q in descending order, and then the corresponding feature vectors β _i＝(β_1i,β_2i,…,β_qi)^T, i=1, 2, …, q are calculated for the sorted feature values α _i;

(5) Calculating the contribution rate of each evaluation index and the accumulated contribution rate:

Wherein U _k represents the variance contribution rate of the k-th main index element, U _k represents the accumulated contribution rate of the k (k is less than or equal to q) first index elements, and when U _k is more than or equal to delta, the k first main index elements are proved to fully reflect the main information of the original index system;

(6) Main index element model:

Wherein F _i represents the i-th main index element, F _i＝β_1i*X₁+β_2i*X₂+…+β_qi*X_q, i=1, 2,3, …, k, and the corresponding variance contribution ratio u _i is taken as the weight of each index element;

further, the evaluation system in step S202 sets weights for the index elements based on the analytic hierarchy process, and checks whether the evaluation index weights are reasonable, so as to determine the guidance and scientificity of the index weights as follows: the evaluation index is the basis of an evaluation system, the index weight is used for determining the importance degree of each index, the focus is emphasized, the weight is determined according to the variation degree of each index or the mutual relation among the indexes, the system utilizes a hierarchical analysis method and a principal component analysis method to set the weight, the hierarchical analysis method firstly carries out hierarchical structure division on a three-level evaluation index system, firstly establishes a hierarchical structure model, refers to a laboratory evaluation index hierarchical structure of fig. 2, corresponds a first-level index to a target layer H, corresponds a second-level index to a criterion layer L1-L3 and corresponds a third-level index to a scheme layer Y1-Y6, then constructs a judgment matrix of each level, carries out pairwise comparison on the criterion layers L1, L2, … and L _k under the target layer H, and evaluates the level according to the importance degree, and the weight calculation and consistency inspection process are as follows:

(1) Let M _k×k＝(m_ij) be the k-order evaluation matrix square matrix, where M _ij is the importance comparison result of the evaluation element i to the evaluation element j, i=1, 2, …, k, j=1, 2, …, k;

(2) Normalize each column vector of M to obtain a matrix V _k×k＝(v_ij), where

(3) Summing V by row yields h= (H ₁,h₂,h₃,…,h_k)^T, where

(4) Normalizing H to give d= (D ₁,d₂,d₃,…,d_k)^T, whereAfter normalization, carrying out hierarchical single sequencing, namely, the elements in D are sequencing weights of the relative importance of the same hierarchical element to the factors of the upper layer;

(5) Calculating γ= (m·d) _i/(k*d_i) as an approximation of the maximum eigenvalue, (m·d) _i represents the i-th component of m·d;

(6) Calculating a consistency index:

CI＝(γ-k)/(k-1)

when ci=0, it means perfect coincidence, the higher the coincidence is when CI is closer to 0, whereas the larger the CI is, the lower the coincidence is;

(7) Calculate the consistency ratio CR:

CR＝CI/RI

Wherein RI is introduced to measure the size of CI, RI represents an average random uniformity index, and when the uniformity ratio CR < τ, the inconsistency of M is generally considered acceptable within the allowable range, passing the uniformity test; otherwise, a comparison matrix M needs to be reconstructed, and M _ij is adjusted;

Further, the step S301 of collecting data in laboratory units by the assessment system through data reported by individuals includes: the evaluation system collects data in an online recording mode and an offline recording mode, for the introduction of early historical data, the evaluation system meets the large-scale basic data batch introduction function, for the later data updating, data updating is mainly carried out by means of an evaluation object, information collection is carried out by taking a laboratory as a unit through data reported by individuals, the starting point of data collection takes staff as a center, and the relevance between people and data is established;

Further, in the step S302, according to the entered information, the related content file stores the following data in units of departments: according to the entered information, the relevant content file is stored locally in a format of excel, csv or json by taking a department as a unit, and after various indexes of an expected evaluation system are determined, the data are stored in a database;

Further, the selecting a corresponding laboratory evaluation rule scheme in step S303 includes: the index system of the system refers to the first-level and second-level index system of the section of the index report of innovation force of science and technology (2017) of China university, the third-level index system relates to a plurality of indexes related to research and development, is not in accordance with the indexes of the project, is replaced and modified, refers to the section of indexes related to science and technology manpower resources, scientific and research material conditions and technological activity investment of the science and technology innovation environment in the science and technology innovation evaluation report 2021 of China area, refers to the section of patent information indexes of the section of the index report 2022 of national science and technology innovation 100 tops, enterprises, universities and research institutions, and designs the index system of the project by combining the positioning and current related indexes of the research institutions on the basis of referring to the index of the innovation force evaluation of the science and technology;

Further, in the step S304, performing data verification on each item of entered information according to the "evaluation index element information base" includes: the system refers to an evaluation index element information base to verify each item of recorded data item by item, and the data which is proved to be missing or invalid is verified, and the data is subtracted according to the quantity, and the final verification data is used as an effective basis for calculating the score of each index;

Further, the analyzing the evaluation index element in the step S401 includes: analyzing the whole data index of a research institution, carrying out statistics and display on all annual data, carrying out statistics and display on index elements of 15 categories, analyzing the data index of each evaluation object, calculating the contribution rate of each evaluation object to the total annual index of the research institute, calculating the respective contribution degree of each evaluation object to the indexes of the 15 categories, displaying the main contribution units of the first five of the 15 category indexes, analyzing the average contribution degree of various indexes, such as average paper number, average patent number and the like, comparing the two dimensions of historical data between each evaluation object and each past year, and displaying the index element change of the research institute and each evaluation unit;

Further, the evaluating the object to perform innovation ability analysis in step S402 includes: according to the effective data value verified in S304, calculating and displaying the overall innovation score of the research institution, innovation scores of all levels of evaluation indexes in a three-level index system, namely, 4 first-level indexes, 12 second-level indexes and 37 third-level indexes, calculating and displaying innovation scores of all levels of evaluation objects, and innovation scores of all levels of indexes of all evaluation objects, calculating and displaying contribution degrees of innovation capacity of all evaluation objects to innovation scores of the research institution, calculating and displaying a per-capitalization innovation index I ₁,I₁ = innovation score/total number of all levels of evaluation objects, considering the size of a department where the evaluation objects are located, the number of research personnel, calculating and displaying talent innovation index I ₂ of all evaluation objects, considering the talent number of all levels of the research personnel, and I ₂ = innovation score/talent comprehensive score, wherein talent comprehensive score is Sigma (talent type number), calculating the average score of the research institution, comparing difference between each evaluation object and the average score, finding out specific indexes of differences, screening out indexes to be lifted, and comparing the history of two-year change history data from each evaluation object and obtaining the innovation history data of each evaluation object;

Further, in the step S403, the calculating the score of each innovation body, and the sorting of the innovation capability of each evaluation object includes: after the information is input, checking the data integrity, and calculating innovation ability values of each evaluation object of the institute and subordinate according to each index weight in the step S201, so as to realize the ordering of innovation ability of each evaluation object;

Further, in the step S404, performing statistical analysis on the quarter and year data according to the obtained evaluation value includes: after the information is input, checking the data integrity, calculating innovation ability values of the research institute and each subordinate evaluation object according to S403, carrying out quarter and year data statistical analysis, and displaying technological innovation ability and change trend of the evaluated object from transverse and longitudinal dimensions;

Further, the step S405 of constructing a linear regression to predict the innovation ability of the evaluation object includes: the method comprises the following specific steps of normalizing the original data through the minimum maximum value, eliminating the influence of different evaluation index score ranges on the result, selecting partial normalized data as a sample set S _m×n, carrying out innovation score prediction on a plurality of evaluation objects, and carrying out multiple sample multiple linear regression:

(1) Normalizing the original data through minimum and maximum normalization:

X_scaled＝(X-X_min)*(max-min)/(X_max-X_min)+min

Wherein, X ^* represents the data normalization result, X _min represents a row vector composed of minimum values in each column, X _max represents a row vector composed of maximum values in each column, and max and min represent the maximum value and the minimum value of the interval range of the data mapping respectively;

(2) Based on multiple sample multiple linear regression, selecting m samples after normalization processing, and n evaluation indexes to form a sample group S= [ S ₁,s₂,…,s_m]^T ], and then predicting the value

Wherein s _i represents the sample feature vector s _i＝[1,s_1i,s_2i,…,s_ni]^T (i=1, 2, …, m), the true value is denoted as y _m×1＝[y₁,y₂,…,y_m]^T, and the predicted value is set asThe weight parameter vector w= [ b, w ₁,w₂,…,w_n]^T, T represents the matrix transpose;

(3) The prediction loss is measured by adopting the mean square error, the prediction fitting degree is estimated, and the average value of the squares of the difference between the predicted value and the actual observed value is calculated to obtain

Further, the visualizing the analysis result obtained by the six modules of the evaluation system in the step S501 includes: referring to fig. 3, the present project is intended to divide a laboratory evaluation system into six main functional modules, according to the evaluation goals and evaluation schemes: basic laboratory information, annual targets, annual data, evaluation indexes, evaluation output and evaluation summary, and meanwhile, the system meets the expandability requirement of the functional modules, and can be added, deleted or adjusted according to actual needs, and each functional module is as follows: laboratory basic information module: for counting laboratory projects and personnel conditions, clicking corresponding sub-modules can check specific information of each project, when personnel change or newly added projects occur, The system can maintain, modify or add related information, when an administrator needs to add employee information, the administrator enters an employee adding interface, fills in employee names and employee numbers according to page prompts, and as the employee numbers are primary keys of the employee information, the system can carry out unique inspection on the employee numbers, then the administrator fills in basic information such as personal names, sexes, academies, birth months, hosting projects and the like of the employees, clicks "submit", the software background verifies whether the input content is available or not, if not, the unavailable content is displayed, and is required to be submitted again after modification, if so, the page content is acquired and is filled into two entity classes, And invoking a corresponding method to perform data persistence operation, and adding employee records in a database, referring to fig. 4; the annual target setting module: the method is used for setting differential targets according to the specific conditions of each laboratory, firstly, the corresponding laboratory is selected by the evaluation targets, related evaluation indexes can be selected, target tasks of the year can be input one by one, the annual targets of the laboratory/department can be imported in batches in an Excel importing mode, after the annual targets are input, after 'confirmation setting' is clicked, the software background reminds the evaluation targets whether all the annual target tasks of the evaluation targets are input, if the annual targets are not all set, the input of the annual targets is continued, if all the annual targets are set, corresponding methods are called to perform target evaluation and detection operations, the annual target data are added in a database, updating the database, importing relevant data into an evaluation system after target setting, facilitating the inspection of all laboratory staff, and simultaneously storing the data for index comparison during annual end evaluation, referring to fig. 5; Annual data module: the method is used for running all data in a statistical laboratory in one year, including personnel, rewards, national level projects, provincial level projects, article patents, newly-added instruments and the like, selecting the corresponding laboratory to introduce evaluation data of Excel into the laboratory, manually adding part of the data to perform qualification detection on the introduced data, submitting the data again after modification if the data are unqualified, and calling a corresponding method to save the data into a corresponding database if the data are qualified, wherein the data are not all finally used as evaluation indexes, and the storage of the data is spare, so that the management efficiency of the laboratory can be improved, and referring to FIG. 6; And an evaluation index module: the performance evaluation index system of the project adopts a three-level index system, the three-level index system jointly forms a two-level index, the two-level index jointly forms a first-level index for setting or revising the evaluation index and the weight, when index information needs to be added, a user enters an index information adding interface, fills in index names and corresponding weight after selecting index levels, repeatedly detects the added index, if the index is repeated, the index needs to be revised and re-input, if the index is not repeated, the new index can be submitted by clicking, the system can perform qualification check after the new index is submitted, the qualified evaluation index acquires page content and fills in two entity classes, And invoking a corresponding method to perform data persistence operation, and refreshing index information in a database, referring to fig. 7; And an evaluation output module: the system is used for checking the qualification of the evaluation index by selecting the consulted evaluation index according to the statistical evaluation index and the annual achievement of the laboratory, saving the index passing the check, checking the integrity of the saved evaluation data, selecting the laboratory corresponding to the evaluation index and selecting the evaluation data after evaluating all the indexes, calculating and summarizing the evaluation achievement based on the performance calculation model, and finally generating an evaluation report, and referring to fig. 8; and an evaluation summarization module: the module home page displays the summary of the comparison of all laboratory evaluation results of the year, comprises laboratory names, laboratory levels, grading results and evaluation levels, can sort the summarized evaluation results according to the evaluation results, calculates the evaluation levels according to the laboratory levels, can click a 'select laboratory', compares and analyzes the previous annual evaluation results of the laboratory, calculates the evaluation levels, divides the evaluation results into four levels of excellent, good, qualified and unqualified, and finally outputs an evaluation summary report to analyze which aspects have advanced and which aspects need further effort, and refer to fig. 9;

Further, in step S502, the evaluation object checks the evaluation condition through the system, and feeds back the evaluation result as follows: based on personal achievements and scores, calculating and displaying each index presentation condition of a department according to the department to which the person belongs through system visualization, presenting the overall condition of a research institute based on the achievements and scores of each department, setting monitoring functions for each innovation subject according to evaluation task requirements, monitoring and reminding incomplete tasks, and intensively exporting an evaluation index system and analysis and calculation results through an evaluation system by each evaluation subject to generate an evaluation report;

Further, in step S503, the evaluation system will collect feedback data of each dimension, and optimize the evaluation system to obtain a more accurate algorithm model as follows: the system receives feedback of the evaluation object to the structure, the evaluation system collects feedback data of each dimension, optimizes the evaluation system, such as index filling data, user feedback and evaluation object spot check, and further updates the algorithm model to obtain a more accurate algorithm model.

The beneficial effects of the invention are as follows:

(1) The invention can carry out standardized and modularized management on laboratory information and data indexes required by evaluation, divides a laboratory performance evaluation management system into a plurality of sub-functional modules, constructs an elastically extensible evaluation index system and an evaluation object evaluation element system according to the standardized and modularized management, designs an evaluation model, highlights data visualization, operation facilitation and interaction affinity, develops a scientific innovation quantitative evaluation prototype system based on a B/S architecture, and jointly completes daily management and annual evaluation work of a laboratory.

(2) According to the invention, research emphasis and characteristics of each level laboratory of a research institution are fully considered in the selection of the evaluation indexes, the selection of the indexes refers to the evaluation rules of the same-level laboratory on one hand, and meanwhile, a supervisor unit is allowed to newly increase the evaluation indexes according to scientific research guidance and research characteristics.

(3) The invention combines the subjective weighting method based on the analytic hierarchy process and the objective analytic method based on the principal component analytic process to ensure the guidance and scientificity of the index weight, the traditional subjective assignment method easily causes that the actual result is difficult to reflect the objective actual situation, the weight setting result is constructed into a judgment matrix, and the rationality of the weight is checked by adopting a consistency check method to ensure the scientificity of the whole evaluation system.

(4) The invention can fully reflect the direct demands of the superior management departments, ensure that scientific research activities are matched with national development strategies, and ensure that scientific research results can better serve the countries.

(5) The invention is an effective organization scientific research team, defines main directions and key contents of research, provides guidance for better configuration of various technologies and material resources and scientific and reasonable daily management, and objectively evaluates technological innovation activities and results so as to carry out transverse and longitudinal comparison, accurately evaluate innovation capability and level, excite scientific research innovation vitality and ensure technological innovation sustainability.

Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof.

Drawings

For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings, in which:

FIG. 1 is a flowchart of a quantitative evaluation method of technological innovation according to the present invention.

FIG. 2 is a diagram showing an example of a hierarchical structure of laboratory evaluation indicators according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating an exemplary laboratory evaluation platform module according to an embodiment of the present invention.

Fig. 4 is a diagram illustrating a laboratory employee information addition flow according to an embodiment of the present invention.

FIG. 5 is a diagram illustrating an example of a annual objective setting procedure according to an embodiment of the present invention.

FIG. 6 is a diagram illustrating an example of a laboratory annual data addition process in accordance with an embodiment of the present invention.

Fig. 7 is a diagram illustrating an exemplary evaluation index and weight adding process according to an embodiment of the present invention.

Fig. 8 is a diagram showing an example of an evaluation output flow according to an embodiment of the present invention.

Fig. 9 is an exemplary diagram of an evaluation summary flow according to an embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made with reference to the accompanying drawings so as to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, but is to be construed as a protection of all the invention which utilizes the inventive concept as long as the variations are within the spirit and scope of the present invention as defined and defined by the appended claims to those skilled in the art.

As shown in fig. 1, the present embodiment provides a quantitative evaluation method for technological innovation, which includes the following steps:

Step S1: establishing a scientific and technological innovation quantitative evaluation index element information base of a scientific and technological research institution, and constructing an innovation capability evaluation index system.

In this embodiment, the laboratory evaluation index hierarchy is shown in fig. 2.

In this embodiment, the step S1 includes the following sub-steps:

step S101: the assessment system classifies the metrics collected by the metrics library into a number of different types.

In step S101, the index library of the evaluation element includes all innovative evaluation index elements, referring to the existing laboratory evaluation schemes and evaluation schemes established by research institutions, the evaluation system classifies the indexes collected by the index library into 15 types, which are respectively classified into a personnel information module, a paper module, a patent module, a monograph module, a standard module, an integrated circuit layout module, a scientific and technical prize module, a research project module, a result appraisal module, a result conversion module, a scientific and technical innovation development planning module, a software copyright module, an environment foundation module, a development communication module and a department platform information module, and classifies the information required to be collected by each information module and the grade of the information, wherein the personnel information module is classified into a special talent layer structure and an external talent layer structure, and can be further classified into further grades by the information indexes such as a paper type, a Top journal, a journal grade, a signature and the like, the patent module is further classified into further grades according to the information indexes such as a patent type, a patent state, a ranking and the like, the information indexes such as a work type and a ranking, the information module is further classified into a ranking and an information grade according to the standard grade, the priority and a priority level, and a further classification level is further classified into a source index classification according to the information index such as a priority level and a priority level, and a further classification level is further classified into a priority level according to the information index such as a priority level and a priority level, the achievement identification module is further classified according to achievement levels and ranked information indexes, the successful conversion module is further classified according to conversion types, achievement types, benefits and other information indexes, the software copyright module is further classified according to ranks, the open communication module is further classified according to communication types and participation types, the department platform information module is further classified according to platform types, and important matters are further classified according to important matter names and grades.

Step S102: the evaluation system decomposes the evaluation target into a three-level index system of '4+12+37'.

The index information in step S102 is the basis for calculating the technological innovation capability, the above-mentioned technological innovation capability evaluation index is referred to, and meanwhile, the index system of the present project is designed by referring to the partial first-level and second-level index systems in the "national university technological innovation capability index report (2017)", the partial indexes related to technological human resources, scientific research material conditions and technological activity investment in the "national technological innovation evaluation report (2021)", the partial patent information indexes in the "national technological innovation 100 tops index report 2022-enterprise, university and research institution" are selected, and the relative weights of these indexes are determined by manually multiplying different weights according to the further differentiated grades in accordance with the above-mentioned 15 data type information extraction calculation, and the index system of the present project is designed by combining the positioning and current related indexes of the research institution on the basis of the above-mentioned technological innovation capability evaluation index.

Step S2: the importance degree of each index in the index layer is calculated by using a principal component analysis method, the unimportant evaluation index is reduced, a judgment matrix is obtained by using a hierarchical analysis method, and objective assignment is carried out on the scientific and technological innovation quantitative evaluation index by using the hierarchical analysis method.

In this embodiment, S2 specifically includes the following sub-steps:

Step S201: and calculating the contribution rate of each index element based on a principal component analysis method, and simplifying the evaluation index.

In step S201, important k (k < q) indexes are extracted from q evaluation indexes of P ₁,P₂,P₃,…,P_n n evaluation objects by the effect of dimension reduction by principal component analysis, correlation among indexes is eliminated, a plurality of index information is synthesized into one comprehensive index value for evaluation, and statistical indexes of different sides, layers and different dimensions of the target object are converted into relative evaluation values instead of simple combination of indexes, wherein the specific steps of main index element analysis are as follows:

(1) Constructing a sample matrix P _n×q:

P＝(p_ij)＝(p₁,p₂,…,p_q)

Wherein p _ij represents the original data of the j-th evaluation index in the evaluation object i, p _j represents the original dataset of the j-th evaluation index, i=1, 2, …, n, j=1, 2, …, q;

(2) The sample matrix is normalized to obtain a normalized matrix X _n×q:

X＝(X_ij)＝(X₁,X₂,…,X_p)

Wherein, Represents the average value of the j-th evaluation index, R _j represents the standard deviation of the j-th evaluation index,

(3) Calculate the correlation coefficient matrix CovX _q×q＝(c_ij):

Wherein c _ij denotes a sample correlation coefficient calculated based on the normalized data X _ij;

(4) Calculating a characteristic value alpha _i and a characteristic vector beta _i of the correlation coefficient matrix CovX:

|α_i·E-CovX|＝0

Wherein E is an identity matrix, α _i is a feature value, and the feature values are sorted by α ₁>α₂>…>α_q in descending order, and then the corresponding feature vectors β _i＝(β_1i,β_2i,…,β_qi)^T, i=1, 2, …, q are calculated for the sorted feature values α _i;

(5) Calculating the contribution rate of each evaluation index and the accumulated contribution rate:

Wherein U _k represents the variance contribution rate of the k-th main index element, U _k represents the accumulated contribution rate of the k (k is less than or equal to q) first index elements, and when U _k is more than or equal to delta, the k first main index elements are proved to fully reflect the main information of the original index system;

(6) Main index element model:

Wherein F _i represents the i-th main index element, F _i＝β_1i*X₁+β_2i*X₂+…+β_qi*X_q, i=1, 2,3, …, k, and the corresponding variance contribution ratio u _i is taken as the weight of each index element;

step S202: and setting weights of the index elements based on an analytic hierarchy process, and checking and evaluating whether the weights of the indexes are reasonable.

In the step S101, the evaluation index is the basis of the evaluation system, the index weight is used to determine the importance degree of each index, the focus is emphasized, the weight is determined according to the variation degree of each index or the interrelation between each index, the system uses two methods of the analytic hierarchy process and the principal component analysis to set the weight, the analytic hierarchy process firstly carries out the hierarchical structure division on the three-level evaluation index system, firstly establishes the hierarchical structure model, and referring to the laboratory evaluation index hierarchical structure of fig. 2, the first-level index corresponds to the target layer H, the second-level index corresponds to the criterion layer L1-L3, the third-level index corresponds to the scheme layer Y1-Y6, then constructs the judgment matrix of each layer, carries out the pairwise comparison on the criterion layers L1, L2, …, L _k under the target layer H, and evaluates the level according to the importance degree, the weight calculation and the consistency test process is as follows:

(1) Let M _k×k＝(m_ij) be the k-order evaluation matrix square matrix, where M _ij is the importance comparison result of the evaluation element i to the evaluation element j, i=1, 2, …, k, j=1, 2, …, k;

(2) Normalize each column vector of M to obtain a matrix V _k×k＝(v_ij), where

(3) Summing V by row yields h= (H ₁,h₂,h₃,…,h_k)^T, where

(4) Normalizing H to give d= (D ₁,d₂,d₃,…,d_k)^T, whereAfter normalization, carrying out hierarchical single sequencing, namely, the elements in D are sequencing weights of the relative importance of the same hierarchical element to the factors of the upper layer;

(5) Calculating γ= (m·d) _i/(k*d_i) as an approximation of the maximum eigenvalue, (m·d) _i represents the i-th component of m·d;

(6) Calculating a consistency index:

CI＝(γ-k)/(k-1)

when ci=0, it means perfect coincidence, the higher the coincidence is when CI is closer to 0, whereas the larger the CI is, the lower the coincidence is;

(7) Calculate the consistency ratio CR:

CR＝CI/RI

Wherein RI is introduced to measure the size of CI, RI represents an average random uniformity index, and when the uniformity ratio CR < τ, the inconsistency of M is generally considered acceptable within the allowable range, passing the uniformity test; otherwise, a comparison matrix M needs to be reconstructed, and M _ij is adjusted;

In this embodiment, the M _ij is equally important as 1, slightly important as 3, relatively strong as 5, strongly important as 7, extremely important as 9, and the intermediate value between two adjacent judgments is 2/4/6/8, so that the 3-order square matrix m= { (1, 3, 7), (1/3,1,5), (1/7, 1/5, 1) }, the column vector is normalized to v= { (0.678,0.714,0.538), (0.226,0.238,0.385), (0.097,0.048,0.077) } and the row is summed to h= (1.930,0.849,0.221) ^T, and the row is normalized to d= (0.643,0.283,0.074) ^T, so that m·d= (2.008,0.866,0.222) ^T is calculated to obtain a consistency index ci=0.03, γ=3.06, and the random consistency index look-up table ri=0.58, τ=0.1, so that the consistency ratio cr=0.052 < τ is obtained, and the consistency matrix is considered.

Step S3: the evaluation system collects data by taking a laboratory as a unit through data reported by individuals, and stores related content files to a local place by taking the laboratory as a unit; and selecting a corresponding affiliated laboratory according to the entered personal information, checking the integrity of the data by referring to a system evaluation index system, and determining the validity of the acquired data.

In this embodiment, S3 specifically includes the following sub-steps:

step S301: the assessment system collects data in laboratory units, including data reported by individuals.

In step S301, data collection is performed in two ways of online input and offline input, for early historical data input, the evaluation system satisfies a large-scale basic data batch input function, for later data update, data update is performed mainly by means of an evaluation object, information collection is performed by means of data reported by individuals in a laboratory unit, a departure point of data collection is centered on staff, and a person-to-data relevance is established.

Step S302: and storing the recorded information in units of departments.

In step S105, according to the entered information, the relevant content file is stored locally in the format of "excel", "csv" or "json" in units of departments, and after each index of the evaluation system is determined, the data is stored in the form of database.

Step S303: a corresponding laboratory evaluation rules scheme is selected.

The laboratory evaluation rules scheme includes: the index system of the system refers to the first-level and second-level index system of the section of the index report of the innovation of science and technology (2017) of China university, the third-level index system relates to a plurality of indexes related to research and development, the indexes are not in accordance with the indexes of the project, the indexes are replaced and modified, the section of indexes related to the scientific and technological human resources, scientific and research material conditions and technological activity investment in the technological innovation environment of the section of the index report 2021 of the innovation of science and technology in China area is referred to, and the index system of the project is designed by referring to the information indexes of the section of patent in the index report 2022 of the innovation of science and technology 100 tops of China, enterprises, universities and research institutions on the basis of referring to the indexes of the evaluation of the innovation of the science and technology.

Step S304: and carrying out data verification on each item of entered information according to an evaluation index library.

And (3) checking each item of entered data item by referring to an evaluation index library, checking and subtracting the data which is proved to be missing or invalid by quantity, and taking the final checked data as an effective basis for calculating the score of each index.

Step S4: and calculating innovation capability values of all evaluation objects by using linear regression according to the effective evaluation data and the evaluation index weight, and comparing and analyzing.

In this embodiment, S4 specifically includes the following sub-steps:

step S401: and (5) evaluating index elements for analysis.

The evaluation index element analysis includes: analyzing the whole data index of a research institution, carrying out statistics and display on all annual data, carrying out statistics and display on index elements of 15 categories, analyzing the data index of each evaluation object, calculating the contribution rate of each evaluation object to the total annual index of the research institute, calculating the respective contribution degree of each evaluation object to the indexes of the 15 categories, displaying the main contribution units of the first five of the indexes of the 15 categories, analyzing the average contribution degree of various indexes, such as average paper number, average patent number and the like, comparing the two dimensions of historical data between each evaluation object and each past year, and displaying the index element change of the research institute and each evaluation unit.

Step S402: the evaluation object performs innovation ability analysis.

The evaluation object performs innovation ability analysis including: according to the effective data value verified in S304, calculating and displaying the whole innovation score of the research institution, innovation scores of all levels of evaluation indexes in a three-level index system, namely, 4 first-level indexes, 12 second-level indexes and 37 third-level indexes, calculating and displaying innovation scores of all levels of evaluation objects, innovation scores of all levels of indexes of all evaluation objects, calculating and displaying contribution degrees of innovation capacity of all evaluation objects to the innovation score of the whole research institution, calculating and displaying a main contribution unit of a human innovation index I ₁,I₁ = innovation score/total number of all levels of evaluation indexes, considering the size of a department where each evaluation object is located, the number of research personnel, calculating and displaying a human innovation index I ₂ of all evaluation objects, considering the number of human talents owned by the department, I ₂ = innovation/human talent comprehensive score, human talent comprehensive score = (human talent type number), calculating the average innovation of the research institution, comparing the difference between each evaluation object and the average score, finding out specific indexes of differences, finding out indexes of differences, needing to be promoted by each evaluation object, comparing the difference between each evaluation object and the two dimension, and carrying out the change history of the innovation indexes, and improving the change history of each evaluation object.

Step S403: and calculating the score of each innovation body, and sequencing the innovation capacity of each evaluation object.

After the information is input, checking the data integrity, and calculating innovation ability values of each evaluation object of the institute and subordinate according to each index weight in the step S201, so as to realize the ordering of innovation ability of each evaluation object.

Step S404: statistical analysis of quarter and year data statistical analysis is performed according to time based on the obtained evaluation values.

After the information is input, checking the data integrity, calculating innovation ability values of the research institute and each subordinate evaluation object according to S403, carrying out quarter and year data statistical analysis, and displaying technological innovation ability and change trend of the evaluated object from transverse and longitudinal dimensions.

Step S405: linear regression predicts the innovation ability of the evaluation object.

The method comprises the following specific steps of normalizing the original data through the minimum maximum value, eliminating the influence of different evaluation index score ranges on the result, selecting partial normalized data as a sample set S _m×n, carrying out innovation score prediction on a plurality of evaluation objects, and carrying out multiple sample multiple linear regression:

(1) Normalizing the original data through minimum and maximum normalization:

X^*＝(X-X_min)*(max-min)/(X_max-X_min)+min；

Wherein, X ^* represents the data normalization result, X _min represents a row vector composed of minimum values in each column, X _max represents a row vector composed of maximum values in each column, and max and min represent the maximum value and the minimum value of the interval range of the data mapping respectively;

(2) Based on multiple sample multiple linear regression, selecting m samples after normalization processing, and n evaluation indexes to form a sample set S= [ S ₁,s₂,…,s_m]^T ], and then predicting the value

Wherein s _i represents the sample feature vector s _i＝[1,s_1i,s_2i,…,s_ni]^T (i=1, 2, …, m), the true value is denoted as y _m×1＝[y₁,y₂,…,y_m]^T, and the predicted value is set asThe weight parameter vector w= [ b, w ₁,w₂,…,w_n]^T, T represents the matrix transpose;

(3) The prediction loss is measured by adopting the mean square error, the prediction fitting degree is estimated, and the average value of the squares of the difference between the predicted value and the actual observed value is calculated to obtain

Step S5: and visualizing the final evaluation result and result analysis of each evaluation object so that the evaluation object can view the evaluation result.

In this embodiment, the laboratory evaluation platform module is shown in fig. 3, the laboratory employee information adding flow is shown in fig. 4, the annual target setting flow is shown in fig. 5, the laboratory annual data adding flow is shown in fig. 6, the evaluation index and weight adding flow is shown in fig. 7, the evaluation output flow is shown in fig. 8, and the evaluation summarizing flow is shown in fig. 9.

In this embodiment, S5 specifically includes the following sub-steps:

Step S501: and visualizing the obtained analysis result through each functional module of the evaluation system.

Visualizing the analysis results obtained by the functional modules of the evaluation system includes the following:

according to the evaluation objective and the evaluation scheme, fig. 3 in this embodiment divides the laboratory evaluation system into six main functional modules: basic laboratory information, annual targets, annual data, evaluation indexes, evaluation output and evaluation summary, and meanwhile, the system meets the expandability requirement of the functional modules, and can be added, deleted or adjusted according to actual needs, and each functional module is as follows: laboratory basic information module: the method is used for counting laboratory projects and personnel conditions, clicking corresponding sub-modules can check specific information of each project, when personnel change or newly added projects occur, related information can be maintained, modified or added, when an administrator needs to add employee information, Entering an employee adding interface, filling in employee names and employee numbers according to page prompts, and because the employee numbers are primary keys of employee information, the system can check the uniqueness of the employee numbers, then an administrator fills in basic information such as personal names, sexes, academies, birth year months, hosting projects and the like of the employees, after filling in, clicks on "submit", a software background verifies whether the input content is available, if not, the unavailable content is displayed, and is required to be submitted again after modification, if yes, the page content is acquired and is filled into two entity classes, and a corresponding method is called to perform data persistence operation, employee records are added in a database, The information adding flow of the laboratory staff is shown in fig. 4; the annual target setting module: the method is used for setting differential targets according to the specific conditions of each laboratory, firstly, the corresponding laboratory is selected by the evaluation targets, related evaluation indexes can be selected, target tasks of the year can be input one by one, the annual targets of the laboratory/department can be imported in batches in an Excel importing mode, after the annual targets are input, after 'confirmation setting' is clicked, the software background reminds the evaluation targets whether all the annual target tasks of the evaluation targets are input, if the annual targets are not all set, the input of the annual targets is continued, if all the annual targets are set, corresponding methods are called to perform target evaluation and detection operations, the annual target data are added in a database, Updating the database, importing relevant data into an evaluation system after target setting, facilitating the inspection of all laboratory staff, and simultaneously storing the data for index comparison when being used for annual end evaluation, wherein the annual target setting flow is shown in figure 5; Annual data module: the method is used for running all data in a statistical laboratory in one year, including personnel, rewards, national level projects, provincial level projects, article patents, newly-added instruments and the like, selecting the corresponding laboratory to introduce evaluation data of Excel into the laboratory, manually adding part of the data to perform qualification detection on the introduced data, submitting the data again after modification if the data are unqualified, and calling a corresponding method to save the data into a corresponding database if the data are qualified, wherein the data are not all finally used as evaluation indexes, the data are stored for standby, the management efficiency of the laboratory can be improved, and the annual data addition flow of the laboratory is shown in a figure 6; And an evaluation index module: the performance evaluation index system of the project adopts a three-level index system, the three-level index system jointly forms a two-level index, the two-level index jointly forms a first-level index for setting or revising the evaluation index and the weight, when index information needs to be added, a user enters an index information adding interface, fills in index names and corresponding weight after selecting index levels, repeatedly detects the added index, if the index is repeated, the index needs to be revised and re-input, if the index is not repeated, the new index can be submitted by clicking, the system can perform qualification check after the new index is submitted, the qualified evaluation index acquires page content and fills in two entity classes, And calling a corresponding method to perform data persistence operation, refreshing index information in a database, and enabling an evaluation index and weight adding flow to be shown in fig. 7; And an evaluation output module: the system is used for checking the qualification of the evaluation index by selecting the consulted evaluation index according to the statistical evaluation index and the annual achievements of the laboratory, storing the index passing the check, carrying out integrity check on the stored evaluation data, selecting the laboratory corresponding to the evaluation index and carrying out evaluation data selection on the laboratory corresponding to the evaluation index after all indexes are evaluated, calculating and summarizing the evaluation achievements based on the performance calculation model, and finally generating an evaluation report, wherein the evaluation output flow is shown in figure 8; and an evaluation summarization module: the module home page displays the summary of the comparison of all laboratory evaluation results in the year, comprises laboratory names, laboratory levels, grading results and evaluation levels, can sort the summarized evaluation results according to the evaluation results, calculates the evaluation levels according to the laboratory levels, can click a 'select laboratory', compares and analyzes the annual evaluation results before the laboratory, calculates the evaluation levels, divides the evaluation results into four levels of good, qualified and unqualified, and finally outputs an evaluation summary report, analyzes which aspects need further effort, and an evaluation summary flow is shown in fig. 9.

Step S502: and the evaluation object looks up the evaluation result through the system and feeds back the evaluation result.

The evaluation object checks the evaluation condition through the system, and the feedback process of the evaluation result is as follows:

Based on personal achievements and scores, the system visualizes, calculates and displays each index presentation condition of one department according to the department to which the person belongs, presents the overall condition of the research institute based on the achievements and scores of each department, sets monitoring functions for each innovation subject according to the evaluation task requirements, monitors and reminds the tasks which are not completed, and each evaluation object can intensively export an evaluation index system and analysis and calculation results through an evaluation system to generate an evaluation report.

Step S503: the evaluation system collects feedback data of each dimension, and optimizes the evaluation system, so that the algorithm model is more accurate.

The system receives feedback of the evaluation object to the structure, the evaluation system collects feedback data of each dimension, optimizes the evaluation system, such as index filling data, user feedback and evaluation object spot check, and further updates the algorithm model to obtain a more accurate algorithm model.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The principles and embodiments of the present invention have been described in detail with reference to specific examples, which are provided to facilitate understanding of the method and core ideas of the present invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present invention, the present description should not be construed as limiting the present invention in view of the above.

Those of ordinary skill in the art will recognize that the embodiments described herein are for the purpose of aiding the reader in understanding the principles of the present invention and should be understood that the scope of the invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific modifications and combinations from the teachings of the present disclosure without departing from the spirit thereof, and such modifications and combinations remain within the scope of the present disclosure.

Claims (10)

1. The technological innovation quantitative evaluation method is characterized by comprising the following steps of:

Step S1, establishing a scientific and technological innovation quantitative evaluation index element information base of a scientific and technological research institution, and constructing an innovation capability evaluation index system;

S2, calculating importance of each index in an index layer by using a principal component analysis method, reducing unimportant evaluation indexes, obtaining a judgment matrix by using a hierarchical analysis method, and objectively assigning scientific and technological innovation quantitative evaluation indexes by using the hierarchical analysis method;

Step S3, collecting data by using a laboratory as a unit through data reported by individuals, storing the data to a local place, selecting a corresponding laboratory according to the input personal information, checking the integrity of the data by referring to a system evaluation index system, and determining the validity of the acquired data;

s4, calculating innovation ability values of all evaluation objects by using linear regression according to the effective evaluation data and the evaluation index weight, and comparing and analyzing;

And S5, visualizing the final evaluation result of each evaluation object and analyzing the result.

2. The quantitative evaluation method of technological innovation according to claim 1, wherein the step S1 specifically includes:

S101, dividing indexes collected by an index library into a plurality of different types;

S102, decomposing the evaluation target into a three-level index system.

3. The quantitative evaluation method of technological innovation according to claim 1, wherein the step S2 specifically includes:

s201, calculating the contribution rate of each index element based on a principal component analysis method, simplifying an evaluation index and constructing an evaluation index system;

S202, setting weights of index elements by using an analytic hierarchy process, and checking and evaluating whether the weights of the indexes are reasonable.

4. The quantitative evaluation method of technological innovation according to claim 1, wherein the step S3 specifically includes:

s301, collecting data in a laboratory unit, including collecting data reported by individuals;

S302, storing the input information in units of departments;

s303, selecting a corresponding laboratory evaluation rule scheme;

s304, verifying the data of each item of entered information according to an evaluation index library.

5. The quantitative evaluation method of technological innovation according to claim 1, wherein the step S4 specifically includes:

S401, according to the obtained effective evaluation data, referring to the numerical values of all indexes in an evaluation index system, analyzing evaluation index elements;

s402, carrying out innovation capability analysis on an evaluation object by referring to the numerical values of various indexes in an evaluation index system according to the obtained effective evaluation data;

s403, calculating the score of each innovation body, and sequencing innovation capacity of each evaluation object;

S404, carrying out statistical analysis according to time according to the obtained evaluation value;

s405, constructing linear regression to predict innovation ability of the evaluation object.

6. The quantitative evaluation method of technological innovation according to claim 1, wherein the step S5 specifically includes:

S501, visualizing the obtained analysis result through each functional module of the evaluation system;

s502, the evaluation object checks the evaluation result through the system and feeds back the evaluation result;

S503, the evaluation system collects feedback data of each dimension, and optimizes the evaluation system, so that the algorithm model is more accurate.

7. The quantitative evaluation method of technological innovation according to claim 2, wherein the evaluation element index library in step S101 includes all innovation evaluation index elements, and the indexes collected by the index library are classified into different types by referring to the existing laboratory evaluation schemes and the evaluation schemes established by research institutions, and the method comprises the following steps: personnel information, papers, patents, monographs, standards, integrated circuit layout designs, scientific and technical rewards, research projects, result identification, result conversion, technological innovation development planning, software copyright, environment foundation, development communication and department platform information;

In the step S102, the evaluation system decomposes the evaluation target into three levels of index systems as follows: the scientific and technological innovation capability assessment index system of the scientific and technological platform is divided into 4 primary indexes, 12 secondary indexes and 37 tertiary indexes, wherein the 37 tertiary indexes are obtained by extraction and calculation according to the information of different index types.

8. The quantitative evaluation method of technological innovation according to claim 3, wherein in the step S201, the contribution rate of each index element is calculated by combining a principal component analysis method, the evaluation index is simplified, and an evaluation index system is constructed, wherein the principal component analysis method is utilized to reduce the dimension of the evaluation index, the important k indexes are extracted from q evaluation indexes of the n evaluation objects P ₁,P₂,P₃,…,P_n, k is < q, the correlation among the indexes is eliminated, the multiple index information is integrated into one comprehensive index value for evaluation, the statistical indexes of different innovation results and project results of the target object are converted into relative evaluation values instead of the simple combination of the indexes, and the specific steps of the principal component analysis are as follows:

(1) Constructing a sample matrix P _n×q:

P＝(p_ij)＝(p₁,p₂,…,p_q)

Wherein p _ij represents the original data of the j-th evaluation index in the evaluation object i, p _j represents the original dataset of the j-th evaluation index, i=1, 2, …, n, j=1, 2, …, q;

(2) The sample matrix is normalized to obtain a normalized matrix X _n×q:

X＝(X_ij)＝(X₁,X₂,…,X_p)

Wherein, Represents the average value of the j-th evaluation index, R _j represents the standard deviation of the j-th evaluation index,

(3) Calculate the correlation coefficient matrix CovX _q×q＝(c_ij):

Wherein c _ij denotes a sample correlation coefficient calculated based on the normalized data X _ij;

(4) Calculating a characteristic value alpha _i and a characteristic vector beta _i of the correlation coefficient matrix CovX:

|α_i·E-CovX|＝0

Wherein E is an identity matrix, α _i is a feature value, and the feature values are sorted by α ₁>α₂>…>α_q in descending order, and then the corresponding feature vectors β _i＝(β_1i,β_2i,…,β_qi)^T, i=1, 2, …, q are calculated for the sorted feature values α _i;

(5) Calculating the contribution rate of each evaluation index and the accumulated contribution rate:

Wherein U _k represents the variance contribution rate of the k index elements, U _k represents the accumulated contribution rate of the k index elements, k is less than or equal to q, and when U _k is more than or equal to delta, the k main index elements can fully reflect the main information of the original index system;

(6) Main index element model:

Wherein F _i represents the i-th main index element, F _i＝β_1i*X₁+β_2i*X₂+…+β_qi*X_q, i=1, 2,3, …, k, and the corresponding variance contribution ratio u _i is taken as the weight of each index element;

Step S202 is to objectively set the weight of the index elements { u ₁,u₂,u₃,…,u_k } based on the hierarchical analysis method on the basis of selecting the most important index elements from the index system by using the principal component analysis method, the hierarchical analysis method firstly carries out hierarchical structure division on the three-level evaluation index system, firstly establishes a hierarchical structure model, corresponds the first-level index to the target layer H, corresponds the second-level index to the criterion layers L1-L3 and corresponds the third-level index to the scheme layers Y1-Y6, then constructs a judgment matrix of each level, carries out pairwise comparison on the criterion layers L1, L2, … and L _k under the target layer H, and evaluates the level according to the importance degree thereof, and the weight calculation and consistency inspection process is as follows:

(1) Let M _k×k＝(m_ij) be the k-order evaluation matrix square matrix, where M _ij is the importance comparison result of the evaluation element i to the evaluation element j, i=1, 2, …, k, j=1, 2, …, k;

(2) Normalize each column vector of M to obtain a matrix V _k×k＝(v_ij), where

(3) Summing V by row yields h= (H ₁,h₂,h₃,…,h_k)^T, where

(4) Normalizing H to give d= (D ₁,d₂,d₃,…,d_k)^T, whereAfter normalization, carrying out hierarchical single sequencing, namely, the elements in D are sequencing weights of the relative importance of the same hierarchical element to the factors of the upper layer;

(5) Calculating γ= (m·d) _i/(k*d_i) as an approximation of the maximum eigenvalue, (m·d) _i represents the i-th component of m·d;

(6) Calculating a consistency index:

CI＝(γ-k)/(k-1)

when ci=0, it means perfect coincidence, the higher the coincidence is when CI is closer to 0, whereas the larger the CI is, the lower the coincidence is;

(7) Calculate the consistency ratio CR:

CR＝CI/RI

Wherein RI is introduced to measure the size of CI, RI represents an average random uniformity index, and when the uniformity ratio CR < τ, the inconsistency of M is generally considered acceptable within the allowable range, passing the uniformity test; otherwise, the comparison matrix M needs to be reconstructed, and M _ij is adjusted.

9. The quantitative evaluation method of technological innovation according to claim 4, wherein in the step S301, the data acquisition manner in laboratory units includes: the method comprises the steps of collecting data through two modes of online input and offline input, for the introduction of early historical data, enabling an evaluation system to meet a large-scale basic data batch introduction function, for the later data update, mainly performing data update by means of an evaluation object, collecting information by taking a laboratory as a unit through data reported by an individual, and establishing the relevance between the individual and the data through data collection;

in step S302, according to the entered information, the related content file stores the following data in units of departments: according to the entered information, the related content files are stored locally in different file formats by taking departments as units, and after various indexes are expected to be determined, the data are stored by utilizing a database;

In the step S304, the data verification of each item of entered information according to the "evaluation index element information base" includes: and (3) checking each item of entered data item by referring to an evaluation index element information base, checking and subtracting the data which is proved to be missing or invalid by quantity, and taking the final check data as an effective basis for calculating the score of each index.

10. The quantitative evaluation method of technological innovation according to claim 5, wherein the analyzing the evaluation index element in step S401 includes: analyzing the whole data index of a research institution, carrying out statistics and display on annual data, carrying out statistics and display on index elements of different categories, analyzing the data index of each evaluation object, calculating the contribution rate of each evaluation object to the total annual index of the research institution where the evaluation object is located, calculating the contribution rate of each evaluation object to the index of different categories, displaying the main contribution units of the index of different categories before ranking, analyzing the average contribution rate of various indexes, including average paper count and average patent count, comparing the two dimensions of historical data between each evaluation object and in the past, and obtaining the index element change of the research institution and each evaluation unit;

The evaluating the object for innovation ability analysis in step S402 includes: calculating and displaying the overall innovation score of the research institution and the innovation scores of all levels of evaluation indexes in a three-level index system according to the effective data values verified in S304, calculating and displaying the innovation scores of all levels of evaluation objects and the innovation scores of all levels of indexes of all evaluation objects, calculating and displaying the contribution degree of the innovation capacity of all evaluation objects to the innovation scores of the research institution, calculating and displaying the main contribution units of all levels of evaluation indexes, calculating and displaying the average innovation index I ₁,I₁ = innovation score/total number of all evaluation objects, considering the size of departments where the evaluation objects are located, calculating and displaying the talent innovation index I ₂ of all evaluation objects, considering the talent numbers of all levels of talents owned by the departments, I ₂ = innovation score/talent comprehensive score, wherein talent comprehensive = sigma (talent type number) of the research institution, calculating the average innovation score of the research institution, comparing the difference between all evaluation objects and the average score, finding out specific indexes of differences, screening out indexes required to be lifted by all evaluation objects, comparing the two dimensions between all evaluation objects and the historical data, and obtaining the change trend-improving indexes of all evaluation objects;

In the step S403, the score of each innovation body is calculated, and the sorting of the innovation capability of each evaluation object is achieved, including: after the information is input, checking the data integrity, calculating innovation ability values F of a research institution and each subordinate evaluation object according to the weight of each index in the step S201, and sequencing innovation ability of the evaluation objects according to the sequence from big to small;

in step S404, according to the obtained evaluation value, the time is subjected to statistical analysis, which specifically includes: after the information is input, checking the data integrity, calculating innovation capability values of research employment and subordinate evaluation objects according to S403, carrying out statistical analysis according to time, and acquiring technological innovation capability and change trend of the evaluated objects from transverse dimension and longitudinal dimension;

The predicting the innovation ability of the evaluation object by using multiple-sample multiple linear regression in step S405 includes: the method comprises the following specific steps of normalizing the original data through the minimum maximum value, eliminating the influence of different evaluation index score ranges on the result, selecting partial normalized data as a sample set S _m×n, carrying out innovation score prediction on a plurality of evaluation objects, and carrying out multiple sample multiple linear regression:

(1) Normalizing the original data through minimum and maximum normalization:

X^*＝(X-X_min)*(max-min)/(X_max-X_min)+min

Wherein, X ^* represents the data normalization result, X _min represents a row vector composed of minimum values in each column, X _max represents a row vector composed of maximum values in each column, and max and min represent the maximum value and the minimum value of the interval range of the data mapping respectively;

(2) Based on multiple sample multiple linear regression, selecting m samples after normalization processing, and n evaluation indexes to form a sample set S= [ S ₁,s₂,…,s_m]^T ], and then predicting the value

Wherein s _i represents the sample feature vector s _i＝[1,s_1i,s_2i,…,s_ni]^T (i=1, 2, …, m), the true value is denoted as y _m×1＝[y₁,y₂,…,y_m]^T, and the predicted value is set asThe weight parameter vector w= [ b, w ₁,w₂,…,w_n]^T, T represents the matrix transpose;

(3) The prediction loss is measured by adopting the mean square error, the prediction fitting degree is estimated, and the average value of the squares of the difference between the predicted value and the actual observed value is calculated to obtain