Quantitative measurement calculation method for interdisciplinary degree

By determining and calculating the interdisciplinary degree index scores of multiple evaluation indicator items, the problem of difficult to measure the interdisciplinary degree of scholars in the existing technology is solved, and scientific evaluation of the interdisciplinary degree of scholars is achieved, and scientific research development and talent training are promoted.

CN119578956APending Publication Date: 2025-03-07CHINA ACAD OF ART
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411450489.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-17
Publication Date
2025-03-07

AI Technical Summary

Technical Problem

It is difficult for existing technologies to effectively measure and evaluate the interdisciplinary level of scholars, affecting scientific research development and talent training.

Method used

By identifying multiple evaluation indicators, obtaining relevant data, calculating interdisciplinary level indicator scores, and calculating interdisciplinary level values ​​for scholars based on these scores. The specific methods include screening candidate index items, conducting correlation analysis, selecting indicators with high correlation, calculating the interdisciplinary level index scores for participating in academic conferences, publishing papers, academic collaborators, and citing literature, and finally weighted summing to obtain the interdisciplinary level value.

Benefits of technology

It provides a scientific method to directly evaluate the interdisciplinary level of scholars, help understand the interdisciplinary status of academic research, and provide reference for scientific research-oriented decision-making.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119578956A_ABST
    Figure CN119578956A_ABST
Patent Text Reader

Abstract

The invention discloses an interdisciplinary quantitative measurement calculation method. The method comprises the following steps: determining a plurality of evaluation index items; acquiring data of evaluation index items corresponding to the scholar to be evaluated; calculating interdisciplinary degree index scores of the plurality of evaluation index items through the obtained data of the evaluation index items; and calculating an interdisciplinary degree value of the to-be-evaluated scholar based on the interdisciplinary degree index scores of the plurality of evaluation index items. According to the quantitative measurement calculation method for the interdisciplinary degree, related calculation is carried out by scientifically selecting proper evaluation index items, and the evaluation value for directly evaluating the interdisciplinary degree of scholars is obtained.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention specifically relates to a quantitative measurement and calculation method for interdisciplinary degree. Background Art

[0002] With the development of disciplines, single-discipline research is prone to ceiling effects, prompting scholars to turn their attention to other disciplines, thereby drawing on useful ideas, theories and methods to find breakthroughs in scientific research. Modern academic research shows the characteristics of collectivism and interdisciplinary synergy, the diversification of research subjects and organizational diversity. As the country implements the innovation-driven development strategy, interdisciplinarity has become an inevitable trend in academic innovation.

[0003] Scholars are the main body of academic research. On the one hand, interdisciplinarity places higher demands on the scholars' own knowledge structure and scientific research capabilities. On the other hand, in the research process, they need to integrate the thinking, methods and knowledge of different fields and disciplines, and cooperate and communicate with researchers from other disciplines. Therefore, studying the degree of interdisciplinarity of scholars will help us understand the current interdisciplinarity status of Chinese academic research, the role of interdisciplinary scientific research in talent cultivation, and the correlation between high-level academic output and interdisciplinarity. Summary of the invention

[0004] The present invention provides a quantitative measurement and calculation method for interdisciplinary degree to solve the above-mentioned technical problems, and specifically adopts the following technical solutions:

[0005] A quantitative measurement and calculation method for interdisciplinary degree, including:

[0006] Determine multiple evaluation indicators;

[0007] Obtain the data of the evaluation index items corresponding to the scholars to be evaluated;

[0008] Calculate the interdisciplinary degree index scores of multiple evaluation index items through the acquired data of the evaluation index items;

[0009] The interdisciplinary level value of the scholar to be evaluated is calculated based on the interdisciplinary level indicator scores of multiple evaluation indicator items.

[0010] Furthermore, the specific method for determining the multiple indicator items is:

[0011] Preliminary screening of multiple candidate indicators;

[0012] Calculate multiple candidate index items through quantitative correlation analysis method;

[0013] According to the calculation results, candidate indicators with higher correlation are selected as evaluation indicators.

[0014] Furthermore, the specific method for determining the multiple evaluation index items is:

[0015] Select C scholars, and let the personnel related to the scholars score their interdisciplinary degree. Take the average score as the interdisciplinary degree survey score g of the scholar. i ;

[0016] Determine that there are a total of T candidate index items, 0 ≤ t < T. By collecting data and statistically calculating the values of the candidate index items, obtain the candidate index data of C scholars:

[0017] (x i1 , x i2 , x it … x iT )0 ≤ i < C

[0018] Among them, x it is the value of the t-th index of the i-th scholar;

[0019] For each candidate index item, calculate its correlation with the interdisciplinary degree survey score of the scholar. The specific calculation formula is as follows:

[0020]

[0021] Among them, cov(X t , g) represents the covariance between the t-th index and the interdisciplinary degree survey score of the scholar, std(X t ) and std(g) respectively represent the standard deviations of X t and g. The value range of R is (-1, 1). Select the candidate index items with R greater than the predetermined value to obtain T' candidate index items:

[0022] (x i1 , x i2 , x it … x iT′ )0 ≤ i < C

[0023] Calculate the correlation between the T' candidate index items. The formula is as follows:

[0024]

[0025] Among them, cov(X t′ , X t ) represents the covariance between the t'j-th index and the t-th index, std(X t′ ) and std(X t ) respectively represent the standard deviations of X t′ and X t . The value range of R t′t is (-1, 1). Select the candidate index items with R jt within the predetermined range as the final evaluation index items.

[0026] Further, candidate index items with R greater than 0.7 are selected to obtain T′ candidate index items;

[0027] Select R t′t The index item t′j and the index item t in the range of (-1, -0.7) and (0.7, 1) are the final evaluation index items.

[0028] Furthermore, the plurality of index items are academic conferences participated in, conference papers published, academic collaborators and cited literature;

[0029] Obtain data on academic conferences participated in, conference papers published, academic collaborators, and cited literature of the scholars to be evaluated;

[0030] Calculate the interdisciplinary degree index score corresponding to the academic conferences participated in;

[0031] Calculate the interdisciplinary degree index score corresponding to the published conference papers;

[0032] Calculate the interdisciplinary degree index scores corresponding to academic collaborators;

[0033] Calculate the interdisciplinary degree index score corresponding to the cited literature.

[0034] Furthermore, the specific method for calculating the interdisciplinary degree index score corresponding to the academic conference participated in is:

[0035] Obtain the papers published in the academic conferences you participated in in the past three years, and obtain the interdisciplinary distribution of the papers by analyzing all the keywords in the papers:

[0036] S ifk ={S1,S2…,S m}

[0037] Among them, k is the kth paper, i is the i-th scholar, f is conference f, and m is the number of disciplines. By counting all the papers of conference f, we can get the distribution of disciplines of conference f:

[0038] S if ={S1,S2…,S n}

[0039] And get the interdisciplinary number distribution D of all papers corresponding to conference f if ,as follows:

[0040] D if ={d if1 ,d if2 …d ifj …d ifn}

[0041] Where n is the number of disciplines associated with conference f, difj To count the number of times conference f in which scholar i participated appeared in the jth discipline;

[0042] According to D if Calculate the Brillouin index H if As a conference of interdisciplinary diversity, its formula is as follows:

[0043]

[0044] in,

[0045] Brillouin index H if Perform normalization:

[0046]

[0047] Among them, H ifmax The calculation is as follows:

[0048]

[0049] Wherein, c = int(N / n), d = mod(N,n), int is the integer function, mode is the modulo function;

[0050] According to D if Calculate the Gini coefficient B if As the interdisciplinary balance of the conference, the calculation formula is as follows:

[0051]

[0052] According to D if Calculate cosine similarity calculation O if As the interdisciplinary difference of the conference, the calculation formula is as follows:

[0053]

[0054] Among them, S ifjk is the cosine similarity between subject j and subject k, which is calculated as follows:

[0055] S ifjk =Cosine(V ifj ,V ifk )d ifj ≠0,d ifk ≠0

[0056] Among them, V ifj is d ifj Representative SubjectS j The word embedding vector, V ifk is d ifk Representative SubjectS kThe word embedding vector of

[0057] The degree of interdisciplinaryness of a conference is calculated by averaging the diversity, balance, and differences. The calculation formula is as follows:

[0058]

[0059] After calculating the interdisciplinary level of all academic conferences participated by the scholar to be evaluated, the average is added up to obtain the average interdisciplinary level of the academic conferences participated by the scholar, which is used as the interdisciplinary level index score of the scholar's academic conference.

[0060]

[0061] Among them, F is the total number of academic conferences.

[0062] Furthermore, the specific method for calculating the interdisciplinary degree index score corresponding to the published conference papers is as follows:

[0063] Analyze the keywords of conference papers published by scholars to obtain the subject distribution of all papers published by scholars:

[0064] S in ={S1,S2…,S n}

[0065] Count all conference papers published by the researcher and get the interdisciplinary distribution of all conference papers D p ,as follows:

[0066] D ip ={d i1 ,d i2 ……d in}

[0067] Among them, n is the number of disciplines related to conference papers, and i is the i-th scholar;

[0068] According to D p Calculate the Brillouin index H p As for the interdisciplinary diversity of conference papers, the formula is as follows:

[0069]

[0070] in,

[0071] The Brillouin index H p Perform normalization:

[0072]

[0073] According to D ip Calculate the Gini coefficient Bip As the interdisciplinary balance of conference papers, the calculation formula is as follows:

[0074]

[0075] According to D ip Calculate the cosine similarity O ip As the interdisciplinary diversity of conference papers, the calculation formula is as follows:

[0076]

[0077] Among them, S ijk is the cosine similarity between subject j and subject k, which is calculated as follows:

[0078] S ijk =Cosine(V ij ,V ik )d ij ≠0,d ik ≠0

[0079] Among them, V ij is d ij Representative SubjectS j The word embedding vector, V ik is d ik Representative SubjectS k The word embedding vector of

[0080] The interdisciplinary index score corresponding to the published conference papers is calculated by averaging the diversity, balance, and difference. The calculation formula is as follows:

[0081]

[0082] Furthermore, the specific method for calculating the interdisciplinary degree index score corresponding to the academic collaborator is:

[0083] Get the academic collaborators’ disciplines and majors, and get the discipline distribution of all collaborators:

[0084] S in ={S1,S2…,S n}

[0085] At the same time, we get the number distribution of disciplines and majors corresponding to all collaborators D ic ,as follows:

[0086] D ic ={d i1 ,d i2 ……d in}

[0087] Among them, n is the number of disciplines to which the academic collaborators belong, i is the i-th scholar,

[0088] According to D ic Calculate the Brillouin index H ic Interdisciplinary diversity as academic collaborators is formulated as follows:

[0089]

[0090] in,

[0091] Brillouin index H ic Perform normalization:

[0092]

[0093] Among them, H icmax The calculation is as follows:

[0094]

[0095] Among them, c = int (N / n), d = mod (N, n), int is the integer function, mode is the modulus function,

[0096] According to D ic Calculate the Gini coefficient B ic The interdisciplinary balance of academic collaborators is calculated as follows:

[0097]

[0098] According to D ic Calculate the cosine similarity O ic The interdisciplinary differences of academic collaborators are calculated as follows:

[0099]

[0100] Among them, S ijk is the cosine similarity between subject j and subject k, which is calculated as follows:

[0101] S ijk =Cosine(V ij ,V ik )d ij ≠0,d ik ≠0

[0102] Among them, V ij is d ij Representative SubjectS j The word embedding vector, V ik is d ik Representative SubjectS kThe word embedding vector of

[0103] The calculation method for the interdisciplinary degree index score corresponding to academic collaborators is to add up the diversity, balance, and difference. The calculation formula is as follows:

[0104]

[0105] Furthermore, the specific method for calculating the interdisciplinary degree index score corresponding to the cited literature is as follows:

[0106] Perform keyword analysis on the obtained reference set to obtain the subject distribution of cited papers:

[0107] S in ={S1,S2…,S n}

[0108] Combining all the cited literature sets, we can get the distribution of disciplines corresponding to all the cited literature D iq ,as follows:

[0109] D iq ={d i1 ,d i2 ……d in}

[0110] Among them, n is the number of disciplines associated with the cited literature, i is the i-th scholar,

[0111] According to D iq Calculate the Brillouin index H iq As the interdisciplinary diversity of citations, the formula is as follows:

[0112]

[0113] in,

[0114] The Brillouin index H iq Perform normalization:

[0115]

[0116] Among them, H iqmax The calculation is as follows:

[0117]

[0118] Among them, c = int (N / n), d = mod (N, n), int is the integer function, mode is the modulus function,

[0119] According to D iq Calculate the Gini coefficient B iqAs the interdisciplinary balance of cited literature, the calculation formula is as follows:

[0120]

[0121] According to D iq Calculate the cosine similarity O iq As the interdisciplinary difference of cited literature, the calculation formula is as follows:

[0122]

[0123] Among them, S ijk is the cosine similarity between subject j and subject k, which is calculated as follows:

[0124] S ijk =Cosine(V ij ,V ik )d ij ≠0,d ik ≠0

[0125] Among them, V ij is d ij Representative SubjectS j The word embedding vector, V ik is d ik Representative SubjectS k The word embedding vector of

[0126] The calculation method for the interdisciplinary degree index score corresponding to the cited literature is to add up the diversity, balance and difference. The calculation formula is as follows:

[0127]

[0128] Furthermore, the specific method for calculating the interdisciplinary degree value of the scholar to be evaluated based on the interdisciplinary degree index scores of multiple evaluation index items is:

[0129] The weight of each indicator is calculated through information gain entropy, and then the interdisciplinary degree value is obtained by weighted summing up all indicators according to the weight:

[0130] First calculate X ij The proportion of ij :

[0131]

[0132] Then calculate the entropy value e of each indicator j :

[0133]

[0134] Among them, C is the total number of scholars surveyed by this invention, and the weight of the index is finally calculated:

[0135]

[0136] After obtaining the indicator weights, the interdisciplinary degree value is obtained by weighted summing up the scores of each indicator item. The specific algorithm is as follows:

[0137] Calculate the weighted index:

[0138]

[0139] Score i It is a value between 0 and 1. The closer it is to 1, the higher the degree of interdisciplinarity of scholar i.

[0140] The benefit of the present invention lies in the quantitative measurement and calculation method of interdisciplinary degree provided, which obtains the evaluation value for directly evaluating the interdisciplinary degree of scholars by scientifically selecting appropriate evaluation index items and performing relevant calculations, thereby providing a scientific method for evaluating the interdisciplinary degree of scholars. BRIEF DESCRIPTION OF THE DRAWINGS

[0141] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0142] Figure 1 It is a schematic diagram of a quantitative measurement calculation method of interdisciplinary degree of the present invention. DETAILED DESCRIPTION

[0143] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0144] like Figure 1 The present invention shows a quantitative measurement and calculation method for interdisciplinary degree, comprising: S1: determining multiple evaluation index items. S2: obtaining data of evaluation index items corresponding to the scholars to be evaluated. S3: calculating the interdisciplinary degree index scores of multiple evaluation index items through the obtained data of evaluation index items. S4: calculating the interdisciplinary degree value of the scholars to be evaluated based on the interdisciplinary degree index scores of multiple evaluation index items. The above steps are described in detail below.

[0145] For step S1: determine multiple evaluation index items.

[0146] This application analyzes the degree of interdisciplinary of scholars, mainly based on Stirling's three-dimensional measure index of interdisciplinarity: diversity (the number of disciplines), balance (the degree of evenness of discipline distribution), and difference (the degree of difference in discipline nature). Among the three-dimensional measure indexes, diversity provides motivation for scholars' academic research, the difference between disciplines provides innovation elements for scholars, and balance makes academic research tend to be stable.

[0147] In the implementation manner of this application, the specific method for determining multiple index items is as follows:

[0148] Preliminarily screen multiple candidate index items.

[0149] Calculate multiple candidate index items through a quantitative correlation analysis method.

[0150] Select the candidate indexes with higher correlation as evaluation index items according to the calculation results.

[0151] Specifically, the specific method for determining multiple evaluation index items is as follows:

[0152] Select C scholars, and relevant personnel related to the scholars (such as students and colleagues) score their degree of interdisciplinarity. The scoring range is from 0 to 1, and the average score is taken as the survey score g of the scholars' degree of interdisciplinarity. i .

[0153] Determine that there are a total of T candidate index items, 0 ≤ t < T. By collecting data and statistically calculating the values of the candidate index items, the candidate index data of C scholars are obtained:

[0154] (x i1 ,x i2 ,x it …x iT )0 ≤ i < C

[0155] Among them, x it is the value of the t-th index of the i-th scholar.

[0156] For each candidate index item, calculate its correlation with the survey score of the scholars' degree of interdisciplinarity in order to select the most relevant index item as the final index item. The specific calculation formula is as follows:

[0157]

[0158] Among them, cov(X t ,g) represents the covariance between the t-th index and the survey score of the scholars' degree of interdisciplinarity, and std(X t ) and std(g) respectively represent X tThe standard deviation of and g, the value range of R is (-1,1), and the candidate index items whose R is greater than the predetermined value are selected to obtain T′ candidate index items:

[0159] (x i1 ,x i2 ,x it …x iT′ )0≤i <C

[0160] In the implementation manner of the present application, candidate index items with R greater than 0.7 are selected to obtain T′ candidate index items.

[0161] Since there are correlations between the index items, it is necessary to remove duplicates of the index items. The specific method is to calculate the correlation between T′ candidate index items. The formula is as follows:

[0162]

[0163] Among them, cov(X t′ ,X t ) represents the covariance between the t′jth index and the tth index, std(X t′ ) and std(X t ) represent X t′ and X t The standard deviation, R t′t The value range is (-1,1), select R t′t The candidate index items within the predetermined range are the final evaluation index items. Specifically, select R t′t The index item t′j and the index item t in the range of (-1, -0.7) and (0.7, 1) are the final evaluation index items.

[0164] In the implementation manner of the present application, the multiple indicator items are academic conferences participated in, conference papers published, academic collaborators and cited literature.

[0165] Specifically, academic conferences in which scholars participate include two types of conferences: one is conferences where scholars serve as organizers or reviewers, and the other is conferences where scholars publish papers. The degree of interdisciplinarity of conferences in which scholars serve as organizers or reviewers can, to a certain extent, reflect the academic interdisciplinarity of scholars. Scholars as organizers must have interdisciplinary resources and interdisciplinary knowledge to be able to play an organizational or planning role in such academic conferences. The degree of interdisciplinarity of conferences in which scholars publish papers can reflect the degree to which scholars’ academic achievements are accepted by interdisciplinary conferences. The higher the degree of interdisciplinarity of such conferences, the more interdisciplinarity of scholars’ academic achievements can be reflected from the side.

[0166] The published conference papers are the academic achievements of scholars, and their interdisciplinary level directly reflects the scholars' interdisciplinary academic output ability.

[0167] There are two types of academic collaborators: those who collaborate with scholars to publish academic results, and those who undertake scientific research projects with scholars. The distribution of academic collaborators in disciplines shows the breadth and diversity of scholars' interdisciplinary collaboration.

[0168] Cited literature refers to the literature cited in conference papers. The interdisciplinary nature of the cited literature indicates the ability of scholars to read and draw on interdisciplinary ideas, theories and methods.

[0169] For step S2: obtain the data of the evaluation index items corresponding to the scholar to be evaluated.

[0170] Specifically, data on academic conferences participated in, conference papers published, academic collaborators, and cited literature of the scholars to be evaluated are obtained.

[0171] The data for this application is mainly obtained from the personal homepages of scholars. The information on personal homepages is generally maintained by the scholars themselves. On the one hand, the data is more comprehensive and timely, and on the other hand, the data is more accurate.

[0172] To collect information about academic conferences, we first directly mine the conference information maintained on personal websites. Scholars mainly participate in the organization of conferences as editorial board members, chairmen, and editors-in-chief. Natural language processing is used to obtain conference names through deep learning models.

[0173] The second is the conference where scholars publish papers. Since the academic paper model maintained by personal websites is relatively standardized, the list of academic papers is obtained through keywords, and the conference name is obtained from the paper list.

[0174] Obtain academic conferences attended by scholars through the above two methods.

[0175] Data collection of published conference papers was obtained directly through the paper lists maintained on personal websites.

[0176] Academic collaborators are collected through personal websites. First, the published papers are obtained to obtain the list of authors of the papers and obtain the author names by segmentation; second, the author information listed in the published works;

[0177] Cited literature collection is obtained by obtaining the list of cited literature listed in published academic papers.

[0178] Construct a subject library and form a complete subject library S by collecting classification information of higher education subjects.

[0179] For step S3: the interdisciplinary degree index scores of the plurality of evaluation index items are calculated by using the acquired data of the evaluation index items.

[0180] Specifically, the interdisciplinary index scores for participating academic conferences, publishing conference papers, academic collaborators, and cited literature are calculated.

[0181] In the implementation mode of this application, the specific method for calculating the interdisciplinary degree index score corresponding to the academic conference participated in is:

[0182] The degree of interdisciplinarity of academic conferences in which scholars participate is obtained by statistically analyzing the interdisciplinarity of papers published in the academic conferences in the past three years.

[0183] Obtain the papers published in the academic conferences you participated in in the past three years.

[0184] Classify the papers by subject, extract the keywords of the papers, and input the keywords into the pre-trained RNN recurrent neural network of the subject classification;

[0185] The input is the jth keyword vector, and the output is a probability vector y of length n after the sofmax function. j , n is the number of subjects, and the subject to which keyword j belongs is:

[0186] S j =findSubject(S,y j )

[0187] Where findSubject is based on the subject library S and the probability vector yy j Obtain subject information.

[0188] By analyzing all the keywords in the paper, we can get the interdisciplinary distribution of the paper:

[0189] S ifk ={S1,S2…,S m}

[0190] Among them, k is the kth paper, i is the i-th scholar, f is conference f, and m is the number of disciplines. By counting all the papers of conference f, we can get the distribution of disciplines of conference f:

[0191] S if ={S1,S2…,S n}

[0192] And get the interdisciplinary number distribution D of all papers corresponding to conference f if ,as follows:

[0193] D if ={d if1 ,d if2 …d ifj …difn}

[0194] Where n is the number of disciplines associated with conference f, d ifj To count the number of times conference f in which scholar i participated appeared in the jth discipline.

[0195] According to D if Calculate the Brillouin index H if As the interdisciplinary diversity of the conference, the Brillouin index is a classic indicator for measuring diversity. It is based on the calculation principle of information entropy and its formula is as follows:

[0196]

[0197] in,

[0198] Since the Brillouin index does not have a value between 0 and 1, the Brillouin index H if Perform normalization:

[0199]

[0200] Among them, H ifmax The calculation is as follows:

[0201]

[0202] Among them, c=int(N / n), d=mod(N,n), int is the integer function, and mode is the modulo function.

[0203] According to D if Calculate the Gini coefficient B if As the interdisciplinary balance of the conference, the calculation formula is as follows:

[0204]

[0205] According to D if Calculate cosine similarity calculation O if As the interdisciplinary difference of the conference, the calculation formula is as follows:

[0206]

[0207] Among them, S ifjk is the cosine similarity between subject j and subject k, which is calculated as follows:

[0208] S ifjk =Cosine(V ifj ,V ifk )d ifj ≠0,d ifk ≠0

[0209] Among them, V ifj is d ifj Representative SubjectS j The word embedding vector, V ifk is d ifk Representative SubjectS k The word embedding vector of .

[0210] The degree of interdisciplinaryness of a conference is calculated by averaging the diversity, balance, and differences. The calculation formula is as follows:

[0211]

[0212] After calculating the interdisciplinary level of all academic conferences participated by the scholar to be evaluated, the average is added up to obtain the average interdisciplinary level of the academic conferences participated by the scholar, which is used as the interdisciplinary level index score of the scholar's academic conference.

[0213]

[0214] Among them, F is the total number of academic conferences.

[0215] In the implementation mode of the present application, the specific method for calculating the interdisciplinary degree index score corresponding to the published conference paper is:

[0216] Analyze the keywords of conference papers published by scholars to obtain the subject distribution of all papers published by scholars:

[0217] S in ={S1,S2…,S n}

[0218] Count all conference papers published by the researcher and get the interdisciplinary distribution of all conference papers D p ,as follows:

[0219] D ip ={d i1 ,d i2 ……d in}

[0220] Among them, n is the number of disciplines related to the conference papers, and i is the i-th scholar.

[0221] According to D p Calculate the Brillouin index H p As for the interdisciplinary diversity of conference papers, the formula is as follows:

[0222]

[0223] in,

[0224] The Brillouin index H p Perform normalization:

[0225]

[0226] According to D ip Calculate the Gini coefficient B ip As the interdisciplinary balance of conference papers, the calculation formula is as follows:

[0227]

[0228] According to D ip Calculate the cosine similarity O ip As the interdisciplinary diversity of conference papers, the calculation formula is as follows:

[0229]

[0230] Among them, S ijk is the cosine similarity between subject j and subject k, which is calculated as follows:

[0231] S ijk =Cosine(V ij ,V ik )d ij ≠0,d ik ≠0

[0232] Among them, V ij is d ij Representative SubjectS j The word embedding vector, V ik is d ik Representative SubjectS k The word embedding vector of

[0233] The interdisciplinary index score corresponding to the published conference papers is calculated by averaging the diversity, balance, and difference. The calculation formula is as follows:

[0234]

[0235] In the implementation mode of this application, the specific method for calculating the interdisciplinary degree index score corresponding to the academic collaborator is:

[0236] Get the academic collaborators’ disciplines and majors, and get the discipline distribution of all collaborators:

[0237] S in ={S1,S2…,S n}

[0238] At the same time, we get the number distribution of disciplines and majors corresponding to all collaborators D ic ,as follows:

[0239] D ic ={d i1 ,d i2 ……d in}

[0240] Among them, n is the number of disciplines to which the academic collaborators belong, i is the i-th scholar,

[0241] According to D ic Calculate the Brillouin index H ic Interdisciplinary diversity as academic collaborators is formulated as follows:

[0242]

[0243] in,

[0244] The Brillouin index H ic Perform normalization:

[0245]

[0246] Among them, H icmax The calculation is as follows:

[0247]

[0248] Among them, c = int (N / n), d = mod (N, n), int is the integer function, mode is the modulus function,

[0249] According to D ic Calculate the Gini coefficient B ic The interdisciplinary balance of academic collaborators is calculated as follows:

[0250]

[0251] According to D ic Calculate the cosine similarity O ic The interdisciplinary differences of academic collaborators are calculated as follows:

[0252]

[0253] Among them, S ijk is the cosine similarity between subject j and subject k, which is calculated as follows:

[0254] S ijk =Cosine(V ij ,V ik )d ij ≠0,d ik ≠0

[0255] Among them, Vij is d ij Representative SubjectS j The word embedding vector, V ik is d ik Representative SubjectS k The word embedding vector of

[0256] The calculation method for the interdisciplinary degree index score corresponding to academic collaborators is to add up the diversity, balance, and difference. The calculation formula is as follows:

[0257]

[0258] In the implementation mode of the present application, the specific method for calculating the interdisciplinary degree index score corresponding to the cited literature is:

[0259] Perform keyword analysis on the obtained reference set to obtain the subject distribution of cited papers:

[0260] S in ={S1,S2…,S n}

[0261] Combining all the cited literature sets, we can get the distribution of disciplines corresponding to all the cited literature D iq ,as follows:

[0262] D iq ={d i1 ,d i2 ……d in}

[0263] Among them, n is the number of disciplines associated with the cited literature, i is the i-th scholar,

[0264] According to D iq Calculate the Brillouin index H iq As the interdisciplinary diversity of citations, the formula is as follows:

[0265]

[0266] in,

[0267] The Brillouin index H iq Perform normalization:

[0268]

[0269] Among them, H iqmax The calculation is as follows:

[0270]

[0271] Among them, c = int (N / n), d = mod (N, n), int is the integer function, mode is the modulus function,

[0272] According to D iq Calculate the Gini coefficient B iq As the interdisciplinary balance of cited literature, the calculation formula is as follows:

[0273]

[0274] According to D iq Calculate the cosine similarity O iq As the interdisciplinary difference of cited literature, the calculation formula is as follows:

[0275]

[0276] Among them, S ijk is the cosine similarity between subject j and subject k, which is calculated as follows:

[0277] S ijk =Cosine(V ij ,V ik )d ij ≠0,d ik ≠0

[0278] Among them, V ij is d ij Representative SubjectS j The word embedding vector, V ik is d ik Representative SubjectS k The word embedding vector of

[0279] The calculation method for the interdisciplinary degree index score corresponding to the cited literature is to add up the diversity, balance and difference. The calculation formula is as follows:

[0280]

[0281] For step S4: the interdisciplinary level value of the scholar to be evaluated is calculated based on the interdisciplinary level indicator scores of multiple evaluation indicator items.

[0282] In the implementation manner of the present application, the specific method for calculating the interdisciplinary degree value of the scholar to be evaluated based on the interdisciplinary degree index scores of multiple evaluation index items is:

[0283] The weight of each indicator is calculated through information gain entropy, and then the interdisciplinary degree value is obtained by weighted summing up all indicators according to the weight:

[0284] First calculate X ij The proportion of ij :

[0285]

[0286] Then calculate the entropy value e of each indicator j :

[0287]

[0288] Among them, C is the total number of scholars surveyed by this invention, and the weight of the index is finally calculated:

[0289]

[0290] After obtaining the indicator weights, the interdisciplinary degree value is obtained by weighted summing up the scores of each indicator item. The specific algorithm is as follows:

[0291] Calculate the weighted index:

[0292]

[0293] Score i It is a value between 0 and 1. The closer it is to 1, the higher the interdisciplinary degree of scholar i. This interdisciplinary degree score can be used to analyze the current distribution of interdisciplinary degrees of Chinese scholars, and to guide policies by analyzing the current situation; by studying the correlation between the interdisciplinary degree of scholars and their academic ability and level, we can explore whether there is a correlation and provide a reference for scientific research-oriented decision-making.

[0294] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the present invention in any form, and any technical solution obtained by equivalent replacement or equivalent transformation falls within the protection scope of the present invention.

Claims

1. A quantitative measurement and calculation method for interdisciplinary degree, characterized by: Including: Determine multiple evaluation index items; Obtain the data of the evaluation index items corresponding to the scholar to be evaluated; Calculate the interdisciplinary degree index scores of multiple evaluation index items based on the obtained data of the evaluation index items; Calculate the interdisciplinary degree value of the scholar to be evaluated based on the interdisciplinary degree index scores of multiple evaluation index items.

2. The method for quantitatively measuring the interdisciplinary degree according to claim 1, wherein: The specific method for determining multiple index items is: Preliminarily screen multiple candidate index items; Calculate multiple candidate index items by means of quantitative correlation analysis; Select the candidate indexes with higher correlation as evaluation index items according to the calculation results.

3. The method for quantitatively measuring the interdisciplinary degree according to claim 2, wherein: The specific method for determining multiple evaluation index items is: Select C scholars, and have people related to the scholars rate their interdisciplinary level, and take the average score as the scholar's interdisciplinary level survey score g i ; It is determined that there are a total of T candidate index items, 0 ≤ t < T. By collecting data and statistically calculating the values of the candidate index items, the candidate index data of C scholars are obtained: (x i1 ,x i2 ,x it …x iT )0≤i<C Among them, x it is the value of the tth indicator of the ith scholar; For each candidate index item, calculate its correlation with the interdisciplinary degree survey score of the scholar. The specific calculation formula is as follows: Among them, cov(X t ,g) represents the covariance between the tth indicator and the interdisciplinary degree survey score of scholars, std(X t ) and std(g) represent X t The standard deviation of and g, the value range of R is (-1,1), and the candidate index items whose R is greater than the predetermined value are selected to obtain T′ candidate index items: (x i1 ,x i2 ,x it …x iT′ )0≤i<C Calculate the correlation between T' candidate index items. The formula is as follows: Among them, cov(X t′ ,X t ) represents the covariance between the t′jth index and the tth index, std(X t′ ) and std(X t ) represent X t′ and X t The standard deviation, R t′t The value range is (-1,1), select R t′t The candidate index items within the predetermined range are the final evaluation index items.

4. The method for quantitatively measuring the interdisciplinary degree according to claim 3, wherein: Select the candidate index items with R > 0.7 to obtain T' candidate index items; Select R t′t The index item t′j and the index item t in the range of (-1, -0.7) and (0.7, 1) are the final evaluation index items.

5. The method for quantitatively measuring the interdisciplinary degree according to claim 3, wherein: The multiple said index items are academic conferences participated in, conference papers published, academic collaborators, and cited documents; Obtain the data of the academic conferences participated in, conference papers published, academic collaborators, and cited documents corresponding to the scholar to be evaluated; Calculate the interdisciplinary degree index score corresponding to the academic conferences participated in; Calculate the interdisciplinary degree index score corresponding to the conference papers published; Calculate the interdisciplinary degree index score corresponding to the academic collaborators; Calculate the interdisciplinary degree index score corresponding to the cited documents.

6. The method for quantitatively measuring the interdisciplinary degree according to claim 5, wherein: The specific method for calculating the interdisciplinary degree index score corresponding to the academic conferences participated in is: Obtain the papers published in the academic conferences participated in in the past three years. By analyzing all the keywords of the papers, obtain the interdisciplinary type distribution of the papers: S ifk ={S1,S2…,S m } Where k is the kth paper, i is the ith scholar, f is the conference f, m is the number of disciplines. By statistically calculating all the papers of the conference f, obtain the discipline type distribution of the conference f: S if ={S1,S2…,S n } And get the interdisciplinary number distribution D of all papers corresponding to conference f if ,as follows: D if ={d if1 ,d if2 …d ifj …d ifn } Where n is the number of disciplines associated with conference f, d ifj To count the number of times conference f in which scholar i participated appeared in the jth discipline; According to D if Calculate the Brillouin index H if As a conference of interdisciplinary diversity, its formula is as follows: in, The Brillouin index H if Perform normalization: Among them, H ifmax The calculation is as follows: Where c = int(N / n), d = mod(N,n), int is the integer function, and mode is the modulo function; According to D if Calculate the Gini coefficient B if As the interdisciplinary balance of the conference, the calculation formula is as follows: According to D if Calculate cosine similarity calculation O if As the interdisciplinary difference of the conference, the calculation formula is as follows: Among them, S ifjk is the cosine similarity between subject j and subject k, which is calculated as follows: S ifjk =Cosine(V ifj ,V ifk )d ifj ≠0,d ifk ≠0 Among them, V ifj is d ifj Representative SubjectS j The word embedding vector, V ifk is d ifk Representative SubjectS k The word embedding vector of The calculation method of the interdisciplinary degree of the conference is to accumulate and average the diversity, balance, and difference. The calculation formula is as follows: After calculating the interdisciplinary degrees of all the academic conferences participated in by the scholar to be evaluated, accumulate and average them to obtain the average interdisciplinary degree of the academic conferences participated in by the scholar, which is used as the interdisciplinary degree index score of the scholar's academic conferences. Where F is the total number of academic conferences.

7. The method for quantitatively measuring the interdisciplinary degree according to claim 6, wherein: The specific method for calculating the interdisciplinary degree index score corresponding to the published conference papers is: Analyze the keywords of conference papers published by scholars to obtain the subject distribution of all papers published by scholars: S in ={S1,S2…,S n } Count all conference papers published by the researcher and get the interdisciplinary distribution of all conference papers D p ,as follows: D ip ={d i1 ,d i2 ……d in } Among them, n is the number of disciplines related to conference papers, and i is the i-th scholar; According to D p Calculate the Brillouin index H p As for the interdisciplinary diversity of conference papers, the formula is as follows: in, The Brillouin index H p Perform normalization: According to D ip Calculate the Gini coefficient B ip As the interdisciplinary balance of conference papers, the calculation formula is as follows: According to D ip Calculate the cosine similarity O ip As the interdisciplinary diversity of conference papers, the calculation formula is as follows: Among them, S ijk is the cosine similarity between subject j and subject k, which is calculated as follows: S ijk =Cosine(V ij ,V ik )d ij ≠0,d ik ≠0 Among them, V ij is d ij Representative SubjectS j The word embedding vector, V ik is d ik Representative SubjectS k The word embedding vector of The interdisciplinary index score corresponding to the published conference papers is calculated by averaging the diversity, balance, and difference. The calculation formula is as follows:

8. The method for quantitatively measuring interdisciplinary degree according to claim 7 is characterized in that: The specific method for calculating the interdisciplinary degree index score corresponding to academic collaborators is: Get the academic collaborators’ disciplines and majors, and get the discipline distribution of all collaborators: S in ={S1,S2…,S n } At the same time, we get the number distribution of disciplines and majors corresponding to all collaborators D ic ,as follows: D ic ={d i1 ,d i2 ……d in } Among them, n is the number of disciplines to which the academic collaborators belong, i is the i-th scholar, According to D ic Calculate the Brillouin index H ic Interdisciplinary diversity as academic collaborators is formulated as follows: in, The Brillouin index H ic Perform normalization: Among them, H icmax The calculation is as follows: Among them, c = int (N / n), d = mod (N, n), int is the integer function, mode is the modulus function, According to D ic Calculate the Gini coefficient B ic The interdisciplinary balance of academic collaborators is calculated as follows: According to D ic Calculate the cosine similarity O ic The interdisciplinary differences of academic collaborators are calculated as follows: Among them, S ijk is the cosine similarity between subject j and subject k, which is calculated as follows: S ijk =Cosine(V ij ,V ik )d ij ≠0,d ik ≠0 Among them, V ij is d ij Representative SubjectS j The word embedding vector, V ik is d ik Representative SubjectS k The word embedding vector of The calculation method for the interdisciplinary degree index score corresponding to academic collaborators is to add up the diversity, balance, and difference. The calculation formula is as follows:

9. The method for quantitatively measuring interdisciplinary degree according to claim 8, characterized in that: The specific method for calculating the interdisciplinary degree index score corresponding to the cited literature is: Perform keyword analysis on the obtained reference set to obtain the subject distribution of cited papers: S in ={S1,S2…,S n } Combining all the cited literature sets, we can get the distribution of disciplines corresponding to all the cited literature D iq ,as follows: D iq ={d i1 ,d i2 ……d in } Among them, n is the number of disciplines associated with the cited literature, i is the i-th scholar, According to D iq Calculate the Brillouin index H iq As the interdisciplinary diversity of citations, the formula is as follows: in, The Brillouin index H iq Perform normalization: Among them, H iqmax The calculation is as follows: Among them, c = int (N / n), d = mod (N, n), int is the integer function, mode is the modulus function, According to D iq Calculate the Gini coefficient B iq As the interdisciplinary balance of cited literature, the calculation formula is as follows: According to D iq Calculate the cosine similarity O iq As the interdisciplinary difference of cited literature, the calculation formula is as follows: Among them, S ijk is the cosine similarity between subject j and subject k, which is calculated as follows: S ijk =Cosine(V ij ,V ik )d ij ≠0,d ik ≠0 Among them, V ij is d ij Representative SubjectS j The word embedding vector, V ik is d ik Representative SubjectS k The word embedding vector of The calculation method for the interdisciplinary degree index score corresponding to the cited literature is to add up the diversity, balance and difference. The calculation formula is as follows:

10. The method for quantitatively measuring interdisciplinary degree according to claim 9, characterized in that: The specific method for calculating the interdisciplinary degree value of the scholar to be evaluated based on the interdisciplinary degree index scores of multiple evaluation index items is: The weight of each indicator is calculated through information gain entropy, and then the interdisciplinary degree value is obtained by weighted summing up all indicators according to the weight: First calculate X ij The proportion of ij : Then calculate the entropy value e of each indicator j : Among them, C is the total number of scholars surveyed by this invention, and the weight of the index is finally calculated: After obtaining the indicator weights, the interdisciplinary degree value is obtained by weighted summing up the scores of each indicator item. The specific algorithm is as follows: Calculate the weighted indicators: Score i It is a value between 0 and 1. The closer it is to 1, the higher the degree of interdisciplinarity of scholar i.