CN-122022112-A - Technological innovation talent flow analysis method and device based on multilayer heterogeneous network

CN122022112ACN 122022112 ACN122022112 ACN 122022112ACN-122022112-A

Abstract

The application discloses a scientific and technological innovation talent flow analysis method and device based on a multilayer heterogeneous network. The method comprises the steps of constructing a multi-layer heterogeneous network model based on talent flow data, carrying out node space-time distribution feature analysis based on node distribution features of the multi-layer heterogeneous network model to obtain node space-time distribution indexes, carrying out evolution measure feature analysis based on node distribution features and edge distribution features of the multi-layer heterogeneous network model to obtain evolution measure indexes, and carrying out technological innovation talent flow analysis based on the node space-time distribution indexes and the evolution measure indexes by adopting a preset correlation analysis model to obtain technological innovation talent flow analysis results. The method can improve the accuracy of talent flow analysis of technological innovation, and lays a foundation for formulating talent strategies and optimizing talent resource allocation.

Inventors

YE ANQI
PAN XIAOQING
HU DANNI
SHEN DAYONG
QIAO FENGCAI
Ding Zizhuo
PAN LIN
FU WENXUAN
LI BO
YAO FENG
LUO YUN
ZHANG ZHONGSHAN
LIN XIANG
Gong Haochen

Assignees

中国人民解放军国防科技大学

Dates

Publication Date: 20260512
Application Date: 20251204

Claims (10)

1. A technological innovation talent flow analysis method based on a multilayer heterogeneous network is characterized by comprising the following steps: constructing a multi-layer heterogeneous network model based on talent flow data; carrying out node space-time distribution characteristic analysis based on the node distribution characteristics of the multi-layer heterogeneous network model to obtain node space-time distribution indexes; performing evolution measure feature analysis based on node distribution features and edge distribution features of the multi-layer heterogeneous network model to obtain an evolution measure index; and carrying out technological innovation talent flow analysis by adopting a preset correlation analysis model based on the node space-time distribution index and the evolution measure index to obtain technological innovation talent flow analysis results.
2. The method of claim 1, wherein the constructing a multi-layer heterogeneous network model based on talent flow data specifically comprises: carrying out data conversion on talent flow data to obtain point elements and edge elements of the multilayer heterogeneous network; And constructing models for different network layers by utilizing a target graph data structure according to the point elements and the edge elements to obtain the multi-layer heterogeneous network model.
3. The method of claim 1, wherein the node spatio-temporal distribution feature analysis is performed based on the node distribution features of the multi-layer heterogeneous network model to obtain a node spatio-temporal distribution index, and specifically comprises: Calculating based on the first talent number of the unit area of the evaluation area of the multi-layer heterogeneous network model and the first area of the evaluation area to obtain talent density indexes; calculating based on the talent density index, the total talent number of the multi-layer heterogeneous network model and the total evaluation area of the multi-layer heterogeneous network model to obtain a talent aggregation index; calculating based on the first talent number, the first area, the total talent number and the total evaluation area to obtain a uniform index; screening the organization nodes of different evaluation areas in the node distribution characteristics of the multi-layer heterogeneous network model, and calculating based on the number of the organization nodes of the screened different evaluation areas to obtain a first index; Calculating based on the total talent number of different evaluation areas, the bit sequence of the evaluation areas and the bit sequence scale dimension number in the node distribution characteristics of the multi-layer heterogeneous network model to obtain bit sequence scale indexes; Calculating based on the second talent number of the evaluation area of the unit area in the node distribution characteristics of the multi-layer heterogeneous network model to obtain a specific gravity index; calculating based on the area number of different evaluation areas, the average talent number of different evaluation areas and the second talent number of different evaluation areas to obtain a coefficient index; The node space-time distribution characteristics comprise talent density index, talent concentration index, uniformity index, first position index, bit sequence scale index, specific gravity index and coefficient of foundation index.
4. The method of claim 1, wherein the performing evolution measure feature analysis based on the node distribution feature and the edge distribution feature of the multi-layer heterogeneous network model to obtain an evolution measure index specifically comprises: Extracting key nodes of the multi-layer heterogeneous network model; performing evaluation index calculation according to the key nodes to obtain key node measurement indexes, wherein the key node measurement indexes comprise a degree centrality index, a medium centrality index and a PageRank index; Performing evaluation index calculation based on the edge distribution characteristics of the multi-layer heterogeneous network model to obtain talent flow tendency measurement indexes, wherein the flow tendency measurement indexes comprise clustering coefficient indexes and co-allocation coefficient indexes; according to node distribution characteristics and edge distribution characteristics in the multi-layer heterogeneous network model, a preset intervention opportunity model is adopted to perform evaluation index calculation, and talent flow gravitation measurement indexes are obtained, wherein the talent flow gravitation measurement indexes comprise talent passing rate indexes and talent attractive indexes; The evolution measure index comprises a key node measure index, a talent flow tendency measure index and a talent flow attraction measure index.
5. The method of claim 4, wherein the calculating the evaluation index according to the key node to obtain the key node measure index specifically comprises: Respectively calculating an ingress parameter and an egress parameter of the key node based on the adjacency matrix of the key node to obtain a degree centrality index of the key node; calculating according to the first shortest path number between any two key nodes and the second shortest path number passing through any target key node between the two key nodes to obtain a medium number centrality index of the target key node; and performing evaluation index calculation on the key nodes by adopting a random walk model formed by a follow-up link propagation network and a random jump fairness network to obtain PageRank index indexes of the key nodes at different moments.
6. The method of claim 4, wherein the computing the evaluation index based on the edge distribution characteristics of the multi-layer heterogeneous network model to obtain the talent flow tendency measure index specifically comprises: Performing evaluation index calculation based on the actual edge number between the current node element and the neighbor node element number to obtain a clustering coefficient index; And performing evaluation index calculation based on the joint probability distribution of any two key nodes and the redundancy distribution of each key node to obtain a homoleptic coefficient index.
7. The method of claim 4, wherein the calculating the evaluation index by using a preset intervention opportunity model according to the node distribution characteristics and the edge distribution characteristics in the multi-layer heterogeneous network model to obtain the talent flow gravitation measurement index specifically comprises: performing evaluation index calculation according to the third talent number and the total flow talent number flowing through any two key nodes to obtain a global talent flow index; Performing evaluation index calculation according to the third talent number and the total talent number of the multi-layer heterogeneous network model to obtain a local talent flow index; Aiming at any technological innovation talents, adopting the preset intervention opportunity model fusing a radiation opportunity model and an exploration opportunity model to make a decision so as to obtain attractive indexes of a target place to the technological innovation talents relative to an initial place; The talent passing rate index comprises the global talent flow index and the local talent flow index.
8. The method of claim 1, wherein the performing the scientific and innovative talent flow analysis based on the node spatiotemporal distribution index and the evolution measure index using a preset correlation analysis model to obtain a scientific and innovative talent flow analysis result specifically comprises: preprocessing the node space-time distribution index and the evolution measure index to obtain a first standard characteristic index corresponding to the node space-time distribution index and a second standard characteristic index corresponding to the evolution measure index; Splicing the first standard characteristic index and the second standard characteristic index to obtain characteristic vectors of all nodes; And carrying out correlation analysis on talent flow results by adopting the preset correlation analysis model based on each feature vector to obtain technological innovation talent flow analysis results.
9. Technological innovation talent flow analysis device based on multilayer heterogeneous network, characterized by comprising: The construction module is used for constructing a multi-layer heterogeneous network model based on talent flow data; the first analysis module is used for carrying out node space-time distribution characteristic analysis based on the node distribution characteristics of the multi-layer heterogeneous network model to obtain node space-time distribution indexes; The second analysis module is used for carrying out evolution measure feature analysis based on the node distribution features and the edge distribution features of the multi-layer heterogeneous network model to obtain an evolution measure index; and the association analysis module is used for carrying out technological innovation talent flow analysis by adopting a preset correlation analysis model based on the node space-time distribution index and the evolution measure index to obtain technological innovation talent flow analysis results.
10. An electronic device comprising at least a memory, a processor, said memory having stored thereon a computer program, said processor, when executing said computer program on said memory, implementing the steps of the innovative talent flow analysis method based on a multi-layer heterogeneous network according to any of the preceding claims 1-8.

Description

Technological innovation talent flow analysis method and device based on multilayer heterogeneous network Technical Field The invention relates to the field of large data analysis of scientific and technological innovation talents and complex network modeling, in particular to a method and a device for flow analysis of the scientific and technological innovation talents based on a multi-layer heterogeneous network. Background The current technological innovation talent flow has the technical defect that the existing method can only process single type data (such as treatise treatises or patent cooperations) and cannot synchronously integrate multidimensional flow tracks of academia, technology, occupation and the like. Traditional static network models are difficult to describe dynamic characteristics of talent flow evolution along with time (such as staged outbreaks of cross-regional flow), a main stream analysis tool does not consider cross-level coupling effects of a macroscopic policy layer (such as talent introduction plan) and a microscopic individual layer (such as occupational selection), and the computational complexity is exponentially increased due to data sparsity by the traditional heterogeneous network construction method (such as a splicing technology based on an adjacency matrix). Disclosure of Invention In view of the above, the invention provides a technology innovation talent flow analysis method and device based on a multi-layer heterogeneous network and electronic equipment, and mainly aims to solve the problem that technology innovation talent flow analysis is inaccurate at present. In order to solve the above problems, the present application provides a technological innovation talent flow analysis method based on a multi-layer heterogeneous network, comprising: constructing a multi-layer heterogeneous network model based on talent flow data; carrying out node space-time distribution characteristic analysis based on the node distribution characteristics of the multi-layer heterogeneous network model to obtain node space-time distribution indexes; performing evolution measure feature analysis based on node distribution features and edge distribution features of the multi-layer heterogeneous network model to obtain an evolution measure index; and carrying out technological innovation talent flow analysis by adopting a preset correlation analysis model based on the node space-time distribution index and the evolution measure index to obtain technological innovation talent flow analysis results. Optionally, the constructing a multi-layer heterogeneous network model based on talent flow data specifically includes: carrying out data conversion on talent flow data to obtain point elements and edge elements of the multilayer heterogeneous network; And constructing models for different network layers by utilizing a target graph data structure according to the point elements and the edge elements to obtain the multi-layer heterogeneous network model. Optionally, the node space-time distribution feature analysis is performed based on the node distribution feature of the multi-layer heterogeneous network model to obtain a node space-time distribution index, which specifically includes: Calculating based on the first talent number of the unit area of the evaluation area of the multi-layer heterogeneous network model and the first area of the evaluation area to obtain talent density indexes; calculating based on the talent density index, the total talent number of the multi-layer heterogeneous network model and the total evaluation area of the multi-layer heterogeneous network model to obtain a talent aggregation index; calculating based on the first talent number, the first area, the total talent number and the total evaluation area to obtain a uniform index; screening the organization nodes of different evaluation areas in the node distribution characteristics of the multi-layer heterogeneous network model, and calculating based on the number of the organization nodes of the screened different evaluation areas to obtain a first index; Calculating based on the total talent number of different evaluation areas, the bit sequence of the evaluation areas and the bit sequence scale dimension number in the node distribution characteristics of the multi-layer heterogeneous network model to obtain bit sequence scale indexes; Calculating based on the second talent number of the evaluation area of the unit area in the node distribution characteristics of the multi-layer heterogeneous network model to obtain a specific gravity index; calculating based on the area number of different evaluation areas, the average talent number of different evaluation areas and the second talent number of different evaluation areas to obtain a coefficient index; The node space-time distribution characteristics comprise talent density index, talent concentration index, uniformity index, first position index, bit sequence scale