An innovation network dynamic evolution analysis method based on time slicing and collaborative change

Abstract

The application relates to the technical field of innovative network analysis, in particular to an innovative network dynamic evolution analysis method based on time slicing and collaborative change, which comprises the following steps: S1, innovative subject classification; S2, time slicing network construction; S3, network structure index calculation; S4, collaborative change measurement; S5, evolution mode identification and risk early warning. By constructing a multi-dimensional innovative subject classification and scoring model, the application realizes objective and quantitative division of exploratory and developmental innovation institutions, and avoids the deviation of subjective qualitative classification. By adopting time slicing technology to construct a dynamic innovation network sequence and introducing a weighted adjacency matrix to fully consider the cooperation strength difference, the application can accurately capture the space-time evolution characteristics of the innovation network. The application proposes special measurement indexes such as a cross-type connection ratio, a collaborative change index and a balance index, and quantifies the interaction strength and collaborative evolution state of the two types of innovative subjects.

Description

Technical Field

[0001] This invention relates to the field of innovation network analysis technology, specifically to a method for dynamic evolution analysis of innovation networks based on time slicing and collaborative changes. Background Technology

[0002] In recent years, with the development of innovation system theory and social network analysis methods, innovation network research has gradually expanded from static structural analysis to dynamic evolutionary analysis. Traditional innovation network research mainly uses cross-sectional data for static analysis, focusing on network topological characteristics at a certain moment, such as network density, centrality distribution, and community structure. However, innovation networks are essentially complex systems that evolve dynamically, and the cooperative relationships among innovation agents change constantly over time. Static analysis is insufficient to capture the dynamic characteristics and inherent laws of network evolution.

[0003] In existing technologies, some studies have begun to use the time-slicing method to construct dynamic innovation networks. This involves dividing the research period into multiple time windows, constructing a static network within each window, and then analyzing the temporal evolution trend of the network structure. However, existing technologies have the following shortcomings: First, the classification of innovation entities lacks a unified quantitative standard and mostly adopts subjective qualitative classification, making it difficult to accurately distinguish between exploratory and developmental innovation institutions. Then, the weight differences in cooperation strength were not fully considered during the construction of the dynamic network, which led to biases in the network structure analysis. Summary of the Invention

[0004] The purpose of this invention is to provide an innovative network dynamic evolution analysis method based on time slicing and cooperative changes, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: An innovative network dynamic evolution analysis method based on time slicing and co-variation includes the following steps: S1. Classification of Innovation Entities: Based on multi-dimensional features, a classification and scoring model for innovation entities is constructed, dividing each innovation entity into exploratory innovation institutions and developmental innovation institutions; S2. Construction of Time Slice Network: Divide the innovation cooperation data into slices according to a preset time interval. At each time slice, construct a static innovation network corresponding to the time point with the innovation subject as the node and the cooperation relationship as the edge. S3. Calculation of network structure indicators: For each time slice of the innovative network, calculate the network density, node centrality, clustering coefficient, modularity and structural hole index. S4. Measurement of Collaborative Change: Construct an exploratory-development collaborative change measurement model to quantify the changes in connection strength and interactive balance between the two types of innovation subjects in different time slices; S5. Evolutionary Pattern Identification and Risk Warning: Based on time series analysis, identify the evolutionary patterns of innovation networks, construct path locking risk index and bubble risk index, and realize dynamic assessment and early warning of the sustainability of innovation networks.

[0006] Preferably, in step S1, the process of constructing an innovation entity classification and scoring model based on multi-dimensional features to divide each innovation entity into exploratory innovation institutions and developmental innovation institutions is as follows: Ten dimensions were selected as classification features: patent type, R&D investment intensity, technological novelty, venture capital ratio, knowledge flow pattern, talent mobility rate, degree of open innovation, proportion of basic research, organizational level and incentive mechanism. Normalize the features of each dimension and calculate the feature values ​​of each innovation subject in each dimension. Construct a classification scoring function, assign weights to each dimension, and calculate the comprehensive classification score S; The formula for constructing the classification scoring function, assigning weights to each dimension, and calculating the comprehensive classification score S is as follows: in, The proportion of basic patents, For a gradual increase in patent ratio, The ratio of R&D investment to revenue; For the proportion of interdisciplinary patents, For the proportion of venture capital, External knowledge citation rate; To reduce the inter-agency mobility of R&D personnel, The proportion of external cooperation projects, The percentage of basic research expenditure; For the number of organizational levels, The proportion of employee innovation rewards; to These are the weight coefficients corresponding to each dimension; Set a classification threshold: if S>0, it is judged as an exploratory innovation institution; if S≤0, it is judged as a developmental innovation institution.

[0007] Preferably, in S2, the construction of the time-sliced ​​network, the process of dividing the innovation cooperation data into slices according to a preset time interval, and constructing a static innovation network at each time segment with innovation entities as nodes and cooperation relationships as edges, is as follows: Collect data on innovative collaborations, including joint patents, joint research projects, joint papers, and strategic cooperation data, with each data point marked with a precise timestamp; The research period is divided into T consecutive time slices based on annual or quarterly time intervals, denoted as T0. ; For each time slice Extract all cooperative relationships within this time period and construct an adjacency matrix. In the process, edge weight calculation is introduced. The calculation formula is: in, Time slice inner subject and Number of collaborations and The main body and Total number of collaborations within that time slice; in Representing the subject With the main body If cooperation exists during this time period, then the value is 0; otherwise, it is 0. Each node is labeled with an innovation type tag, marked as exploratory or developmental, forming a multi-time-section innovation network sequence with node attributes.

[0008] Preferably, in S3, the calculation of network structure indicators, the process of calculating network density, node centrality, clustering coefficient, modularity, and structural hole index for the innovative network under each time slice is as follows: Calculate network density It measures the overall connectivity of the network; Calculate network density The calculation formula is: in, Time slice The actual number of edges in the network. This represents the total number of nodes in that time slice. Calculate the degree centrality of each node. Betweenness centrality With proximity centrality Analyze the status and influence of nodes in the network; The betweenness centrality of each node is calculated. The calculation formula is: in, The total number of shortest paths from node s to node t. For the nodes The number of shortest paths; Calculate the average clustering coefficient of the network It measures the connectivity between a node's neighbors; The Louvain algorithm is used for community detection, and the modularity index is calculated. Assess the salience of network community structure; Calculate the structural hole constraint index for each node. Identify key nodes that connect different groups.

[0009] Preferably, in S4, the process of constructing an exploratory-developmental collaborative change measurement model to quantify the changes in connection strength and interactive equilibrium between the two types of innovation subjects across different time slices is as follows: Count the number of connections between exploratory innovation entities within each time slice. Number of connections between development-oriented innovation entities and cross-type connection count ; Calculate cross-type connection ratio This measures the degree of integration between the two types of entities. Constructing a Cooperative Change Index Quantify the intensity of interaction between exploratory and developmental innovation entities; Calculate the balance index To assess the relative positions and developmental balance of the two types of entities in the network; Based on the index differences between adjacent time slices, the rate of coordinated change is calculated to identify abrupt changes in the interaction pattern.

[0010] Preferably, the calculation of cross-type connection ratio The calculation formula is: The construction of the collaborative change index The calculation formula is: The calculation of the balance index The calculation formula is: in, To increase the number of exploratory innovation entities, The number of entities engaged in development-oriented innovation.

[0011] Preferably, in S5, evolutionary pattern identification and risk warning, the process of identifying innovation network evolutionary patterns based on time series analysis, constructing path-locking risk indices and bubble risk indices, and realizing dynamic assessment and early warning of innovation network sustainability is as follows: Time series analysis was performed on the network indicators for each time slice, and the trend of indicator changes was calculated using the sliding window method. Construct a path-locking risk index The calculation is based on a weighted average of the concentration, internal connectivity density, and centrality advantage of developmental innovation entities. Constructing a bubble risk index The calculation is based on the isolation degree of exploratory innovation subjects, insufficient cross-type connections, and expansion speed. Set a risk warning threshold, and trigger a warning signal of the corresponding level when the risk index exceeds the threshold; Based on historical evolution data, a time series forecasting model is used to predict future risk trends.

[0012] Preferably, the construction path locking risk index The calculation formula is: in, and The average degree centrality of developmental and exploratory subjects, respectively. and The internal network density of the two types of subjects are respectively. , , These are the weighting coefficients.

[0013] An innovative network dynamic evolution analysis system based on time slicing and co-variation includes: The innovation entity classification module is used to construct an innovation entity classification and scoring model based on multi-dimensional features, dividing each innovation entity into exploratory innovation institutions and development innovation institutions; The time-slice network construction module is used to divide innovation cooperation data into slices according to preset time intervals. At each time slice, the innovation subject is used as a node and the cooperation relationship is used as an edge to construct a static innovation network at the corresponding time. The network structure index calculation module is used to calculate network density, node centrality, clustering coefficient, modularity, and structural hole index for innovative networks under each time slice. The collaborative change measurement module is used to construct an exploratory-development collaborative change measurement model to quantify the changes in connection strength and interaction balance between the two types of innovation subjects across time slices. The evolution pattern identification and risk warning module is used to identify the evolution patterns of innovation networks based on time series analysis, construct path locking risk index and bubble risk index, and realize dynamic assessment and early warning of the sustainability of innovation networks.

[0014] Compared with the prior art, the beneficial effects of the present invention are: This invention constructs a multi-dimensional classification and scoring model for innovation entities, achieving an objective and quantitative division between exploratory and developmental innovation institutions, avoiding the biases of subjective qualitative classification. It employs time-slicing technology to construct dynamic innovation network sequences and introduces a weighted adjacency matrix to fully consider differences in cooperation intensity, accurately capturing the spatiotemporal evolution characteristics of the innovation network. It proposes specialized metrics such as cross-type connection ratio, synergistic change index, and balance index, systematically quantifying the interaction intensity and synergistic evolution status of the two types of innovation entities. Furthermore, it constructs path-locking risk and bubble risk indices, establishing a quantitative assessment and early warning mechanism for innovation sustainability, enabling timely identification of potential risks in the evolution of the innovation network. This invention's method can provide scientific quantitative decision support for regional innovation policy formulation, industrial cluster development planning, and innovation ecosystem optimization. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the overall process of the innovative network dynamic evolution analysis method based on time slicing and collaborative changes of the present invention; Figure 2 This is a schematic diagram of the module architecture of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Please see Figure 1 As shown, this invention is a method for dynamic evolution analysis of innovative networks based on time slicing and cooperative changes, comprising the following steps: S1. Classification of Innovation Entities: Based on multi-dimensional characteristics, a classification and scoring model for innovation entities is constructed, dividing each innovation entity into exploratory innovation institutions and developmental innovation institutions. The specific implementation process is as follows: First, ten dimensions were selected as classification features, including: patent type dimension (basic patent ratio). Gradual patent ratio ), R&D investment dimension (R&D investment as a percentage of revenue) ), Technological novelty dimension (proportion of interdisciplinary patents) ), funding source dimension (proportion of venture capital) ), knowledge flow dimension (external knowledge citation rate) ), talent mobility dimension (inter-institutional turnover rate of R&D personnel) external cooperation project ratio ), Research type dimension (basic research expenditure ratio) Organizational structure dimension (number of organizational levels) ), incentive mechanism dimension (employee innovation reward ratio) ).

[0018] Secondly, each feature dimension is normalized, mapping all feature values ​​to the [0, 1] interval to eliminate dimensional differences. For the ... The first subject The feature is given by the normalization formula: Then, a classification scoring function is constructed, and the weights of each dimension are determined using the Analytic Hierarchy Process (AHP). to Calculate the comprehensive classification score The calculation formula is: in, The proportion of basic patents, For a gradual increase in patent ratio, The ratio of R&D investment to revenue; For the proportion of interdisciplinary patents, For the proportion of venture capital, External knowledge citation rate; To reduce the inter-agency mobility of R&D personnel, The proportion of external cooperation projects, The percentage of basic research expenditure; For the number of organizational levels, The proportion of employee innovation rewards; to These are the weight coefficients for each dimension; Finally, a classification threshold of 0 was set. If S>0, the organization was identified as an exploratory innovation institution, and if S≤0, it was identified as a developmental innovation institution. Exploratory innovation institutions are characterized by a high proportion of basic patents, high R&D investment, high interdisciplinary innovation, high reliance on venture capital, high external knowledge citation, high talent mobility, high external cooperation, high investment in basic research, flat organization, and strong innovation incentives. Developmental innovation institutions, on the other hand, exhibit the opposite characteristics.

[0019] S2. Time-slice network construction: The innovation collaboration data is divided into slices according to preset time intervals. At each time slice, a static innovation network is constructed with the innovation subject as the node and the collaboration relationship as the edge. The specific implementation process is as follows: First, we collect multi-source innovation collaboration data, including: joint patent application data from the State Intellectual Property Office (including application number, applicant, application year, and IPC classification number), joint research project data from the Ministry of Science and Technology (including project number, undertaking unit, and year of project approval), joint paper data from Web of Science (including DOI, author affiliation, and year of publication), and strategic cooperation announcement data from enterprises (including cooperating parties and signing time). Each data entry is recorded with precise timestamp information.

[0020] Secondly, the research period (e.g., 2010-2025) is divided into 16 consecutive time slices (T=16) based on annual time intervals, denoted as... Each time slice corresponds to a natural year. For analysis scenarios that require more granular time, quarterly or semi-annual time intervals can be used.

[0021] Then, for each time slice Extract all cooperative relationships within this time period and construct a weighted adjacency matrix. For any two innovation entities and edge weight Calculation using normalized number of collaborations: in, Time slice inner subject and Number of collaborations and The main body and The total number of collaborations within this time slice; this normalization method eliminates the influence of the subject's collaboration activity on the edge weights, and more accurately reflects the relative strength of collaborations.

[0022] Finally, each node is labeled with an innovation type tag (exploratory E or developmental D) to form a multi-time-section innovation network sequence with node attributes. ,in , For a set of nodes, Let be the set of edges. This is the weight matrix. It is a collection of node labels.

[0023] S3. In the calculation of network structure indicators, for each time slice of the innovative network, the network density, node centrality, clustering coefficient, modularity, and structural hole index are calculated. The specific implementation process is as follows: S31. Network density calculation: Network density Measure the overall connectivity of a network and calculate network density. The calculation formula is: in, Time slice The actual number of edges in the network. This represents the total number of nodes in this time slice; the network density ranges from [0, 1], with a larger value indicating a denser network connection.

[0024] S32, Node Centrality Calculation: Degree centrality :node The ratio of the number of connections to the maximum possible number of connections is given by the formula: Betweenness centrality The formula for measuring the importance of a node as a network bridge is: in, The total number of shortest paths from node s to node t. For the nodes The number of shortest paths.

[0025] Proximity centrality The inverse of the average shortest distance from a node to all other nodes, expressed by the formula: in For nodes arrive The shortest path length.

[0026] S33. Clustering Coefficient Calculation: Average Clustering Coefficient of the Network Measuring connectivity between a node's neighbors reflects the degree of network aggregation; nodes The local clustering coefficient is: in, For nodes The degree, For nodes The actual number of edges between neighbors; the average clustering coefficient of the network is the average of the clustering coefficients of all nodes: S34. Modular Computation: The Louvain algorithm is used for community detection by maximizing the modularity function. Identifying the structure of online communities: in, The total number of edges in the network. For nodes The community to which it belongs For indicator functions, when The value is 1 if it is true, and 0 otherwise. The value ranges from [-1, 1], and the larger the value, the more significant the community structure.

[0027] S35. Structural Void Calculation: Calculate the structural void constraint index for each node. This measures the redundancy of nodes in the network. in, For nodes The neighborhood group, For nodes Deploy to nodes Relationship ratio: For weighted networks: For unauthorized networks: The smaller the constraint index, the more structural holes the node occupies, and the stronger its ability to connect different groups.

[0028] S4. In the measurement of collaborative change, an exploratory-developmental collaborative change measurement model is constructed to quantify the changes in the connection strength and interactive balance between the two types of innovation subjects in each time slice. The specific implementation process is as follows: First, the edges are divided into three categories based on node labels: exploratory inter-subject connections. , Development-oriented inter-subject connections Cross-type connections The number and total weight of each type of connection are counted separately.

[0029] Secondly, calculate the cross-type connection ratio. This measures the degree of integration between the two types of entities. Cross-type connection ratio The calculation formula is: The value ranges from [0, 1]. The larger the value, the more frequent the cross-type cooperation and the higher the degree of integration between the two types of entities.

[0030] Then, construct the collaborative change index. Quantify the intensity of interaction between exploratory and developmental innovation entities; Constructing a Cooperative Change Index The calculation formula is: This index measures the relative strength of cross-type connections relative to same-type connections. The higher the value, the more active the interaction between the two types of entities and the stronger the synergistic effect.

[0031] Next, calculate the balance index. To assess the relative positions and developmental balance of the two types of entities in the network; Calculate the balance index The calculation formula is: in, To increase the number of exploratory innovation entities, The number of entities engaged in development-oriented innovation. The value ranges from [0, 1]. The closer the value is to 1, the more balanced the number of the two types of subjects is.

[0032] Finally, the rate of coordinated change is calculated based on the index differences between adjacent time slices: By analyzing the sign and magnitude of the rate of change, we can identify the mutation points and evolutionary trends of interaction patterns.

[0033] In S5, evolutionary pattern identification and risk warning, time series analysis is used to identify innovation network evolution patterns, construct path locking risk index and bubble risk index, and realize dynamic assessment and early warning of innovation network sustainability. The specific implementation process is as follows: First, time series analysis was performed on the network metrics for each time slice. A sliding window with a width of 3 was used to calculate the moving average and trend of the metrics to identify the stage characteristics of network evolution. Secondly, construct a path-locking risk index. Assess the degree of risk that innovation networks may fall into technological path dependence; Construct a path-locking risk index The calculation formula is: in, and The average degree centrality of developmental and exploratory subjects, respectively. and The internal network density of the two types of subjects are respectively. These are the weighting coefficients. The larger the value, the stronger the dominant position of the development entity, and the higher the risk of path lock-in.

[0034] Then, construct a bubble risk index. Assess the risk of a technology bubble emerging in innovation networks; Constructing a bubble risk index The calculation formula is: in, and These represent the changes in the number of two types of subjects between adjacent time slices. These are the weighting coefficients. The higher the value, the higher the degree of isolation of the exploratory subject, the faster the expansion, and the higher the risk of bubble formation.

[0035] Next, three risk warning thresholds are set: low risk (index < 1.0), medium risk (1.0 ≤ index < 1.5), and high risk (index ≥ 1.5). When or When the corresponding threshold is exceeded, a warning signal of the corresponding level is triggered.

[0036] Finally, based on historical evolution data, the ARIMA time series forecasting model is used to predict the risk index for the next 3-5 time slices, providing forward-looking decision support for policymakers.

[0037] An innovative network dynamic evolution analysis system based on time slicing and co-variation includes: The innovation entity classification module is used to construct an innovation entity classification and scoring model based on multi-dimensional features, dividing each innovation entity into exploratory innovation institutions and development innovation institutions; The time-slice network construction module is used to divide innovation cooperation data into slices according to preset time intervals. At each time slice, the innovation subject is used as a node and the cooperation relationship is used as an edge to construct a static innovation network at the corresponding time. The network structure index calculation module is used to calculate network density, node centrality, clustering coefficient, modularity, and structural hole index for innovative networks under each time slice. The collaborative change measurement module is used to construct an exploratory-development collaborative change measurement model to quantify the changes in connection strength and interaction balance between the two types of innovation subjects across time slices. The evolution pattern identification and risk warning module is used to identify the evolution patterns of innovation networks based on time series analysis, construct path locking risk index and bubble risk index, and realize dynamic assessment and early warning of the sustainability of innovation networks.

[0038] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for dynamic evolution analysis of innovative networks based on time slicing and collaborative change, characterized in that, Includes the following steps: S1. Classification of Innovation Entities: Based on multi-dimensional features, a classification and scoring model for innovation entities is constructed, dividing each innovation entity into exploratory innovation institutions and developmental innovation institutions; S2. Construction of Time Slice Network: Divide the innovation cooperation data into slices according to a preset time interval. At each time slice, construct a static innovation network corresponding to the time point with the innovation subject as the node and the cooperation relationship as the edge. S3. Calculation of network structure indicators: For each time slice of the innovative network, calculate the network density, node centrality, clustering coefficient, modularity and structural hole index. S4. Measurement of Collaborative Change: Construct an exploratory-development collaborative change measurement model to quantify the changes in connection strength and interactive balance between the two types of innovation subjects in different time slices; S5. Evolutionary Pattern Identification and Risk Warning: Based on time series analysis, identify the evolutionary patterns of innovation networks, construct path locking risk index and bubble risk index, and realize dynamic assessment and early warning of the sustainability of innovation networks.

2. The method for dynamic evolution analysis of innovation networks based on time slicing and collaborative change according to claim 1, characterized in that, In step S1, the classification of innovation entities, the process of constructing a classification and scoring model for innovation entities based on multi-dimensional features to divide each innovation entity into exploratory innovation institutions and developmental innovation institutions is as follows: Ten dimensions were selected as classification features: patent type, R&D investment intensity, technological novelty, venture capital ratio, knowledge flow pattern, talent mobility rate, degree of open innovation, proportion of basic research, organizational level and incentive mechanism. Normalize the features of each dimension and calculate the feature values ​​of each innovation subject in each dimension. Construct a classification scoring function, assign weights to each dimension, and calculate the comprehensive classification score S; The formula for constructing the classification scoring function, assigning weights to each dimension, and calculating the comprehensive classification score S is as follows: in, The proportion of basic patents, For a gradual increase in patent ratio, The ratio of R&D investment to revenue; For the proportion of interdisciplinary patents, For the proportion of venture capital, External knowledge citation rate; To reduce the inter-agency mobility of R&D personnel, The proportion of external cooperation projects, The percentage of basic research expenditure; For the number of organizational levels, The proportion of employee innovation rewards; to These are the weight coefficients corresponding to each dimension; Set a classification threshold: if S>0, it is judged as an exploratory innovation institution; if S≤0, it is judged as a developmental innovation institution.

3. The method for dynamic evolution analysis of innovation networks based on time slicing and collaborative change according to claim 1, characterized in that, In the S2 time-slice network construction, the process of dividing the innovation cooperation data into slices according to a preset time interval, and constructing a static innovation network at each time segment with innovation entities as nodes and cooperation relationships as edges, is as follows: Collect data on innovative collaborations, including joint patents, joint research projects, joint papers, and strategic cooperation data, with each data point marked with a precise timestamp; The research period is divided into T consecutive time slices based on annual or quarterly time intervals, denoted as T0. ; For each time slice Extract all cooperative relationships within this time period and construct an adjacency matrix. In the process, edge weight calculation is introduced. The calculation formula is: in, Time slice inner subject and Number of collaborations and The main body and Total number of collaborations within that time slice; in Representing the subject With the main body If cooperation exists during this time period, then the value is 0; otherwise, it is 0. Each node is labeled with an innovation type tag, marked as exploratory or developmental, forming a multi-time-section innovation network sequence with node attributes.

4. The method for dynamic evolution analysis of innovation networks based on time slicing and collaborative change according to claim 1, characterized in that, In the S3 network structure index calculation, the process of calculating network density, node centrality, clustering coefficient, modularity, and structural hole index for the innovative network under each time slice is as follows: Calculate network density It measures the overall connectivity of the network; Calculate network density The calculation formula is: in, Time slice The actual number of edges in the network. This represents the total number of nodes in that time slice. Calculate the degree centrality of each node. Betweenness centrality With proximity centrality Analyze the status and influence of nodes in the network; The betweenness centrality of each node is calculated. The calculation formula is: in, The total number of shortest paths from node s to node t. For the nodes The number of shortest paths; Calculate the average clustering coefficient of the network It measures the connectivity between a node's neighbors; The Louvain algorithm is used for community detection, and the modularity index is calculated. Assess the salience of network community structure; Calculate the structural hole constraint index for each node. Identify key nodes that connect different groups.

5. The method for dynamic evolution analysis of innovation networks based on time slicing and collaborative change according to claim 1, characterized in that, In S4, the process of constructing an exploratory-developmental collaborative change measurement model to quantify the changes in connection strength and interactive equilibrium between the two types of innovation subjects across different time slices is as follows: Count the number of connections between exploratory innovation entities within each time slice. Number of connections between development-oriented innovation entities and cross-type connection count ; Calculate cross-type connection ratio This measures the degree of integration between the two types of entities. Constructing a Cooperative Change Index Quantify the intensity of interaction between exploratory and developmental innovation entities; Calculate the balance index To assess the relative positions and developmental balance of the two types of entities in the network; Based on the index differences between adjacent time slices, the rate of coordinated change is calculated to identify abrupt changes in the interaction pattern.

6. The method for dynamic evolution analysis of innovation networks based on time slicing and collaborative changes according to claim 5, characterized in that, The calculation of cross-type connection ratio The calculation formula is: The construction of the collaborative change index The calculation formula is: The calculation of the balance index The calculation formula is: in, To increase the number of exploratory innovation entities, The number of entities engaged in development-oriented innovation.

7. The method for dynamic evolution analysis of innovation networks based on time slicing and collaborative change according to claim 1, characterized in that, In S5, evolutionary pattern identification and risk warning, the process of identifying innovation network evolutionary patterns based on time series analysis, constructing path locking risk index and bubble risk index, and realizing dynamic assessment and early warning of innovation network sustainability is as follows: Time series analysis was performed on the network indicators for each time slice, and the trend of indicator changes was calculated using the sliding window method. Construct a path-locking risk index The calculation is based on a weighted average of the concentration, internal connectivity density, and centrality advantage of developmental innovation entities. Constructing a bubble risk index The calculation is based on the isolation degree of exploratory innovation subjects, insufficient cross-type connections, and expansion speed. Set a risk warning threshold, and trigger a warning signal of the corresponding level when the risk index exceeds the threshold; Based on historical evolution data, a time series forecasting model is used to predict future risk trends.

8. The method for dynamic evolution analysis of innovation networks based on time slices and collaborative changes according to claim 7, characterized in that, The risk index of the construction path lock The calculation formula is: in, and The average degree centrality of developmental and exploratory subjects, respectively. and The internal network density of the two types of subjects are respectively. , , These are the weighting coefficients.

9. The method for dynamic evolution analysis of innovation networks based on time slices and collaborative changes according to claim 7, characterized in that, The construction of a bubble risk index The calculation formula is: in, and These represent the changes in the number of two types of subjects between adjacent time slices. , , These are the weighting coefficients.

10. A dynamic evolution analysis system for innovative networks based on time slicing and collaborative changes, characterized in that, include: The innovation entity classification module is used to construct an innovation entity classification and scoring model based on multi-dimensional features, dividing each innovation entity into exploratory innovation institutions and development innovation institutions; The time-slice network construction module is used to divide innovation cooperation data into slices according to preset time intervals. At each time slice, the innovation subject is used as a node and the cooperation relationship is used as an edge to construct a static innovation network at the corresponding time. The network structure index calculation module is used to calculate network density, node centrality, clustering coefficient, modularity, and structural hole index for innovative networks under each time slice. The collaborative change measurement module is used to construct an exploratory-development collaborative change measurement model to quantify the changes in connection strength and interaction balance between the two types of innovation subjects across time slices. The evolution pattern identification and risk warning module is used to identify the evolution patterns of innovation networks based on time series analysis, construct path locking risk index and bubble risk index, and realize dynamic assessment and early warning of the sustainability of innovation networks.

Images