Interpretation model construction method based on space interaction network

Abstract

The invention discloses an explanatory model construction method based on a space interaction network, which comprises the following steps of: accurately constructing the space interaction network by taking geographic units or functional areas as nodes, taking space interaction among the units as edges and endowing interaction strength as a weight, and performing community division by applying a community discovery algorithm. Community features and grouping structures can be effectively identified, relations and differences among communities can be disclosed, city planning and management can be assisted, network density, node distances and structural features are measured by adopting a social network analysis method, the influence of relational variables is analyzed, network complexity and dynamics are disclosed from multiple angles, node relations and network features are utilized to construct an explanatory model, and city planning and management can be realized. According to the method, the spatial interaction phenomenon is deeply explained, influence factors are revealed, theoretical support is provided for prediction and optimization, a multi-spatial interaction network explanatory model is constructed, the competitive cooperation relation between different travel modes or networks is analyzed, interaction and influence are revealed, and a scientific basis is provided for optimization and integration of an urban traffic system.

Description

Technical Field

[0001] The present invention relates to the technical field of explanatory model construction, and in particular to an explanatory model construction method based on a spatial interaction network. Background Art

[0002] In the field of analysis and application of spatial interaction networks, although researchers have developed a variety of technologies and methods to understand and explain the complex interactions in the network, existing technologies still face a series of challenges and limitations:

[0003] First, when constructing explanatory models of spatial interaction networks, existing technologies usually only focus on the static attributes of nodes (such as cities, communities, etc.), such as population size, economic level, etc., but ignore the dynamics and complexity of the interaction relationships between nodes. This simplified processing method leads to limitations in the model in explaining network evolution, predicting future trends, and evaluating policy effects.

[0004] Secondly, although community segmentation methods have made significant progress in identifying and analyzing the cluster structure of spatial interaction networks, existing technologies often only focus on the static division results of communities and lack in-depth research on the dynamic changes within communities and the interaction relationships between communities. This limits our comprehensive understanding of the characteristics and evolution laws of network cluster structures.

[0005] Furthermore, the application of social network analysis methods in spatial interaction networks also faces challenges. Existing technologies usually only focus on the macro-structural characteristics of the network, such as network density, average path length, etc., but lack a detailed analysis of the role played by individual nodes in the network and the interaction mechanism between nodes. This leads to deficiencies in revealing the relationship variables between the individual behaviors of network members and the overall behavior of the system.

[0006] In addition, when interpreting spatial interaction networks, existing technologies often ignore the competition and coordination relationships between networks under different travel modes or different spatial scales. This neglect may lead to deviations in analyzing network interaction intensity, optimizing network layout, and formulating relevant policies.

[0007] Finally, existing technologies lack a unified analytical framework and effective algorithms when processing and analyzing multiple spatial interaction networks. This limits our in-depth understanding of the mutual influence and mutual constraints between different networks, and also affects our application effects in urban planning, traffic management and other fields. Summary of the invention

[0008] In view of this, the present invention proposes a method for constructing an explanatory model based on a spatial interaction network, which can comprehensively consider node attributes, relationship variables, network structure characteristics, and network competition and coordination relationships under different travel modes or spatial scales to construct a more accurate and comprehensive explanatory model.

[0009] The technical solution of the present invention is achieved in this way:

[0010] A method for constructing an explanatory model based on a spatial interaction network, specifically comprising:

[0011] The basic units of the spatial interaction network are selected, including nodes and edges, where nodes represent geographical units or functional areas, edges represent spatial interactions between units, and the interaction intensity is used as the weight of the edge;

[0012] Apply community discovery algorithms to divide spatial interaction networks into communities, identify and analyze the characteristics and group structures of each community;

[0013] Using social network analysis methods, we measure the density of spatial interaction networks, the distance between nodes, and the characteristics of network structure, and analyze the impact of relationship variables on individual network members and system behaviors;

[0014] Construct explanatory models to explain spatial interaction networks using the relationships between interaction nodes;

[0015] The network structure features are introduced into the explanatory model, and the network centrality of the edges is used to explain the spatial interactions;

[0016] Construct an explanatory model of multiple spatial interaction networks, analyze the competition and coordination between different travel modes or different spatial interaction networks, and analyze their impact on the intensity of spatial interaction.

[0017] As a further optional solution of the method for constructing an explanatory model based on a spatial interaction network, the basic units of the spatial interaction network are selected, including nodes and edges, wherein nodes represent geographical units or functional areas, edges represent spatial interactions between units, and interaction strength is used as the weight of the edge, specifically including:

[0018] Set nodes, edges and interaction strengths;

[0019] The determined nodes and edges are connected according to the actual spatial relationship and interaction intensity to form a spatial interaction network.

[0020] As a further optional solution of the method for constructing an explanatory model based on a spatial interaction network, the application of a community discovery algorithm to divide the spatial interaction network into communities, and to identify and analyze the characteristics and group structure of each community, specifically includes:

[0021] The spatial interaction network is divided into communities based on the community discovery algorithm to obtain the initial community division results;

[0022] The initial division results are evaluated according to the modularity formula to obtain the final community division results;

[0023] Based on the final community division results, identify the characteristics and group structure of each community;

[0024] According to the characteristics and group structure of each community, the functional positioning, interaction mode and possible influencing factors of each community are obtained.

[0025] As a further optional solution of the method for constructing an explanatory model based on a spatial interaction network, the constructing of the explanatory model, using the relationship between the interaction nodes to explain the spatial interaction network, specifically includes:

[0026] According to the relationship between the interactive nodes, the attribute data of the node or the surrounding nodes are obtained;

[0027] Based on the relationship between the interacting nodes, the complementarity and similarity between the nodes are calculated;

[0028] An explanatory model is constructed based on the attribute data of nodes or their surroundings, as well as the complementarity and similarity between nodes.

[0029] As a further optional solution of the method for constructing an explanatory model based on a spatial interaction network, the explanatory model introduces network structure features and uses the network centrality of edges to explain spatial interactions, specifically including:

[0030] Obtain network structure characteristics;

[0031] Calculating the centrality of the network according to the network structure characteristics, the centrality of the network includes proximity centrality, betweenness centrality and directness centrality;

[0032] Through statistical analysis and visualization methods, the relationship between network centrality and spatial interaction phenomena is obtained.

[0033] As a further optional solution of the method for constructing an explanatory model based on a spatial interaction network, the explanatory model of constructing multiple spatial interaction networks analyzes the competition and coordination relationship between different travel modes or different spatial interaction networks, and analyzes their impact on the intensity of spatial interaction, specifically including:

[0034] Collect OD data for multiple travel modes;

[0035] Based on the collected OD data, a spatial interaction network of different travel modes is constructed;

[0036] The competition and coordination index between different travel modes is calculated by calculating the correlation coefficient of the interaction flow between the two travel modes;

[0037] Combining competition and coordination indicators, the impact of competition and coordination relationships among different travel modes on spatial interaction intensity is analyzed.

[0038] As a further optional solution of the method for constructing an explanatory model based on a spatial interaction network, the method further includes:

[0039] Select spatial units and data for explanatory models of multiple spatial interaction networks;

[0040] Interpretability methods were applied to evaluate and validate the selected spatial units and data.

[0041] A spatial interaction network-based explanatory model building system, comprising:

[0042] A selection module is used to select basic units of the spatial interaction network, including nodes and edges, where nodes represent geographical units or functional areas, edges represent spatial interactions between units, and interaction strength is used as the weight of the edge;

[0043] The partitioning module is used to apply the community discovery algorithm to divide the spatial interaction network into communities, identify and analyze the characteristics and group structure of each community;

[0044] The measurement module is used to measure the density of spatial interaction networks, the distance between nodes and the network structure characteristics using social network analysis methods, and to analyze the impact of relationship variables on individual network members and system behaviors;

[0045] The first building module is used to build an explanatory model to explain the spatial interaction network using the relationships between interaction nodes;

[0046] An introduction module is used to introduce network structure features into the explanatory model and explain spatial interactions using the network centrality of edges;

[0047] The second building block is used to construct an explanatory model of multiple spatial interaction networks, analyze the competition and synergy between different travel modes or different spatial interaction networks, and analyze their impact on the intensity of spatial interaction.

[0048] A computing device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for constructing an explanatory model based on a spatial interaction network when executing the computer program.

[0049] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for constructing an explanatory model based on a spatial interaction network.

[0050] The beneficial effects of the present invention are as follows: by taking geographical units or functional areas as nodes, spatial interactions between units as edges, and assigning interaction intensity as edge weights, the technical solution can accurately construct a spatial interaction network, providing a solid foundation for subsequent analysis; applying a community discovery algorithm to divide the spatial interaction network into communities can not only effectively identify the characteristics and group structures of each community, but also reveal the connections and differences between different communities, providing valuable references for urban planning and management; using a social network analysis method to measure the density of the spatial interaction network, the distance between nodes and the network structure characteristics, and analyzing the influence of relationship variables on individual network members and system behaviors, it can reveal the complexity and dynamics of the spatial interaction network from multiple perspectives; using the relationship between interaction nodes and network structure characteristics to construct an explanatory model, it can explain the spatial interaction phenomenon in more depth, reveal how different factors affect the spatial interaction intensity, and provide theoretical support for the prediction and optimization of spatial interaction; constructing an explanatory model of multiple spatial interaction networks, analyzing the competition and synergy between different travel modes or different spatial interaction networks, it can reveal the interaction and influence between different travel modes, and provide a scientific basis for the optimization and integration of urban transportation systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention 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 invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0052] Figure 1 A schematic diagram of a flow chart of a method for constructing an explanatory model based on a spatial interaction network according to the present invention;

[0053] Figure 2 A schematic diagram of the composition of an explanatory model building system based on a spatial interaction network of the present invention;

[0054] Figure 3 The figure is a schematic diagram of the composition of a computing device according to the present invention. DETAILED DESCRIPTION

[0055] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] refer to Figures 1 to 3, a method for constructing an explanatory model based on a spatial interaction network, specifically including:

[0057] The basic units of the spatial interaction network are selected, including nodes and edges, where nodes represent geographical units or functional areas, edges represent spatial interactions between units, and the interaction intensity is used as the weight of the edge;

[0058] Apply community discovery algorithms to divide spatial interaction networks into communities, identify and analyze the characteristics and group structures of each community;

[0059] Using social network analysis methods, we measure the density of spatial interaction networks, the distance between nodes, and the characteristics of network structure, and analyze the impact of relationship variables on individual network members and system behaviors;

[0060] Construct explanatory models to explain spatial interaction networks using the relationships between interaction nodes;

[0061] The network structure features are introduced into the explanatory model, and the network centrality of the edges is used to explain the spatial interactions;

[0062] Construct an explanatory model of multiple spatial interaction networks, analyze the competition and coordination between different travel modes or different spatial interaction networks, and analyze their impact on the intensity of spatial interaction.

[0063] In this embodiment, by taking geographic units or functional areas as nodes, spatial interactions between units as edges, and assigning interaction intensity as edge weights, the technical solution can accurately construct a spatial interaction network, providing a solid foundation for subsequent analysis. The application of community discovery algorithms to divide the spatial interaction network into communities can not only effectively identify the characteristics and group structures of each community, but also reveal the connections and differences between different communities, providing valuable references for urban planning and management. The social network analysis method is used to measure the density of the spatial interaction network, the distance between nodes, and the network structure characteristics, and analyze the impact of relationship variables on individual network members and system behaviors. It can reveal the complexity and dynamics of the spatial interaction network from multiple perspectives. The relationship between interaction nodes and network structure characteristics are used to construct an explanatory model, which can explain the spatial interaction phenomenon in more depth, reveal how different factors affect the spatial interaction intensity, and provide theoretical support for the prediction and optimization of spatial interaction. An explanatory model of multiple spatial interaction networks is constructed, and the competition and synergy between different travel modes or different spatial interaction networks is analyzed. It can reveal the interaction and influence between different travel modes, and provide a scientific basis for the optimization and integration of urban transportation systems.

[0064] Preferably, the basic unit of the selected spatial interaction network includes nodes and edges, wherein nodes represent geographical units or functional areas, edges represent spatial interactions between units, and interaction strength is used as the weight of the edge, specifically including:

[0065] Set nodes, edges and interaction strengths;

[0066] The determined nodes and edges are connected according to the actual spatial relationship and interaction intensity to form a spatial interaction network.

[0067] In this embodiment, by setting nodes as geographic units or functional areas and edges as spatial interactions between these units, the spatial interactions within or between cities can be accurately depicted. This definition method allows each element in the network to have clear geographic or functional attributes, which is convenient for subsequent analysis and interpretation. By taking the interaction intensity as the weight of the edge, not only the existence of spatial interaction but also its strength is considered. This weighted processing method enables the network to more truly reflect the actual spatial interaction situation and provides richer information for subsequent analysis. By applying the community discovery algorithm to divide the spatial interaction network into communities, the group structure in the network can be efficiently identified. These group structures often correspond to geographic units with similar functions or attributes, which is of great significance for understanding the spatial structure and functional layout of the city. On the basis of community division, the characteristics of each community can be further analyzed in depth, including the interaction intensity within the community, the interaction mode between communities, and the status and role of the community in the entire network. These analyses help to reveal The complexity and diversity of urban spatial interaction can be shown by the social network analysis method. Multiple aspects of the spatial interaction network can be measured, such as network density, node centrality, edge weight distribution, etc. These indicators can reflect the structural characteristics and interaction patterns of the network from different angles, and provide comprehensive data support for subsequent analysis. Through social network analysis, the structural characteristics of the network can be deeply analyzed, such as the clustering of nodes and the distribution of edges. These analyses help to reveal the complex relationships and interaction mechanisms within the network, and provide strong support for understanding the dynamic changes of urban spatial interaction. This technical solution not only considers the relationship between nodes and edges, but also integrates a variety of explanatory methods such as community division and social network analysis. This integration enables the model to explain the spatial interaction phenomenon more comprehensively and improves the accuracy and credibility of the explanation. By adjusting the definition of nodes and edges and the choice of explanatory methods, this technical solution can adapt to spatial interaction analysis of different scales and needs, including the interaction between different functional areas within the city, the interaction between cities, and the interaction between different travel modes.

[0068] Preferably, the application of community discovery algorithm to divide the spatial interaction network into communities, identify and analyze the characteristics and group structure of each community, specifically includes:

[0069] The spatial interaction network is divided into communities based on the community discovery algorithm to obtain the initial community division results;

[0070] The initial division results are evaluated according to the modularity formula to obtain the final community division results;

[0071] Based on the final community division results, identify the characteristics and group structure of each community;

[0072] According to the characteristics and group structure of each community, the functional positioning, interaction mode and possible influencing factors of each community are obtained.

[0073] In this embodiment, by applying the community discovery algorithm, the spatial interaction network can be efficiently divided into communities. A community refers to an organic organization with a certain collaborative relationship. Community division helps to clearly analyze the closeness of connections between cities and reveal the cluster structure of urban spatial interactions. The modularity formula is used to evaluate the initial community division results to ensure the accuracy and effectiveness of the division. The higher the modularity, the more obvious the community structure and the closer the internal connection of the community, thereby improving the reliability of community division. According to the final community division results, the characteristics and cluster structure of each community can be identified, and then the functional positioning of each community can be obtained, which helps to understand the role of different communities in the city. The roles and functions in the spatial interaction network provide strong support for urban planning and management; through community division, we can deeply analyze the interaction patterns between communities, including the frequency, intensity and direction of interaction, which helps to reveal the dynamic changes of urban spatial interaction and provide a scientific basis for urban transportation planning, land use, etc.; according to the characteristics and group structure of each community, we can further explore the factors that may affect community interaction. These factors may include geographical location, traffic conditions, economic development level and other aspects. Through in-depth analysis of these factors, we can provide strong support for the formulation of targeted urban policies and measures and promote the coordinated development of urban spatial interaction.

[0074] It should be noted that the initial division results are evaluated according to the modularity formula to obtain the final community division results, which include:

[0075] Calculate modularity: Use the modularity formula to calculate, which takes into account multiple factors such as the number of complex network edges, adjacent matrix, node degree, and the community where the node is located. Through calculation, a quantitative indicator can be obtained to evaluate the quality of community division;

[0076] Evaluation of the division results: Based on the value of modularity, the initial community division results are evaluated. When Q>0, the community structure begins to appear; when Q>0.3, the community structure begins to become obvious. Generally, the value of the Q function is between 0.3 and 0.7. The closer the value of Q is to 1, the more obvious the community structure is and the closer the connection within the community is.

[0077] Determine the final division: Based on the evaluation results, select the division with higher modularity as the final community division result. This ensures that the obtained community division results have high accuracy and reliability, and provides strong support for the subsequent identification and analysis of the characteristics and group structure of each community.

[0078] In addition, the specific steps for calculating modularity include:

[0079] Step 1. Determine the parameters:

[0080] m: the number of edges in the complex network;

[0081] A_i j: The edge connecting the adjacent matrices i and j. When A_i j = 0, it means that there is no edge between the adjacent matrices i and j.

[0082] K_i, K_j: the degrees of node i and node j respectively;

[0083] C_i, C_j: the communities where nodes i and j are located respectively. If and only if nodes i and j are in the same community, δ(C_i, C_j) = 1, otherwise δ(C_i, C_j) = 0;

[0084] Step 2: Apply the modularity formula:

[0085] The definition of modularity is:

[0086]

[0087] Among them, this formula comprehensively considers the structural characteristics of complex networks and evaluates the quality of the division by calculating the degree of deviation between the connection between nodes and the random connection;

[0088] Step 3: Perform calculations:

[0089] Traverse all node pairs (i, j) and calculate A_i j and K_i K_j / (2m).

[0090] Depending on whether nodes i and j are in the same community (i.e., the value of δ(C_i,C_j)), the corresponding degree of deviation is accumulated.

[0091] Finally, divide the accumulated result by 2m to get the value of modularity Q;

[0092] Step 4: Interpret the results:

[0093] When Q>0, it means that the community structure begins to emerge.

[0094] When Q>0.3, the community structure begins to emerge clearly.

[0095] The closer the Q value is to 1, the more obvious the community structure is and the closer the connections within the community are.

[0096] Preferably, the construction of the explanatory model, using the relationship between the interaction nodes to explain the spatial interaction network, specifically includes:

[0097] According to the relationship between the interactive nodes, the attribute data of the node or the surrounding nodes are obtained;

[0098] Based on the relationship between the interacting nodes, the complementarity and similarity between the nodes are calculated;

[0099] An explanatory model is constructed based on the attribute data of nodes or their surroundings, as well as the complementarity and similarity between nodes.

[0100] In this embodiment, by obtaining attribute data of nodes or their surroundings (such as land price, population, volume ratio, POI density, land use diversity, etc.), the technical solution can more comprehensively understand the basic background of spatial interaction, and these data provide a rich information basis for analyzing spatial interaction; calculating the complementarity and similarity between nodes is one of the core steps of the technical solution. Complementarity reflects the ability of nodes to meet each other's needs, such as the complementarity between residential and work functions, while similarity measures the degree of similarity of functions or attributes between nodes. This quantitative analysis helps to reveal the internal mechanism of spatial interaction; combined with the attribute data of nodes or their surroundings, as well as the complementarity and similarity between nodes, the constructed explanatory model can more accurately describe and predict the spatial interaction phenomenon. This model not only helps to understand the current spatial interaction pattern, but also provides a scientific basis for future urban planning, transportation planning, etc.; through the constructed explanatory model, the technical solution can provide decision makers with more accurate and scientific support. For example, in urban planning, the model can be used to predict the interaction intensity of different functional areas, thereby optimizing the functional layout. In transportation planning, the model can be used to analyze the synergy and competition relationship between different transportation modes and formulate a reasonable transportation plan.

[0101] Preferably, the explanatory model introduces network structure features and uses the network centrality of edges to explain spatial interactions, specifically including:

[0102] Obtain network structure characteristics;

[0103] Calculating the centrality of the network according to the network structure characteristics, the centrality of the network includes proximity centrality, betweenness centrality and directness centrality;

[0104] Through statistical analysis and visualization methods, the relationship between network centrality and spatial interaction phenomena is obtained.

[0105] In this embodiment, by calculating the centrality of the network (including proximity centrality, betweenness centrality and directness centrality), the impact of the network structure on spatial interaction can be quantified, providing more accurate quantitative indicators for analyzing spatial interaction. Through statistical analysis and visualization methods, the relationship between network centrality and spatial interaction phenomena is intuitively presented, which helps to reveal the inherent connection and laws between network structure and spatial interaction. This technical solution can provide a scientific basis for urban planning and traffic management. By identifying key sections and nodes with high network centrality, the layout of the urban road network can be optimized, traffic efficiency and accessibility can be improved, and traffic congestion can be alleviated.

[0106] It should be noted that the network centrality is calculated based on the network structure characteristics. The network centrality includes proximity centrality, betweenness centrality and direct centrality, specifically including:

[0107] Proximity centrality calculation: Calculate the sum of the shortest path distances from each node to all other nodes, take the inverse of the sum of the distances, and get the proximity centrality value of each node. The larger the proximity centrality value, the higher the relative accessibility of the node in the network.

[0108] Betweenness centrality calculation: Calculate the total number of shortest paths between all node pairs, count the number of shortest paths passing through each node, and obtain the betweenness centrality value of each node. The higher the betweenness centrality value, the stronger the node's transit and connection capabilities in the network;

[0109] Direct centrality calculation: Calculate the degree of deviation between the shortest path and the Euclidean distance between each node pair. The smaller the deviation, the better the direct centrality between the nodes and the higher the connection efficiency. Summarize the direct centrality values ​​of all node pairs to obtain the direct centrality characteristics of the entire network.

[0110] Preferably, the explanatory model of constructing multiple spatial interaction networks, analyzing the competition and coordination relationship between different travel modes or different spatial interaction networks, and analyzing their impact on the intensity of spatial interaction, specifically includes:

[0111] Collect OD data for multiple travel modes;

[0112] Based on the collected OD data, a spatial interaction network of different travel modes is constructed;

[0113] The competition and coordination index between different travel modes is calculated by calculating the correlation coefficient of the interaction flow between the two travel modes;

[0114] Combining competition and coordination indicators, the impact of competition and coordination relationships among different travel modes on spatial interaction intensity is analyzed.

[0115] In this embodiment, by collecting OD data of multiple travel modes, the technical solution can construct a spatial interaction network covering multiple travel modes, thereby more comprehensively reflecting the complexity of urban spatial interaction, which helps to deeply understand the role and mutual influence of different travel modes in spatial interaction; by calculating the correlation coefficient of the interaction flow between two travel modes, the technical solution can quantify the competition and synergy relationship between different travel modes. This quantitative analysis provides an objective basis for evaluating the mutual influence between different travel modes and helps to reveal the deep-seated mechanism behind spatial interaction; combined with competition and synergy indicators, the technical solution can analyze the impact of competition and synergy between different travel modes on the intensity of spatial interaction, which helps to identify the most competitive travel modes in different regions and different time periods, and which travel modes have synergy effects. This information is of great significance for optimizing public transportation planning and improving the overall efficiency and service level of public transportation systems; by constructing an explanatory model of multiple spatial interaction networks, the technical solution provides a scientific basis for urban transportation planning and policy formulation. By deeply analyzing the competition and synergy relationship between different travel modes, it can more accurately predict the changing trend of future traffic flow and provide strong support for urban transportation planning and policy adjustments.

[0116] Preferably, the method further comprises:

[0117] Select spatial units and data for explanatory models of multiple spatial interaction networks;

[0118] Interpretability methods were applied to evaluate and validate the selected spatial units and data.

[0119] In this embodiment, by carefully selecting spatial units (such as social management grids, subway station areas, etc.), the spatial interaction characteristics of different areas can be captured more accurately, making the analysis results more targeted and practical. The precise selection of data (such as OD data of multiple travel modes such as subways, buses, taxis, and shared bicycles) ensures the comprehensiveness and accuracy of the analysis, and helps to deeply explore the competition and synergy relationship between different travel modes; applying interpretable methods (such as linear regression, geographically weighted regression, etc.) to evaluate and verify spatial units and data not only improves the credibility of the model, but also makes the analysis results easier to understand and accept. Applying interpretable methods (such as linear regression, geographically weighted regression, etc.) to evaluate and verify spatial units and data not only improves the credibility of the model, but also makes the analysis results easier to understand and accept; by constructing an explanatory model of multiple spatial interaction networks and deeply analyzing the competition and synergy relationship between different travel modes, the complexity and dynamics of spatial interaction networks can be more clearly understood.

[0120] A spatial interaction network-based explanatory model building system, comprising:

[0121] A selection module is used to select basic units of the spatial interaction network, including nodes and edges, where nodes represent geographical units or functional areas, edges represent spatial interactions between units, and interaction strength is used as the weight of the edge;

[0122] The partitioning module is used to apply the community discovery algorithm to divide the spatial interaction network into communities, identify and analyze the characteristics and group structure of each community;

[0123] The measurement module is used to measure the density of spatial interaction networks, the distance between nodes and the network structure characteristics using social network analysis methods, and to analyze the impact of relationship variables on individual network members and system behaviors;

[0124] The first building module is used to build an explanatory model to explain the spatial interaction network using the relationships between interaction nodes;

[0125] An introduction module is used to introduce network structure features into the explanatory model and explain spatial interactions using the network centrality of edges;

[0126] The second building block is used to construct an explanatory model of multiple spatial interaction networks, analyze the competition and synergy between different travel modes or different spatial interaction networks, and analyze their impact on the intensity of spatial interaction.

[0127] A computing device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for constructing an explanatory model based on a spatial interaction network when executing the computer program.

[0128] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for constructing an explanatory model based on a spatial interaction network.

[0129] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for constructing an explanatory model based on a spatial interaction network, characterized in that: Specifically include: The basic units of the spatial interaction network are selected, including nodes and edges, where nodes represent geographical units or functional areas, edges represent spatial interactions between units, and the interaction intensity is used as the weight of the edge; Apply community discovery algorithms to divide spatial interaction networks into communities, identify and analyze the characteristics and group structures of each community; Using social network analysis methods, we measure the density of spatial interaction networks, the distance between nodes, and the characteristics of network structure, and analyze the impact of relationship variables on individual network members and system behaviors; Construct explanatory models to explain spatial interaction networks using the relationships between interaction nodes; The network structure features are introduced into the explanatory model, and the network centrality of the edges is used to explain the spatial interactions; Construct an explanatory model of multiple spatial interaction networks, analyze the competition and coordination between different travel modes or different spatial interaction networks, and analyze their impact on the intensity of spatial interaction.

2. The method for constructing an explanatory model based on a spatial interaction network according to claim 1, characterized in that: The basic units of the selected spatial interaction network include nodes and edges, where nodes represent geographical units or functional areas, edges represent spatial interactions between units, and interaction strength is used as the weight of the edge, specifically including: Set nodes, edges and interaction strengths; The determined nodes and edges are connected according to the actual spatial relationship and interaction intensity to form a spatial interaction network.

3. The method for constructing an explanatory model based on a spatial interaction network according to claim 2, characterized in that: The application of community discovery algorithm divides the spatial interaction network into communities, identifies and analyzes the characteristics and group structure of each community, specifically including: The spatial interaction network is divided into communities based on the community discovery algorithm to obtain the initial community division results; The initial division results are evaluated according to the modularity formula to obtain the final community division results; Based on the final community division results, identify the characteristics and group structure of each community; According to the characteristics and group structure of each community, the functional positioning, interaction mode and possible influencing factors of each community are obtained.

4. The method for constructing an explanatory model based on a spatial interaction network according to claim 3, characterized in that: The construction of the explanatory model uses the relationship between the interaction nodes to explain the spatial interaction network, specifically including: According to the relationship between the interactive nodes, the attribute data of the node or the surrounding nodes are obtained; Based on the relationship between the interacting nodes, the complementarity and similarity between the nodes are calculated; An explanatory model is constructed based on the attribute data of nodes or their surroundings, as well as the complementarity and similarity between nodes.

5. The method for constructing an explanatory model based on a spatial interaction network according to claim 4, characterized in that: The explanatory model introduces network structure features and uses the network centrality of edges to explain spatial interactions, including: Obtain network structure characteristics; Calculating the centrality of the network according to the network structure characteristics, the centrality of the network includes proximity centrality, betweenness centrality and directness centrality; Through statistical analysis and visualization methods, the relationship between network centrality and spatial interaction phenomena is obtained.

6. The method for constructing an explanatory model based on a spatial interaction network according to claim 5, characterized in that: The explanatory model of multiple spatial interaction networks is constructed to analyze the competition and coordination relationship between different travel modes or different spatial interaction networks, and analyze their impact on the intensity of spatial interaction, including: Collect OD data for multiple travel modes; Based on the collected OD data, a spatial interaction network of different travel modes is constructed; The competition and coordination index between different travel modes is calculated by calculating the correlation coefficient of the interaction flow between the two travel modes; Combining competition and coordination indicators, the impact of competition and coordination relationships among different travel modes on spatial interaction intensity is analyzed.

7. The method for constructing an explanatory model based on a spatial interaction network according to claim 6, characterized in that: The method further comprises: Select spatial units and data for explanatory models of multiple spatial interaction networks; Interpretability methods were applied to evaluate and validate the selected spatial units and data.

8. An explanatory model building system based on spatial interaction network, characterized in that: include: A selection module is used to select basic units of the spatial interaction network, including nodes and edges, where nodes represent geographical units or functional areas, edges represent spatial interactions between units, and interaction strength is used as the weight of the edge; The partitioning module is used to apply the community discovery algorithm to divide the spatial interaction network into communities, identify and analyze the characteristics and group structure of each community; The measurement module is used to measure the density of spatial interaction networks, the distance between nodes and the network structure characteristics using social network analysis methods, and to analyze the impact of relationship variables on individual network members and system behaviors; The first building module is used to build an explanatory model to explain the spatial interaction network using the relationships between interaction nodes; An introduction module is used to introduce network structure features into the explanatory model and explain spatial interactions using the network centrality of edges; The second building block is used to construct an explanatory model of multiple spatial interaction networks, analyze the competition and synergy between different travel modes or different spatial interaction networks, and analyze their impact on the intensity of spatial interaction.

9. A computing device, characterized in that It comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method for constructing an explanatory model based on a spatial interaction network as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for constructing an explanatory model based on a spatial interaction network as described in any one of claims 1 to 7.

Images