Innovation chain and industry chain fusion method based on knowledge graph

Abstract

The invention relates to the technical field of innovation chain and industry chain fusion, and discloses an innovation chain and industry chain fusion method based on a knowledge graph. According to the method, entities and relationships thereof in an innovation chain and an industry chain are collected, a graph neural network model is used for constructing a knowledge graph, and the entities and the relationships are subjected to embedded representation. Then, connecting points between the two are recognized, connection is represented in a graph structure, and the knowledge graph is updated; and dynamically adjusting the importance of innovation chain and industry chain information by adopting a fusion model based on an attention mechanism to realize the fusion of the knowledge graph. And finally, on the basis of the fused knowledge graph, upstream and downstream support information of an industry chain is provided for technical innovation, and rapid conversion and industrialization of technical achievements are promoted. In conclusion, the efficiency and accuracy of information processing are effectively improved, the relevance and interactivity between the innovation chain and the industry chain are enhanced, and the method has remarkable effects of promoting deep fusion of the innovation chain and the industry chain and accelerating technical achievement transformation and industry upgrading.

Description

Technical Field

[0001] The present invention relates to the technical field of integration of innovation chain and industrial chain, and specifically to a method for integrating innovation chain and industrial chain based on knowledge graph. Background Art

[0002] In the current scientific and technological innovation and industrial development environment, the effective integration of the innovation chain and the industrial chain has become the key to promoting technological progress and industrial upgrading. However, traditional integration methods often have problems such as information islands and data non-interoperability, resulting in insufficient connection between the innovation chain and the industrial chain, affecting the rapid transformation and industrialization of technological achievements. In order to solve this problem, the industry has begun to explore technical means based on knowledge graphs to achieve a comprehensive combing and deep integration of entities and their relationships in the innovation chain and the industrial chain. However, most of the existing knowledge graph-based fusion methods only stay at the data integration level, lack intelligent information processing and dynamic adjustment mechanisms, and are difficult to accurately capture the complex connections and potential values ​​between the innovation chain and the industrial chain. Therefore, there is an urgent need for a new method that can efficiently integrate the knowledge graphs of the innovation chain and the industrial chain and provide upstream and downstream support information for the industrial chain to promote the rapid development of technological innovation and the in-depth promotion of industrialization applications. Summary of the invention

[0003] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to provide a method for integrating innovation chain and industrial chain based on knowledge graph, so as to solve the problems of information islands and data non-interoperability in the integration of innovation chain and industrial chain in the existing technology.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a method for integrating an innovation chain and an industrial chain based on a knowledge graph, the method comprising: Step S100: Collect entities and their relationships in the innovation chain, use a graph neural network model to represent the entities and their relationships in the innovation chain as graph structure data, build an innovation chain knowledge graph, and embed entities and relationships in the innovation chain knowledge graph; Step S200: Collect entities and their relationships in the industrial chain, use a graph neural network model to represent the entities and their relationships in the industrial chain as graph structure data, build an industrial chain knowledge graph, and embed the entities and relationships in the industrial chain knowledge graph; Step S300: According to the connection points between the innovation chain and the industrial chain, the connection between the innovation chain and the industrial chain is represented in the graph structure, and the entity and relationship representation in the innovation chain knowledge graph and the industrial chain knowledge graph are updated; Step S400: Adopting a fusion model based on the attention mechanism, dynamically adjusting the importance of the innovation chain and the industrial chain information according to the weight distribution of the attention mechanism, fusing the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph, and generating a fused knowledge graph representation; Step S500: Based on the integrated knowledge graph, provide support information from upstream and downstream of the industrial chain for technological innovation.

[0005] Preferably, in a possible implementation of the first aspect, the entities and their relationships in the innovation chain include technological inventions, R&D teams, R&D projects, technology transfer and achievement transformation; the entities and their relationships in the industrial chain include raw material supply, production and manufacturing, logistics and transportation, marketing and consumer feedback.

[0006] Preferably, in a possible implementation of the first aspect, the graph neural network model includes: an input layer for receiving entities and their relationship data in an innovation chain or an industrial chain; a graph convolution layer for performing convolution operations on graph structure data to extract graph structure features; a pooling layer for performing dimensionality reduction processing on the output of the graph convolution layer; and an output layer for outputting graph structure data.

[0007] Preferably, in a possible implementation manner of the first aspect, the embedding method uses a TransR method.

[0008] Preferably, in a possible implementation manner of the first aspect, step S300 specifically includes: Identify potential connection points based on entities and their relationships in the innovation and industrial chains; According to the type of connection point, corresponding edges or nodes are added to the graph structure to represent the connection between the innovation chain and the industrial chain; Update the entity and relationship representations in the innovation chain knowledge graph and the industrial chain knowledge graph, including updating the attribute information of the entities and the weight information of the relationships.

[0009] Preferably, in a possible implementation of the first aspect, the fusion model based on the attention mechanism includes: The encoder layer is used to encode the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph; The attention mechanism layer is used to calculate the attention weights of each entity and its relationship in the innovation chain and industrial chain, and adjust the importance of the embedded representation according to the weights; The fusion layer is used to fuse the weight-adjusted embedding representations to generate a fused knowledge graph representation.

[0010] Preferably, in a possible implementation manner of the first aspect, step S400 specifically includes: Input the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph into the encoder layer for encoding; The encoded embedding representation is input into the attention mechanism layer to calculate the attention weights of each entity and its relations; Adjust the importance of the embedded representation according to the attention weight to obtain the adjusted embedded representation; The adjusted embedding representation is input into the fusion layer for fusion to generate the fused knowledge graph representation.

[0011] Preferably, in a possible implementation manner of the first aspect, step S500 specifically includes: Extract supporting information from upstream and downstream of the industrial chain based on the fused knowledge graph representation; Guide technological innovation and achievement transformation based on support information from upstream and downstream of the industrial chain; Monitor the application effect of technological innovation in the industrial chain and make optimization adjustments based on feedback information.

[0012] Preferably, in a possible implementation of the first aspect, the method further includes visually displaying the fused knowledge graph so as to intuitively understand the relationship between the innovation chain and the industrial chain.

[0013] The beneficial effect of the present invention is that the entities and their relationships in the innovation chain and the industrial chain are embedded and represented by a graph neural network model, which effectively improves the efficiency and accuracy of information processing. At the same time, by representing the connection between the innovation chain and the industrial chain according to the connection points between the innovation chain and the industrial chain in a graph structure, the relevance and interactivity between the innovation chain and the industrial chain are further enhanced. In addition, the present invention also introduces a fusion model based on the attention mechanism, dynamically adjusts the importance of the innovation chain and industrial chain information according to the allocation of attention weights, and realizes the intelligent fusion and optimization of the information of the two. This method not only provides support information for the upstream and downstream of the industrial chain for technological innovation, promotes the rapid transformation and industrialization of technological achievements, but also ensures the accuracy and timeliness of the fused knowledge map through means such as visual display and update maintenance. Therefore, the present invention has significant beneficial effects in promoting the deep integration of the innovation chain and the industrial chain, accelerating the transformation of technological achievements and industrial upgrading. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] 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.

[0015] Figure 1A method flow chart based on the integration of knowledge graph innovation chain and industrial chain is provided for this application. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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.

[0017] Embodiment 1: Figure 1 As shown, the present invention provides a method for integrating innovation chain and industrial chain based on knowledge graph, and the method includes: Step S100: Collect entities and their relationships in the innovation chain, use a graph neural network model to represent the entities and their relationships in the innovation chain as graph structure data, build an innovation chain knowledge graph, and embed the entities and relationships in the innovation chain knowledge graph.

[0018] Specifically, entities and their relationships in the innovation chain include technological inventions, R&D teams, R&D projects, technology transfer, and achievement transformation. The graph neural network model includes: an input layer for receiving entity and relationship data in the innovation chain or industrial chain; a graph convolution layer for performing convolution operations on graph structure data and extracting graph structure features; a pooling layer for performing dimensionality reduction processing on the output of the graph convolution layer; and an output layer for outputting graph structure data. Construct an innovation chain knowledge graph and use the TransR method to embed entities and relationships in the innovation chain knowledge graph.

[0019] In this embodiment, the entities and their relationship data in the innovation chain are stored in the form of triplets, each of which contains a head entity, a relationship, and a tail entity. Next, a graph neural network model is defined. The model includes an input layer, a graph convolution layer, a pooling layer, and an output layer. The graph convolution layer is used to extract feature information in the graph structure, the pooling layer is used to reduce the feature dimension, and the output layer outputs the final node representation. Specifically, a class named GNNModel is defined, which inherits from nn.Module. In the class initialization method, two graph convolution layers conv1 and conv2, and a pooling layer pool are defined. Among them, the number of input channels of conv1 is the same as the node feature dimension, and the number of output channels is the hidden layer dimension; the number of input channels of conv2 is the same as the hidden layer dimension, and the number of output channels is the output layer dimension. The pooling layer uses adaptive average pooling to reduce the feature dimension to 1. In the forward propagation method of the class, the node feature matrix x and the edge index matrix edge_index in the input data are first obtained. Then, the node feature matrix and the edge index matrix are passed to the first graph convolution layer conv1, and the ReLU activation function is applied. Then, the processed feature matrix is ​​passed to the second graph convolution layer conv2. Finally, the feature matrix is ​​pooled and the final node representation is returned. In addition, a function named triples_to_graph_data is defined to convert the triple data into a graph data format. The function first builds an index map of entities and relationships, then randomly initializes the node feature matrix, and builds an edge list. Finally, the node feature matrix, edge index matrix, and node feature dimension are used as attributes of the graph data object and the object is returned.

[0020] The TransR method is a knowledge graph embedding method that embeds entities and relationships by projecting them into different spaces. In this embodiment, a class called TransRModel is used, which also inherits from nn.Module. In the class initialization method, the entity embedding matrix entity_embedding, the relationship embedding matrix relation_embedding, and the projection matrix projection_matrix are defined. Among them, the dimensions of the entity embedding matrix and the relationship embedding matrix are the same as the number of entities and relationships and the embedding dimension respectively; the input dimension of the projection matrix is ​​the same as the entity embedding dimension, and the output dimension is the projection dimension. In the forward propagation method of the class, the indexes of the head entity, relationship, and tail entity are first obtained, and the corresponding entity embedding and relationship embedding are obtained respectively. Next, the head entity embedding and the tail entity embedding are projected into the relationship space through the projection matrix. Then, the L2 distance between the head entity projection, the relationship embedding, and the tail entity projection is calculated as the score. Initialize the TransR model and set parameters such as the entity dimension, the relationship dimension, and the projection dimension. Then, the model is trained using the optimizer and the L2 loss function.

[0021] Finally, the output of the graph neural network is used as part of the entity embedding to initialize the entity embedding matrix of TransR. First, the trained GNN model is used to forward propagate the graph data to obtain the representation of the node. Then the representation is converted into a tensor and the entity embedding matrix of the TransR model is initialized. Through the above steps, the graph neural network model and TransR method are applied to the processing and analysis of entities and their relationship data in the innovation chain. The feature information in the graph structure is effectively extracted and embedded in the knowledge graph.

[0022] Step S200: Collect entities and their relationships in the industrial chain, use a graph neural network model to represent the entities and their relationships in the industrial chain as graph structure data, build an industrial chain knowledge graph, and embed the entities and relationships in the industrial chain knowledge graph.

[0023] Specifically, the entities and their relationships in the industrial chain include raw material supply, production and manufacturing, logistics and transportation, market sales, and consumer feedback. The process of representing the entities and their relationships in the industrial chain as graph structure data and constructing the industrial chain knowledge graph is the same as the process of representing the entities and their relationships in the innovation chain as graph structure data and constructing the innovation chain knowledge graph in step S100.

[0024] Step S300: According to the connection points between the innovation chain and the industrial chain, the connection between the innovation chain and the industrial chain is represented in the graph structure, and the entity and relationship representations in the innovation chain knowledge graph and the industrial chain knowledge graph are updated.

[0025] Specifically, after identifying the entities and their relationships in the innovation chain and the industrial chain, the potential connection points between the two are identified. These connection points involve the connection between technology research and development and raw material supply, the application of technological innovation in production and manufacturing, the support of logistics and transportation for the distribution of technological products, the satisfaction of consumer needs by technological innovation in market sales, and the guidance of consumer feedback on the direction of technological research and development.

[0026] According to the type of connection point, add corresponding edges or nodes to the graph structure of the graph neural network model. For example, when a technical invention in the innovation chain is related to the production and manufacturing link in the industrial chain, add an edge from the technical invention entity to the production and manufacturing entity in the graph structure to indicate the direct connection between the two. Similarly, if the market sales link in the industrial chain effectively promotes the technical achievements in the innovation chain, add corresponding edges in the graph to reflect this relationship.

[0027] While adding edges or nodes, update the entity and relationship representations in the innovation chain knowledge graph and the industrial chain knowledge graph, including updating the attribute information of the entity, such as the maturity of the technological invention, the process level of the manufacturing link, the data feedback of the market sales, etc., and updating the weight information of the relationship to reflect the importance and closeness of the connection point.

[0028] In this embodiment, the connection points are stored in the form of a dictionary, with the key being the connection point type and the value being the connected entity pair list. For example, the entity pair list corresponding to the connection point type tech_to_production from technology invention to production manufacturing contains the corresponding relationship between technology invention nodes and production manufacturing nodes; the entity pair list corresponding to the connection point type marketing_to_tech from marketing to promotion of technological achievements contains the corresponding relationship between marketing nodes and technology invention nodes.

[0029] Then update the graph structure to add the edges corresponding to the connection points. By traversing the connection point dictionary, add the edges corresponding to the connection points to the graph structure data of the innovation chain and the industrial chain. Finally, use the trained graph neural network model to forward propagate the updated graph structure to obtain the new node representation. Two convolutional layers are used. The first convolutional layer takes the node feature matrix x and the edge index matrix edge_index as input, and outputs the node feature matrix after convolution operation; the second convolutional layer takes the output of the first convolutional layer as input and further performs convolution operation. Through the operation of two convolutional layers, a new node representation is obtained.

[0030] Step S400: Adopt a fusion model based on the attention mechanism, dynamically adjust the importance of the innovation chain and industrial chain information according to the weight distribution of the attention mechanism, fuse the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph, and generate a fused knowledge graph representation.

[0031] Specifically, the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph are input into the encoder layer for encoding; the encoded embedded representation is input into the attention mechanism layer to calculate the attention weights of each entity and its relationship; the importance of the embedded representation is adjusted according to the attention weight to obtain the adjusted embedded representation; the adjusted embedded representation is input into the fusion layer for fusion to generate the fused knowledge graph representation.

[0032] The fusion model based on the attention mechanism includes: an encoder layer, which is used to encode the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph; an attention mechanism layer, which is used to calculate the attention weights of each entity and their relationships in the innovation chain and the industrial chain, and adjust the importance of the embedded representation according to the weight; a fusion layer, which is used to fuse the embedded representations after weight adjustment to generate a fused knowledge graph representation.

[0033] In this embodiment, the encoder layer encodes the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph. Define a class named EncoderLayer, and the forward method implements the forward propagation process of the encoder layer, including using the convolution layer to perform convolution operations on the input features and performing nonlinear transformations through the ReLU activation function to obtain the encoded feature representation.

[0034] In order to calculate the attention weights of each entity and its relationship in the innovation chain and industrial chain, and adjust the importance of the embedded representation according to the weights, a class named AttentionLayer is defined, including a fully connected layer, which is used to calculate the attention score of the input feature. The forward method implements the forward propagation process of the attention mechanism layer, including calculating the attention score of the input feature, applying the softmax function to obtain the attention weights, and adjusting the importance of the embedded representation according to these weights, usually by weighted averaging.

[0035] The function of the fusion layer is to fuse the weighted embedded representations to generate a fused knowledge graph representation. Define a class named FusionLayer, and the forward method implements the forward propagation process of the fusion layer to perform a weighted average of the representations of the two knowledge graphs.

[0036] The above encoder layer, attention mechanism layer and fusion layer are combined to form a complete fusion model. A class named FusionModel is defined, including two encoder layers (used to process the knowledge graphs of the innovation chain and the industrial chain respectively), two attention mechanism layers and a fusion layer. The forward method implements the forward propagation process of the fusion model. The specific steps include encoding the knowledge graphs of the innovation chain and the industrial chain, calculating the attention weights and the adjusted embedding representation, and finally fusing the representations of the two knowledge graphs to obtain the fused knowledge graph representation.

[0037] The specific steps of using the fusion model for fusion are as follows: First, initialize the fusion model, set the input feature dimension in_channels, the hidden layer dimension hidden_channels, and the output dimension (i.e., the fusion representation dimension) out_channels. Then, pass the node feature matrix x1, x2 and the edge index matrix edge_index1, edge_index2 as input to the fusion model. Through the forward propagation process of the fusion model, we can get the fused knowledge graph representation fused_representation, as well as the attention weights attn_weights1 and attn_weights2 of the innovation chain and the industrial chain. Finally, output the fused knowledge graph representation and attention weights.

[0038] Step S500: Based on the integrated knowledge graph, provide support information from upstream and downstream of the industrial chain for technological innovation.

[0039] Specifically, based on the integrated knowledge graph representation, extract support information from the upstream and downstream of the industrial chain; guide technological innovation and achievement transformation based on the support information from the upstream and downstream of the industrial chain; monitor the application effect of technological innovation in the industrial chain, and make optimization adjustments based on feedback information.

[0040] In this embodiment, the various links in the industrial chain are first identified through the integrated knowledge graph, including raw material supply, production and manufacturing, logistics and transportation, market sales, and consumer feedback, as well as the close connection between these links and technological inventions, R&D teams, and R&D projects in the innovation chain.

[0041] In the process of extracting supporting information from upstream and downstream of the industrial chain, we focus on various resources and potential market opportunities required for technological innovation. For example, by analyzing the relationship between the raw material supply link and the technological innovation node in the knowledge graph, we can quickly locate the raw material supplier suitable for the technological innovation, thereby ensuring a stable supply of raw materials for technological innovation. Similarly, by analyzing the relationship between the production and manufacturing links and the technological innovation nodes, we can understand which production and manufacturing links can efficiently transform technological innovation into actual products, thereby promoting the rapid industrialization of technological achievements.

[0042] In the process of guiding technological innovation and achievement transformation, we use the extracted upstream and downstream support information of the industrial chain to provide targeted advice and guidance to the technological innovation team. This includes adjusting the direction and focus of research and development according to market demand and technological development trends in the technology research and development stage; in the achievement transformation stage, we select the most suitable production and marketing strategies to maximize the transformation efficiency of technological achievements and the success rate of industrialization.

[0043] In addition, through real-time monitoring and feedback mechanisms, feedback information from all links of the industrial chain is collected. This feedback information not only helps to evaluate the actual effect and market value of technological innovation, but also provides valuable experience and lessons for subsequent technological innovation and achievement transformation.

[0044] Based on this feedback information, we can further optimize and adjust the strategies and methods of technological innovation and achievement transformation. For example, we can timely adjust the direction and focus of technological research and development, or improve production and marketing strategies in response to problems and bottlenecks that arise during the application of the industrial chain, so as to ensure that technological innovation can continuously and stably promote the development and upgrading of the industrial chain.

[0045] In one embodiment, the method further includes visually displaying the fused knowledge graph so as to intuitively understand the relationship between the innovation chain and the industrial chain.

[0046] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for integrating innovation chain and industrial chain based on knowledge graph, characterized in that: The method comprises: Step S100: Collect entities and their relationships in the innovation chain, use a graph neural network model to represent the entities and their relationships in the innovation chain as graph structure data, build an innovation chain knowledge graph, and embed entities and relationships in the innovation chain knowledge graph; Step S200: Collect entities and their relationships in the industrial chain, use a graph neural network model to represent the entities and their relationships in the industrial chain as graph structure data, build an industrial chain knowledge graph, and embed the entities and relationships in the industrial chain knowledge graph; Step S300: According to the connection points between the innovation chain and the industrial chain, the connection between the innovation chain and the industrial chain is represented in the graph structure, and the entity and relationship representation in the innovation chain knowledge graph and the industrial chain knowledge graph are updated; Step S400: Adopting a fusion model based on the attention mechanism, dynamically adjusting the importance of the innovation chain and the industrial chain information according to the weight distribution of the attention mechanism, fusing the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph, and generating a fused knowledge graph representation; Step S500: Based on the integrated knowledge graph, provide support information from upstream and downstream of the industrial chain for technological innovation.

2. The method for integrating the innovation chain and the industrial chain as claimed in claim 1, characterized in that: The entities and their relationships in the innovation chain include technological inventions, R&D teams, R&D projects, technology transfer and transformation of results; the entities and their relationships in the industrial chain include raw material supply, production and manufacturing, logistics and transportation, marketing and consumer feedback.

3. The method for integrating the innovation chain and the industrial chain as claimed in claim 1, characterized in that: The graph neural network model includes: an input layer for receiving entities and their relationship data in the innovation chain or industrial chain; a graph convolution layer for performing convolution operations on graph structure data and extracting graph structure features; a pooling layer for performing dimensionality reduction processing on the output of the graph convolution layer; and an output layer for outputting graph structure data.

4. The method for integrating the innovation chain and the industrial chain as claimed in claim 1, characterized in that: The embedding method uses the TransR method.

5. The method for integrating the innovation chain and the industrial chain as claimed in claim 1, characterized in that: Step S300 specifically includes: Identify potential connection points based on entities and their relationships in the innovation and industrial chains; According to the type of connection point, corresponding edges or nodes are added to the graph structure to represent the connection between the innovation chain and the industrial chain; Update the entity and relationship representations in the innovation chain knowledge graph and the industrial chain knowledge graph, including updating the attribute information of the entities and the weight information of the relationships.

6. The method for integrating the innovation chain and the industrial chain as claimed in claim 1, characterized in that: The fusion model based on the attention mechanism includes: The encoder layer is used to encode the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph; The attention mechanism layer is used to calculate the attention weights of each entity and its relationship in the innovation chain and industrial chain, and adjust the importance of the embedded representation according to the weights; The fusion layer is used to fuse the weight-adjusted embedding representations to generate a fused knowledge graph representation.

7. The method for integrating the innovation chain and the industrial chain as claimed in claim 6, characterized in that: Step S400 specifically includes: Input the embedded representations in the innovation chain knowledge graph and the industrial chain knowledge graph into the encoder layer for encoding; The encoded embedding representation is input into the attention mechanism layer to calculate the attention weights of each entity and its relations; Adjust the importance of the embedded representation according to the attention weight to obtain the adjusted embedded representation; The adjusted embedding representation is input into the fusion layer for fusion to generate the fused knowledge graph representation.

8. The method for integrating the innovation chain and the industrial chain as claimed in claim 1, characterized in that: Step S500 specifically includes: Extract supporting information from upstream and downstream of the industrial chain based on the fused knowledge graph representation; Guide technological innovation and achievement transformation based on support information from upstream and downstream of the industrial chain; Monitor the application effect of technological innovation in the industrial chain and make optimization adjustments based on feedback information.

9. The method for integrating the innovation chain and the industrial chain as claimed in claim 1, characterized in that: The method also includes visually displaying the integrated knowledge graph so as to intuitively understand the relationship between the innovation chain and the industrial chain.

Images