Medical institution safety production risk reasoning method, device and equipment and storage medium

Abstract

The invention provides a medical institution safety production risk reasoning method, device and equipment and a storage medium, and relates to the technical field of wisdom medical institution, real-time monitoring data is input into a medical institution safety production knowledge graph, an implicit risk relationship is mined in the medical institution safety production knowledge graph, and a time-varying risk conduction weight is calculated; outputting a risk conduction path and a medical institution safety production risk priority based on the time-varying risk conduction weight; wherein the medical institution safety production knowledge graph is constructed based on a triple structure extracted from multi-source data, the triple comprises entities, relationships and attributes, implicit risk association is found through graph reasoning, and an indirect risk conduction path is dynamically output; manual rule writing is not needed, and the method adapts to regulation update and scene change.

Description

Technical Field

[0001] This invention relates to the field of smart healthcare technology, and in particular to a method, apparatus, equipment, and storage medium for risk reasoning in medical institution safety production. Background Technology

[0002] Currently, the application of knowledge graphs in the field of medical institution safety production mainly focuses on static data storage and simple queries, such as searching departmental fire equipment lists, lacking complex logical reasoning capabilities. Reasoning methods are based on rules such as "fire equipment exceeding its service life corresponds to high risk," which struggles to handle dynamically changing risk scenarios. Rule maintenance is costly, it cannot cover all scenarios, and it cannot adapt to logical changes following new regulations or equipment updates. Rule-based inference methods cannot identify implicit relationships, leading to delays in hazard discovery. Furthermore, medical institution safety production data is scattered across unstructured documents such as regulatory documents, inspection records, and accident cases, failing to form a structured knowledge network and lacking the ability to integrate cross-source data, thus hindering the discovery of deep risk correlations. Summary of the Invention

[0003] This invention provides a method, apparatus, equipment, and storage medium for risk reasoning in medical institutions to address the shortcomings of traditional risk reasoning methods in medical institutions, such as complex reasoning, high rule maintenance costs, and delayed hazard detection.

[0004] This invention provides a method for reasoning about safety production risks in medical institutions, comprising: Real-time monitoring data is input into the medical institution safety production knowledge graph, hidden risk relationships are mined in the medical institution safety production knowledge graph, and time-varying risk transmission weights are calculated. Based on the time-varying risk transmission weights, the risk transmission path and the risk priority of medical institution safety production are output; The knowledge graph for safe production in medical institutions is constructed based on a triple structure extracted from multi-source data, wherein the triple includes entities, relations, and attributes.

[0005] According to the medical institution safety production risk reasoning method provided by the present invention, the method for constructing the medical institution safety production knowledge graph includes: Entities and relationships are extracted from multi-source data, including medical regulatory texts, equipment inspection records, and accident reports. The entities and relationships are input into a graph database, with the entities serving as nodes and the relationships as edges, to generate a knowledge graph of safe production in medical institutions. The nodes include departments, equipment, and regulations, while the edges include affiliation, risk association, and spatiotemporal constraints.

[0006] According to the medical institution safety production risk reasoning method provided by the present invention, the step of mining implicit risk relationships in the medical institution safety production knowledge graph and calculating time-varying risk transmission weights includes: The features of the target entity's neighbor nodes are aggregated using a graph neural network model; The time-varying risk transmission weights between entities are calculated based on the characteristics of the target entity's neighboring nodes.

[0007] According to the medical institution safety production risk reasoning method provided by the present invention, the training method of the graph neural network model includes: Using historical accident data as positive samples, negative samples are generated by randomly masking graph edges and node attributes. The positive and negative samples are input into the graph neural network model to obtain the prediction results; A contrastive learning loss function is constructed based on the predicted results and the actual results. The model parameters are then optimized based on the contrastive learning loss function to obtain a trained graph neural network model.

[0008] According to the medical institution safety production risk reasoning method provided by the present invention, the step of calculating the time-varying risk transmission weight between entities based on the features of the target entity's neighbor nodes includes: Obtain the neighboring nodes of the target entity, extract the key features of the neighboring nodes, including device status, historical failure count and regulatory compliance, and encode the key features into a feature vector of uniform dimension through a neural network; Node feature similarity is obtained based on the feature vectors; The spatiotemporal decay factor is calculated based on the degree of deviation between the equipment detection time difference and the standard cycle; The graph attention mechanism is used to calculate the time-varying risk transmission weight between the target entity and each of its neighboring nodes based on the node feature similarity and the spatiotemporal decay factor.

[0009] According to the medical institution safety production risk reasoning method provided by the present invention, the step of outputting the risk transmission path based on the time-varying risk transmission weight includes: The path search algorithm generates multi-hop risk transmission paths based on the time-varying risk transmission weights between the entities.

[0010] According to the medical institution safety production risk reasoning method provided by the present invention, the step of outputting the medical institution safety production risk priority based on the time-varying risk transmission weight includes: The risk characteristics of neighboring nodes are summed up by attention weights, and the sum is then superimposed on the target entity to generate a comprehensive risk score. The risk priority for safe production in medical institutions is output based on the comprehensive risk score.

[0011] This invention also provides a knowledge graph-based device for reasoning about safety production risks in medical institutions, comprising: The mining module is used to input real-time monitoring data into the medical institution safety production knowledge graph, mine hidden risk relationships in the medical institution safety production knowledge graph, and calculate the time-varying risk transmission weights; The output module is used to output the risk transmission path and the risk priority of medical institution safety production based on the time-varying risk transmission weight. The knowledge graph for safe production in medical institutions is constructed based on a triple structure extracted from multi-source data, wherein the triple includes entities, relations, and attributes.

[0012] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the medical institution safety production risk reasoning method as described in any of the preceding claims.

[0013] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the medical institution safety production risk reasoning method described in any of the above claims.

[0014] The present invention provides a method, apparatus, equipment, and storage medium for risk reasoning in medical institution safety production. By inputting real-time monitoring data into a medical institution safety production knowledge graph, it mines implicit risk relationships within the knowledge graph and calculates time-varying risk transmission weights. Based on these weights, it outputs risk transmission paths and safety production risk priorities. The medical institution safety production knowledge graph is constructed based on a triple structure extracted from multi-source data. Each triple includes an entity, a relationship, and an attribute. Implicit risk associations are discovered through graph reasoning, and indirect risk transmission paths are dynamically output. No manual rule writing is required, adapting to regulatory updates and scenario changes. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 This is a flowchart of the medical institution safety production risk reasoning method provided in the embodiments of the present invention; Figure 2This is a functional structure diagram of a knowledge graph-based medical institution safety production risk reasoning device provided in an embodiment of the present invention; Figure 3 This is a functional structure diagram of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0018] Figure 1 A flowchart of the medical institution safety production risk reasoning method provided in the embodiments of the present invention is shown below. Figure 1 As shown, the medical institution safety production risk reasoning method provided in this embodiment of the invention includes: Step 101: Input the real-time monitoring data into the medical institution safety production knowledge graph, mine the hidden risk relationships in the medical institution safety production knowledge graph, and calculate the time-varying risk transmission weight; Step 102: Output the risk transmission path and the risk priority of medical institution safety production based on the time-varying risk transmission weight; The knowledge graph for safe production in medical institutions is constructed based on a triple structure extracted from multi-source data, wherein the triple includes entities, relations, and attributes.

[0019] The application of knowledge graphs in the field of safety production in traditional medical institutions mainly focuses on static data storage and simple queries, such as searching departmental fire equipment lists, lacking complex logical reasoning capabilities. Reasoning methods are rule-based, making it difficult to handle dynamically changing risk scenarios. Rule maintenance is costly, it cannot cover all scenarios, and it cannot adapt to logical changes after new regulations or equipment updates. Rule-based inference methods cannot identify implicit relationships, resulting in delayed hazard discovery. Furthermore, safety production data in medical institutions is scattered across unstructured documents such as regulatory documents, inspection records, and accident cases, failing to form a structured knowledge network and lacking the ability to integrate cross-source data, thus hindering the discovery of deep risk relationships.

[0020] Traditional medical institution safety production systems rely on fixed rule chains for risk assessment. This invention uses a time-varying weighted neural network to adaptively calculate time-varying risk transmission weights and dynamically generate risk transmission paths based on these weights.

[0021] The medical institution safety production risk reasoning method provided in this embodiment of the invention inputs real-time monitoring data into a medical institution safety production knowledge graph, mines implicit risk relationships in the knowledge graph, calculates time-varying risk transmission weights, and outputs risk transmission paths and medical institution safety production risk priorities based on the time-varying risk transmission weights. The medical institution safety production knowledge graph is constructed based on a triple structure extracted from multi-source data, where each triple includes an entity, a relationship, and an attribute. Implicit risk associations are discovered through graph reasoning, and indirect risk transmission paths are dynamically output. No manual rule writing is required, adapting to regulatory updates and scenario changes.

[0022] Based on any of the above embodiments, the method for constructing the medical institution safety production knowledge graph includes: Step 201: Extract entities and relationships from multi-source data, including but not limited to medical regulatory texts, equipment inspection records, and accident reports; Step 202: Input the entities and relationships into the graph database, use the entities as nodes in the graph database, and the relationships as edges in the graph database to generate a knowledge graph of medical institution safety production. The nodes include departments, equipment and regulations, and the edges include affiliation, risk association and spatiotemporal constraint relationships.

[0023] For example, extract entities such as "safety exits" and "fire extinguishers" and relationships such as "must be equipped in" and "regular inspection items" from regulations (such as the "Nine Provisions on Fire Safety Management of Medical Institutions"); store them using graph databases such as Neo4j, where nodes are entities such as departments, equipment, and regulatory clauses, and edges are relationships such as "belong to" and "associated risks".

[0024] In some embodiments of the present invention, in scenarios where medical data is missing (such as new equipment having no historical test records), the embodiments of the present invention employ common sense knowledge graph completion technology for reasoning enhancement: External knowledge injection: Extracting common rules from medical industry knowledge bases (such as clinical guidelines and equipment manuals), for example: "Fire hydrants corrode at a rate 20% faster in humid environments" → Automatically generate hidden edges: (Fire hydrant) - [Environmental corrosion risk] -> (High humidity area) Alternatively, embedding models such as TransE can be used to predict missing relationships, and the completed graph can be used for more comprehensive risk reasoning.

[0025] In some embodiments of the present invention, in order to integrate unstructured data (such as device images and surveillance videos), this embodiment extends the knowledge graph to the multimodal domain: Visual data fusion uses CNN to extract rust and damage features from fire equipment images, transforming them into graph node attributes. A graph attention mechanism is then used to link visual features with structured data (e.g., rust level → failure probability mapping). Real-time behavior recognition (e.g., "materials blocking fire lanes") is performed on surveillance videos, generating dynamic risk event nodes. These nodes are then linked with static graphs to trigger real-time inference, such as "lane blockage + surgery in progress → emergency evacuation risk."

[0026] Based on any of the above embodiments, the step of mining implicit risk relationships in the medical institution's safety production knowledge graph and calculating time-varying risk transmission weights includes: Step 301: Aggregate the features of the target entity's neighbor nodes using a graph neural network model; Step 302: Calculate the time-varying risk transmission weights between entities based on the characteristics of the target entity's neighboring nodes.

[0027] In this embodiment of the invention, the reasoning process for mining implicit risk relationships in the medical institution's safety production knowledge graph includes: (1) Input the data to be inferred (e.g., "fire hydrants in a certain department have not been inspected within the time limit"); (2) Aggregate the features of neighboring nodes (such as the department's historical accident records and the risk levels of adjacent departments) through graph neural network models (such as GraphSAGE); (3) Calculate the risk transmission weights between entities based on the graph attention mechanism to generate the risk probability distribution; (4) Output reasoning results (such as "the fire risk level of this department has been raised to medium level") and related paths (such as "overdue inspection → equipment failure → risk of fire spread").

[0028] This invention discovers hidden risk associations through graph reasoning, such as the indirect risk transmission path between "overloaded operating room circuits" and "design defects in fire escape routes on floors"; it eliminates the need for manual rule writing, automatically learning risk logic through algorithms to adapt to regulatory updates and scenario changes; and it provides visualized risk transmission paths to help managers locate root causes (such as "system loopholes → execution deviations → equipment malfunctions → accidents") and formulate targeted measures.

[0029] Based on any of the above embodiments, the training method of the graph neural network model includes: Step 401: Using historical accident data as positive samples, generate negative samples by randomly masking graph edges and node attributes; Step 402: Input the positive samples and the negative samples into the graph neural network model to obtain the prediction results; Step 403: Construct a contrastive learning loss function based on the predicted results and the actual results, optimize the model parameters based on the contrastive learning loss function, and obtain a trained graph neural network model.

[0030] In some embodiments of the present invention, dynamic risk reasoning can also be implemented based on a fusion of rule engines and machine learning. In risk reasoning for safe production in medical institutions, interpretability and accuracy are equally important. This embodiment introduces an expert rule engine as a constraint condition based on the original graph neural network reasoning model. The specific implementation method is as follows: Expert rule embedding: Transforming the rigid requirements in medical institution safety regulations (such as the "Fire Safety Standard for Medical Institutions") into logical constraints; Constraint-based optimization training: During the training of graph neural network models, rule constraints are introduced through the loss function to ensure that the model output conforms to industry standards.

[0031] Based on any of the above embodiments, the step of calculating the time-varying risk transmission weights between entities based on the features of the target entity's neighbor nodes includes: Step 501: Obtain the neighbor nodes of the target entity, extract the key features of the neighbor nodes, including device status, historical failure count and regulatory compliance, and encode the key features into a feature vector of uniform dimension through a neural network. Step 502: Obtain node feature similarity based on the feature vector; Step 503: Calculate the spatiotemporal decay factor based on the degree of deviation between the equipment detection time difference and the standard cycle; Step 504: Using a graph attention mechanism, calculate the time-varying risk transmission weights between the target entity and each neighboring node based on the node feature similarity and the spatiotemporal decay factor.

[0032] Based on any of the above embodiments, the step of outputting the risk transmission path based on the time-varying risk transmission weight includes: The path search algorithm generates multi-hop risk transmission paths based on the time-varying risk transmission weights between the entities.

[0033] Multi-hop risk transmission paths include, for example, equipment aging → process loopholes → personnel injury (3-hop transmission).

[0034] The path reasoning process includes: Initial event: The sensor detects that the temperature of the operating room power distribution box exceeds the standard (node ​​A); First-step reasoning: Graph association distribution box → power supply line → ventilator (node ​​B), calculate overload risk probability (aggregate electrical load characteristics); The second jump in reasoning: It was discovered that the ventilator's backup battery (node ​​C) was not tested monthly as required by the "Medical Equipment Maintenance Specifications", triggering a system loophole flag; The third jump inference: Combining scheduling data, predicting the risk of patient hypoxia caused by intraoperative power outage + backup failure (node ​​D); Output: Risk path: Overloaded power distribution box → Ventilator power interruption → Backup battery failure → Patient hypoxia risk; Risk level: Emergency (red); Recommended action: Immediately inspect the power distribution box, test the backup battery, and revise the monthly inspection system.

[0035] Based on any of the above embodiments, the step of outputting the safety production risk priority of medical institutions based on the time-varying risk transmission weight includes: Step 601: Sum the risk features of neighboring nodes according to their attention weights, and then overlay the sum onto the target entity to generate a comprehensive risk score. Step 602: Output the corresponding safety production risk priority of the medical institution based on the comprehensive risk score. The safety production risk priority of the medical institution is shown in Table 1.

[0036] Table 1. Safety Risk Priority in Medical Institutions

[0037] The medical institution safety production risk reasoning method provided in this invention constructs a medical institution safety production knowledge graph: integrating multi-source data such as regulations and standards, equipment information, and accident cases to build a knowledge network containing "entity-relationship-attribute" (e.g., entity: fire hydrant, relationship: installed in → department, attribute: inspection time, model); using graph neural networks (GNN) or path search algorithms (e.g., PRA) to mine implicit risk relationships in the knowledge graph (e.g., the association path between "equipment overdue for inspection" and "historical fire accidents"); inputting real-time monitoring data (e.g., equipment status, inspection results), predicting potential risks through graph reasoning, and outputting risk transmission paths and priorities. Through the deep integration of knowledge graphs and dynamic reasoning, a paradigm upgrade of medical institution safety production management from "passive response" to "proactive prevention" is achieved.

[0038] The knowledge graph-based medical institution safety production risk reasoning device provided by the present invention will be described below. The knowledge graph-based medical institution safety production risk reasoning device described below can be referred to in correspondence with the medical institution safety production risk reasoning method described above.

[0039] Figure 2 A schematic diagram of the structure of the knowledge graph-based medical institution safety production risk reasoning device provided in an embodiment of the present invention is shown below. Figure 2As shown, the knowledge graph-based medical institution safety production risk reasoning device provided in this embodiment of the invention includes: The mining module 201 is used to input real-time monitoring data into the medical institution safety production knowledge graph, mine hidden risk relationships in the medical institution safety production knowledge graph, and calculate the time-varying risk transmission weights. Output module 202 is used to output the risk transmission path and the risk priority of medical institution safety production based on the time-varying risk transmission weight; The knowledge graph for safe production in medical institutions is constructed based on a triple structure extracted from multi-source data, wherein the triple includes entities, relations, and attributes.

[0040] The knowledge graph-based risk reasoning device for medical institution safety production provided in this invention inputs real-time monitoring data into a medical institution safety production knowledge graph, mines implicit risk relationships within the knowledge graph, and calculates time-varying risk transmission weights. Based on these time-varying risk transmission weights, it outputs risk transmission paths and safety production risk priorities for medical institutions. The medical institution safety production knowledge graph is constructed based on a triple structure extracted from multi-source data. Each triple includes an entity, a relation, and an attribute. Implicit risk associations are discovered through graph reasoning, and indirect risk transmission paths are dynamically output. No manual rule writing is required, adapting to regulatory updates and scenario changes.

[0041] Figure 3 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 3 As shown, the electronic device may include: a processor 310, a communication interface 320, a memory 330, and a communication bus 340. The processor 310, communication interface 320, and memory 330 communicate with each other via the communication bus 340. The memory 330 includes computer programs, an operating system, and acquired data. The processor 310 can call logical instructions in the memory 330 to execute a medical institution safety production risk reasoning method. This method includes: inputting real-time monitoring data into a medical institution safety production knowledge graph; mining implicit risk relationships in the medical institution safety production knowledge graph; calculating time-varying risk transmission weights; and outputting risk transmission paths and medical institution safety production risk priorities based on the time-varying risk transmission weights. The medical institution safety production knowledge graph is constructed based on a triple structure extracted from multi-source data, where each triple includes an entity, a relation, and an attribute.

[0042] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to related technologies, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0043] On the other hand, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the medical institution safety production risk reasoning method provided by the above methods. The method includes: inputting real-time monitoring data into a medical institution safety production knowledge graph; mining implicit risk relationships in the medical institution safety production knowledge graph; calculating time-varying risk transmission weights; and outputting risk transmission paths and medical institution safety production risk priorities based on the time-varying risk transmission weights. The medical institution safety production knowledge graph is constructed based on a triple structure extracted from multi-source data, wherein the triple includes entities, relationships, and attributes.

[0044] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0045] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of software products. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0046] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for reasoning about safety production risks in medical institutions, characterized in that, include: Real-time monitoring data is input into the medical institution safety production knowledge graph, hidden risk relationships are mined in the medical institution safety production knowledge graph, and time-varying risk transmission weights are calculated. Based on the time-varying risk transmission weights, the risk transmission path and the risk priority of medical institution safety production are output; The knowledge graph for safe production in medical institutions is constructed based on a triple structure extracted from multi-source data, wherein the triple includes entities, relations, and attributes.

2. The method for reasoning about safety production risks in medical institutions according to claim 1, characterized in that, The method for constructing the knowledge graph of medical institution safety production includes: Entities and relationships are extracted from multi-source data, which includes relevant regulations on safety production in medical institutions, business management records, equipment operation and maintenance records, and accident reports. The entities and relationships are input into a graph database, with the entities serving as nodes and the relationships as edges, to generate a knowledge graph of safe production in medical institutions. The nodes include departments, equipment, and regulations, while the edges include affiliation, risk association, and spatiotemporal constraints.

3. The method for reasoning about safety production risks in medical institutions according to claim 1, characterized in that, The process of mining implicit risk relationships in the knowledge graph of safe production in the medical institution and calculating time-varying risk transmission weights includes: The features of the target entity's neighbor nodes are aggregated using a graph neural network model; The time-varying risk transmission weights between entities are calculated based on the characteristics of the target entity's neighboring nodes.

4. The method for reasoning about safety production risks in medical institutions according to claim 3, characterized in that, The training method for the graph neural network model includes: Using historical accident data as positive samples, negative samples are generated by randomly masking graph edges and node attributes. The positive and negative samples are input into the graph neural network model to obtain the prediction results; A contrastive learning loss function is constructed based on the predicted results and the actual results. The model parameters are then optimized based on the contrastive learning loss function to obtain a trained graph neural network model.

5. The method for reasoning about safety production risks in medical institutions according to claim 3, characterized in that, The calculation of time-varying risk transmission weights between entities based on the features of the target entity's neighbor nodes includes: Obtain the neighboring nodes of the target entity, extract the key features of the neighboring nodes, including device status, historical failure count and regulatory compliance, and encode the key features into a feature vector of uniform dimension through a neural network; Node feature similarity is obtained based on the feature vectors; The spatiotemporal decay factor is calculated based on the degree of deviation between the equipment detection time difference and the standard cycle; The graph attention mechanism is used to calculate the time-varying risk transmission weight between the target entity and each of its neighboring nodes based on the node feature similarity and the spatiotemporal decay factor.

6. The method for reasoning about safety production risks in medical institutions according to claim 5, characterized in that, The step of outputting the risk transmission path based on the time-varying risk transmission weight includes: The path search algorithm generates multi-hop risk transmission paths based on the time-varying risk transmission weights between the entities.

7. The method for reasoning about safety production risks in medical institutions according to claim 5, characterized in that, The method of outputting the safety production risk priority of medical institutions based on the time-varying risk transmission weight includes: The risk characteristics of neighboring nodes are summed in a weighted manner according to attention weights, and the weighted sum is superimposed on the target entity to generate a comprehensive risk score. The risk priority for safe production in medical institutions is output based on the comprehensive risk score.

8. A knowledge graph-based device for reasoning about safety production risks in medical institutions, characterized in that, include: The mining module is used to input real-time monitoring data into the medical institution safety production knowledge graph, mine hidden risk relationships in the medical institution safety production knowledge graph, and calculate the time-varying risk transmission weights; The output module is used to output the risk transmission path and the risk priority of medical institution safety production based on the time-varying risk transmission weight. The knowledge graph for safe production in medical institutions is constructed based on a triple structure extracted from multi-source data, wherein the triple includes entities, relations, and attributes.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the medical institution safety production risk reasoning method as described in any one of claims 1 to 7.

10. A non-transitory readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the medical institution safety production risk reasoning method as described in any one of claims 1 to 7.

Images