A method for constructing a semantic virtual geographic environment of wetlands based on hierarchical knowledge graph

Abstract

This application belongs to the intersection of computer technology, geographic information science, and environmental science, and discloses a method for constructing a semantic virtual geographic environment of wetlands based on a hierarchical knowledge graph. The method includes: constructing a data graph, by constructing wetland monitoring entities, converting a multi-source heterogeneous wetland monitoring dataset into a knowledge graph with entities as nodes and semantic relationships as edges; constructing a computational graph, integrating wetland executable computational entities into the knowledge graph, and enabling the computational entities to perform computations and conversions on the entities in the data graph based on algorithmic rules; and constructing a scenario graph, representing a knowledge graph composed of entities of different wetland application scenarios, wherein the application scenario entities associate data entities from the data graph with computational entities in the computational graph. This method can solve the problems of "both too much and too little" caused by the lack of semantic information of wetland monitoring data in existing wetland virtual geographic environments, and "both strong and too weak" caused by the lack of intelligent computational entities.

Description

Technical Field

[0001] The present invention relates to the intersection of computer technology, geographic information science and environmental science, and in particular to virtual geographic environment, semantic technology and knowledge graph technology. In particular, a method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph is disclosed. Background Art

[0002] Wetland ecosystems are extremely sensitive to environmental changes and are characterized by their fragility, dynamism, and complexity. This poses significant challenges to their monitoring, understanding, management, and conservation. New technologies and methods are urgently needed to deepen our understanding of wetlands and enable intelligent monitoring and management.

[0003] Virtual Geographic Environment (VGE), a multidimensional dynamic digital geographic space platform based on geographical theory and integrating virtual reality, network communication, geographic information system and other technologies, has been applied to wetland ecosystem research.

[0004] However, current system-integrated wetland virtual geographic environments are generally "visual" but "unintelligent," making it difficult to meet the needs of wetland ecosystems for computation, management, and decision-making. The main reason is that the current architecture of wetland virtual geographic environments lacks semantic information, leading to the following two prominent contradictions:

[0005] 1. The paradox of "excessive yet scarce" wetland monitoring data: With the development of sensor network technology, the volume of wetland monitoring data has exploded, enabling data integration at the physical level. However, this data is multi-source, heterogeneous, and spans different spatial and temporal scales, lacking semantic connections and forming "logical" data silos. This makes it difficult for both machines (including large language models) and humans (even experts) to deeply understand and reason about this physically integrated data to generate valuable knowledge, thus failing to fully realize the potential benefits of massive data.

[0006] 2. The paradox of wetland computing capabilities being both strong and weak: With the rapid development of geographic modeling platforms and environmental intelligence technologies, the number and types of wetland computing entities (algorithms, models, services) have rapidly increased, and service-based architectures have become mainstream. While service encapsulation facilitates the sharing and reuse of wetland computing entities, these entities still exist as "individuals" and cannot directly interact with the virtual environment. Parameters must be manually configured before execution, and feedback is delayed. Existing execution methods based on service chains or predefined rules are unable to fully adapt to the highly dynamic and unpredictable nature of wetland environments. This makes it difficult to fully consider all possible scenarios during the rule design phase, thus limiting the system's flexibility and intelligence.

[0007] To address these issues, there is an urgent need to establish a mechanism that can transform wetland monitoring data from physical to logical integration and enable computational entities to automatically adapt to dynamic changes in the wetland environment, thereby enhancing the intelligence of wetland virtual geographic environments. The application of knowledge graphs in virtual geographic environments is still in the conceptual exploration and preliminary experimental stages, and their potential for formal modeling and explicit expression has not yet been fully utilized.

[0008] In summary, faced with the two contradictions of "both abundant and scarce" wetland monitoring data and "both strong and weak" wetland computing capabilities in the wetland virtual geographic environment, this paper will explore a knowledge graph-driven wetland semantic virtual geographic environment construction method, and use the knowledge graph to associate wetland monitoring data and wetland computing entities in the form of a graph structure to provide support for wetland ecosystem monitoring and management. Summary of the Invention

[0009] The present invention aims to solve the technical problems of "both too much and too little" caused by the lack of semantic information of wetland monitoring data in the existing wetland virtual geographic environment and "both strong and too weak" caused by insufficient intelligence of computing entities, thereby constructing an intelligent wetland semantic virtual geographic environment.

[0010] In order to solve the above technical problems, the present invention provides a method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph, and its technical solution is:

[0011] Constructing a data graph, a computation graph, and a scenario graph, wherein the data graph is used to integrate wetland monitoring data entities, the computation graph is used to integrate wetland executable computation entities, and the scenario graph is used to represent different wetland application scenario entities;

[0012] Construct a data graph based on wetland monitoring data entities, and transform the multi-source heterogeneous wetland monitoring dataset into a knowledge graph with entities as nodes and semantic relationships as edges;

[0013] Constructing a computational graph, integrating wetland executable computational entities into the knowledge graph, and enabling the computational entities to perform computation and transformation operations on entities in the data graph based on algorithmic rules;

[0014] A scenario graph is constructed to represent a knowledge graph composed of different wetland application scenario entities, where the application scenario entities are associated with data entities from the data graph and computational entities from the computational graph.

[0015] As an optional implementation of the first aspect of the present application, the step of constructing the data graph specifically includes: establishing the wetland monitoring entity, and formally describing the wetland discrete entities, continuous entities and monitoring information based on the existing ontology extension; constructing a hybrid ontology strategy, and realizing the hybrid ontology architecture by adopting a hierarchical semantic modeling method combining global ontology and local ontology; integrating multi-source heterogeneous wetland monitoring data, and realizing heterogeneous data integration through a hybrid integration of materialization strategy and virtualization strategy.

[0016] As an optional implementation method of the first aspect of the present application, in the step of realizing heterogeneous data integration through the hybrid integration of materialization strategy and virtualization strategy, the materialization strategy includes: adopting the materialization strategy for data greater than a preset query analysis threshold, less than a preset change frequency and less than a preset update cost; completing the materialization integration through the ETL process, the ETL process includes collecting data from the input source, mapping the data from the source mode to the target mode and saving the converted data to the triple storage; in the conversion stage, mapping the source data to RDF by constructing RML mapping rules; in the loading stage, using the materialization engine according to the RML mapping rules, the multi-source heterogeneous wetland monitoring data is materialized into RDF triples and persisted in the graph database.

[0017] As an optional implementation of the first aspect of the present application, in the step of realizing heterogeneous data integration through hybrid integration of materialization strategy and virtualization strategy, the virtualization strategy includes: virtualizing and integrating batch real-time monitoring data using an ontology-based data access technology model; providing a unified data view through an intermediary architecture to achieve the goal of not moving or copying data to a unified location; and translating queries on the knowledge graph into queries on the data source in real time online.

[0018] As an optional implementation of the first aspect of the present application, the step of constructing the data graph further includes: creating a federated storage using the federated storage function of the graph database; configuring multiple local materialized or remote virtualized SPARQL endpoints; and integrating multiple virtualized and materialized RDF graphs in one of the knowledge graphs.

[0019] As an optional implementation of the first aspect of the present application, the step of constructing a computational graph specifically includes: designing an execution mode of an executable knowledge graph, the execution mode being divided into synchronous and asynchronous parts; constructing a knowledge graph interaction mediator for the computing entity API-P based on the API-P framework, and realizing job management and monitoring by persisting API-P jobs; developing a knowledge graph job manager, which is responsible for job management and monitoring of the computing entity, and which realizes knowledge graph persistence of the job by sending SPARQL statements to the endpoint of the knowledge graph; designing an input mediation algorithm and an output mediation algorithm to realize data interaction between the computing entity API-P and the knowledge graph, the input mediation algorithm converts the entities of the knowledge graph into a list of input parameters supported by the API-P, and the output mediation algorithm integrates the execution results of the API-P into the knowledge graph.

[0020] As an optional implementation of the first aspect of the present application, the step of constructing a scene graph specifically includes: representing each wetland application scenario entity as a named graph, which is a triple graph and has a uniform resource identifier; using the wetland application scenario entity external information semantic relationship set to define the basic attributes of each wetland application scenario entity and the semantic connection between entities; using the wetland application scenario entity internal structure description set, described in the form of a quadruple, based on the triple information containing the data graph and the computational graph, specifically specifying the URI of the named graph to clarify the data entities, computing entities and their dependencies used by the wetland application scenario entity; using the named graph to realize the separation and version control of complex wetland monitoring data and computing entities.

[0021] In a second aspect, the present application provides a system for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph, including:

[0022] Construct a data graph module to integrate wetland monitoring data entities, including converting multi-source heterogeneous wetland monitoring datasets into a knowledge graph with entities as nodes and semantic relationships as edges based on wetland monitoring data entities;

[0023] Constructing a computational graph module for integrating wetland executable computational entities, including integrating the wetland executable computational entities into the knowledge graph and enabling the computational entities to perform computation and transformation operations on entities in the data graph based on algorithmic rules;

[0024] Construct a scene graph module for representing different wetland application scenario entities; including a knowledge graph representing the composition of different wetland application scenario entities, wherein the application scenario entities are associated with data entities from the data graph and computational entities from the computational graph.

[0025] In a third aspect, an embodiment of the present application provides an electronic device comprising a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein when the program or instruction is executed by the processor, the steps of the method described in the first aspect are implemented.

[0026] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] (1) In terms of data application, the wetland semantic virtual geographic environment can integrate multi-source heterogeneous data from both physical and logical levels with the help of ontology, eliminate the semantic gap, and achieve more efficient integration. It can then be combined with machine learning technology to extract valuable information and knowledge from massive wetland monitoring data, significantly improving the value density and utilization efficiency of the data.

[0029] (2) In terms of adaptive computing, the wetland semantic virtual geographic environment can realize the automatic conversion and matching of computing entity parameters and data entities based on the executable knowledge graph architecture; during operation, it can trace back the data source and computing process through traceability computing, and dynamically adjust the computing process and parameter configuration according to actual conditions, so as to better adapt to the highly dynamic characteristics of wetland ecosystems and improve the intelligence level of computing and the autonomy of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A flowchart of a method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph provided by an embodiment of the present invention;

[0031] Figure 2 Provide an overall design map for the semantic virtual geographical environment of the wetland;

[0032] Figure 3 This is a schematic diagram of the semantic integrated ontology structure;

[0033] Figure 4 Schematic diagram of materialized semantic integration and virtualized semantic integration;

[0034] Figure 5 Schematic diagram of the process of materializing and integrating wetland monitoring data;

[0035] Figure 6 A structural diagram of a wetland semantic virtual geographic environment construction system based on a hierarchical knowledge graph provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0037] The terms "first", "second", etc. in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the application can be implemented in a sequence other than those illustrated or described here. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally represents that the objects associated before and after are in a kind of "or" relationship. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically limited.

[0038] Example 1

[0039] See also Figure 1 , which is a flowchart of a method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph proposed in the first embodiment of the present invention. Figure 2 The overall design framework of the wetland semantic virtual geographic environment constructed by the present invention is shown, the core of which is the layered knowledge graph, including the data graph (G D ), computational graph (G C ) and scene graph (G S ).

[0040] It should be noted that the wetland semantic virtual geographic environment consists of a data layer, a computing layer, and a scene layer, which correspond to integrated data entities, computing entities, and scene entities respectively. The three layers use semantic technology to integrate data entities, computing entities, and scene entities.

[0041] The method for constructing a semantic virtual geographic environment of a wetland comprises the following steps:

[0042] Step 1: Construct a data graph. By constructing wetland monitoring entities, the multi-source heterogeneous wetland monitoring dataset is converted into a knowledge graph with entities as nodes and semantic relationships as edges.

[0043] In some embodiments, multi-source heterogeneous data is converted into an RDF knowledge graph by constructing a wetland monitoring entity. This ontology reuses existing ontologies and is improved based on wetland characteristics to describe discrete entities, continuous entities, and monitoring information. D) adopts a hybrid ontology strategy, with a global ontology defining core semantics and local ontologies describing domain-specific data. Data integration employs a hybrid approach of materialization and virtualization: Highly queried, low-change, and high-update-cost data is materialized into RDF storage through ETL and RML mapping; For large, real-time data, virtualization integration is achieved through the Ontology-Based Data Access (OBDA) model. Finally, federated storage integrates materialized and virtualized endpoints, enabling the integration of multiple virtualized and materialized RDF graphs within a single knowledge graph.

[0044] Specifically, the data graph (G D ) can be defined as a multi-relationship graph ,in Represents a collection of entities related to wetlands or wetland monitoring data. Represents the set of relations between wetland monitoring data entities. Represents a set of facts consisting of entities and relationships. It consists of a series of triples of the form ,in Represent the head and tail entities respectively, Represents a relationship from a head entity to a tail entity. Through clear ontology definitions and formalized semantic relationships, the knowledge graph explicitly expresses the information implicit in wetland monitoring data and organizes the entities and their relationships in the wetland monitoring data in an orderly manner, making the previously scattered information centrally displayed and easy to understand.

[0045] Furthermore, knowledge graphs can leverage embedding techniques to map wetland monitoring data entities and relationships into a low-dimensional vector space. This improves computational efficiency while preserving the semantics of the original data and supports downstream data-driven applications. For RDF graphs, the RDF2Vec graph embedding method can be used. RDF2Vec traverses paths in the RDF graph and utilizes word embedding concepts from natural language processing (such as Word2Vec) to convert entities and relationships in the data graph into low-dimensional vectors. Therefore, combining data graphs with embedding techniques like RDF2Vec makes wetland monitoring data not only visual but also computable.

[0046] From a structural perspective, there are three strategies to achieve data semantic integration and knowledge graph construction: single ontology, multiple ontology, and hybrid ontology. Figure 3 As shown, the data graph is constructed using Figure 3 (A) single entity, such as Figure 3 (B) Multiple Ontologies and Figure 3(C) Three strategic concepts of hybrid ontology. The single ontology method uses a single global ontology, which has a simple structure but is difficult to adapt to the dynamic changes of wetland data; the multi-ontology method uses multiple independent ontologies, which solves the heterogeneity problem but leads to poor data interoperability; the hybrid ontology method innovatively combines the global ontology with the local ontology, establishes a unified semantic framework to define the core elements through the global ontology, and uses the local ontology to describe the data in specific fields, which not only maintains the consistency of the system but also enhances the flexibility, effectively solving the key problem of semantic integration of wetland monitoring data. Therefore, based on this strategy, the global ontology can be used as the public semantic foundation of the knowledge graph to define the basic concepts, attributes, relationships and rules in multi-source heterogeneous wetland monitoring data; the local ontology is used to describe the wetland monitoring data of specific themes, fields, scenarios and needs, thereby enhancing the flexibility and diversity of the knowledge graph.

[0047] In terms of form, Figure 4 Figure 2 shows two hybrid integration strategies used in data graph construction: materialization and virtualization. The materialization approach physically integrates data into the graph database through an ETL process. While this approach offers high query efficiency, it also incurs high storage and maintenance costs, making it particularly unsuitable for large datasets such as raster data. The virtualization approach uses an intermediary architecture to achieve a unified data view, maintaining data storage in situ. While this approach saves storage space and maintenance costs, query performance is significantly impacted by the need to dynamically convert relational data into RDF triples in real time.

[0048] In order to construct a data graph to solve the problem of integrating heterogeneous wetland monitoring data from different fields and multiple sources in a virtual geographic environment, the wetland monitoring data is first analyzed, wetland monitoring entities are established, and the discrete entities, continuous entities and monitoring information in the wetland monitoring data are formally described. The wetland monitoring entity can be represented as ,in , C represents the set of wetland monitoring related classes, D represents the set of wetland monitoring data related data types, P represents the attribute set, and ,in It is an object attribute, and the domain and range are both objects instantiated by the class; It is a data type attribute, the domain is the object instantiated by the class, and the value range is the literal. Representing class relationships Relationship with data types A collection of is a set of class relations in the form of ; is a set of data type relations in the form .

[0049] If the virtual geographic environment is regarded as a container, the discrete entities in the wetland monitoring data are distributed in the container in the form of object models. Based on the summary of the research on existing object models and according to the characteristics of wetland monitoring data, the wetland monitoring data based on discrete entities is represented as a container composed of types ,state ,time ,space ,property , relationship characteristics The six-tuple formed is:

[0050]

[0051] Where, Represents discrete geographic entities in wetland ecosystems. GeoSPARQL provides a standard way to represent geospatial data, including discrete objects such as points, lines, polygons, and the spatial relationships between them. The discrete entities in wetland monitoring data correspond to the GeoSPARQL ontology. Class, that is Therefore, the wetland monitoring entity constructs a semantic representation of discrete entities based on the reuse of GeoSPARQL and the domain characteristics of wetland monitoring data.

[0052] The wetland monitoring dataset can be represented as , where S represents the set of monitored samples and T represents the set of labels. It is the smallest unit of the dataset, representing a monitoring of a certain area or object in the wetland. Each label Represents a key-value pair , Corresponding to wetland properties, Corresponding monitoring results.

[0053] Semantic integration is achieved through a series of mapping rules from wetland monitoring database schema to wetland monitoring data ontology. , the process of converting a known data set D into a structured data graph based on a unified ontology O.

[0054] The present invention adopts a hybrid integration strategy to construct a wetland monitoring data map to balance performance, cost and flexibility. Figure 5The ETL process for materializing wetland monitoring data involves collecting data from input sources (extraction), cleaning and mapping the data from the source schema (the schema of the original data source) to the target schema (transformation), and saving the transformed data to a triple store (loading). During the extraction phase, wetland remote sensing information extraction generally includes tasks such as wetland land cover classification, geographic object detection, and ecological parameter extraction. Structured data tables generally require no further processing. Wetland monitoring data published via a Web API can be processed using the Python Requests library and extracted as hierarchical files in XML or JSON format. For vector data, the GeoPandas library can be used to read vector data in formats such as Shapefile, GeoJSON, and GML and convert it into an in-memory GeoDataFrame. During the transformation phase, source data is mapped to RDF by constructing RML mapping rules. During the loading phase, a materialization engine is used to materialize the heterogeneous, multi-source wetland monitoring data into RDF triples according to the RML mapping rules and persist them in a graph database.

[0055] Step 2: Build a computational graph, integrate wetland executable computational entities into the knowledge graph, and enable the computational entities to perform computation and transformation operations on entities in the data graph based on algorithmic rules.

[0056] In some embodiments, the wetland executable computing entity is integrated into the knowledge graph to form a computing graph (G C The specific implementation steps include: first, designing synchronous and asynchronous execution modes to handle simple and complex computational tasks, respectively; then, developing an API-P interaction mediator based on the PygeoAPI framework to implement persistent job management; then, building a knowledge graph job manager (KGManager) to implement graph-based job storage using SPARQL statements; and finally, designing input / output intermediary algorithms. The former converts knowledge graph entities into API-P input parameters, while the latter integrates computational results into the knowledge graph to complete data graph updates. This solution enables computational entities to automatically execute workflows and adaptively adjust, and can perform computational transformation operations on data graph entities based on algorithmic rules.

[0057] Specifically, the computation graph (G c ) is represented as a knowledge graph that integrates executable computational entities, and each computational entity can perform computation or transformation operations on entities in the knowledge graph based on specific algorithms or rules, thereby enriching or updating the knowledge graph. The computational graph can be viewed as a binary .in Represents a data graph, including a collection of concepts (classes) and wetland monitoring data entity collection . Represents a set of executable computational entities in the knowledge graph. Represents a collection of wetland computing services, each of which can be represented as , Represents constraints on service inputs and outputs. Represents a collection of task entities in the knowledge graph. Each task It's a service Specific implementation examples, Represents the specific input and output data of the task. As the core feature of the semantic virtual geographic environment, the computational graph realizes the functional evolution of the knowledge graph by dynamically integrating wetland computing entities. This technology deeply integrates the computing process with knowledge expression, transforming the static data graph into a dynamic knowledge graph with continuous evolution capabilities. The computing entity has dual capabilities: it can automatically perform tasks such as data processing and simulation operations through workflows, and can adaptively adjust parameters to respond to changes in the wetland environment or virtual environment. The ultimate goal of the computational graph is to develop autonomous computing, which can independently make decisions and perform corresponding operations under specific conditions, and realize intelligent optimization of computing and virtual-real collaboration.

[0058] Step 3: Construct a scenario graph to represent the knowledge graph composed of different wetland application scenario entities. The application scenario entities are associated with the data entities in the data graph and the computational entities in the computational graph.

[0059] In some embodiments, the knowledge graph representing entities of different wetland application scenarios involves a series of data entities from the data graph and computational entities from the computational graph. S ) includes: representing each wetland application scenario entity as a named graph (Named Graph), that is, ,in Represented as a triple graph, for The uniform resource identifier (URI) of the wetland application scenario entity is used. The basic attributes of each wetland application scenario entity (such as time, location, purpose, etc.) and the semantic connections between entities are defined using the semantic relationship set of the external information of the wetland application scenario entity. The internal structure description set of the wetland application scenario entity is described in the form of a quadruple. Based on the triple information containing the data graph and the computation graph, the URI of the named graph is specially specified, thereby clarifying the data entities, computation entities, and their dependencies used by the wetland application scenario entity.

[0060] Specifically, the scene graph (G S ) represents the knowledge graph composed of different wetland application scenario entities, involving a series of data entities from the data graph and computational entities in the computational graph. The application graph can be represented as .in, Represents a collection of wetland application scenario entities. Each wetland application entity Represented as a named graph, that is ,in Represented as a triple graph, for The Uniform Resource Identifier (URI) of the It represents the set of semantic relationships of external information of wetland application scenario entities, including a series of triples that define the basic attributes of each wetland application scenario entity (such as time, location, purpose, etc.) and the semantic connections between entities. Represents the internal structure description set of wetland application scenario entities, described in the form of quadruple, including data graph and calculation graph On the basis of the triple information, the URI of the named graph is specially specified, thereby clarifying the data entities, computing entities, and their dependencies used by the wetland application scenario entities. As a scene management unit in the virtual geographic environment, the named graph effectively manages the wetland monitoring scene through data isolation and version control mechanisms. Each named graph serves as an independent container to store the data entities and computing entities of a specific scene, supporting separate maintenance of the scene and coexistence of multiple versions. When the scene needs to be modified, it can be executed directly in the corresponding named graph to avoid overall system disturbance. This mechanism is particularly suitable for hypothetical analysis, allowing different scenario plans to be constructed in independent named graphs for simulation experiments, ensuring that the data of each plan do not interfere with each other, and significantly improving the maintainability and analysis flexibility of complex wetland monitoring systems.

[0061] Example 2

[0062] See also Figure 6 , shown is a schematic diagram of the structure of a wetland semantic virtual geographic environment construction system based on a hierarchical knowledge graph proposed in the second embodiment of the present application. The system includes the following key modules:

[0063] Constructing a data graph module 100 for integrating wetland monitoring data entities, including converting multi-source heterogeneous wetland monitoring data sets into a knowledge graph with entities as nodes and semantic relationships as edges based on the wetland monitoring data entities;

[0064] Constructing a computational graph module 200 for integrating wetland executable computational entities, including integrating the wetland executable computational entities into the knowledge graph and enabling the computational entities to perform computation and transformation operations on entities in the data graph based on algorithmic rules;

[0065] A scenario graph module 300 is constructed to represent different wetland application scenario entities; it includes a knowledge graph representing different wetland application scenario entities, wherein the application scenario entities are associated with data entities from the data graph and computational entities from the computational graph.

[0066] In the embodiments of the present application, a system for constructing a semantic virtual geographic environment of a wetland based on a hierarchical knowledge graph can be a device, or a component, integrated circuit, or chip in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. For example, the mobile electronic device can be a mobile phone, tablet computer, laptop computer, PDA, in-vehicle electronic device, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc. The non-mobile electronic device can be a server, network attached storage (NAS), personal computer (PC), etc., which is not specifically limited in the embodiments of the present application.

[0067] In the embodiments of the present application, a system for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph can be a device having an operating system. The operating system can be an Android operating system, an iOS operating system, or other possible operating systems, which are not specifically limited in the embodiments of the present application.

[0068] The embodiment of the present application provides a wetland semantic virtual geographic environment construction system based on a hierarchical knowledge graph, which can implement the various processes implemented by a wetland semantic virtual geographic environment construction method based on a hierarchical knowledge graph in the method embodiment. To avoid repetition, they will not be repeated here.

[0069] Optionally, an embodiment of the present application also provides an electronic device, including a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the various processes of the above-mentioned embodiment of the method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph are implemented, and the same technical effect can be achieved. To avoid repetition, they will not be described here.

[0070] An embodiment of the present application also provides a readable storage medium, which stores a program or instruction. When the program or instruction is executed by a processor, the various processes of the embodiment of the above-mentioned method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0071] The processor is the processor in the electronic device described in the above embodiment. The readable storage medium includes a computer-readable storage medium, such as a computer read-only memory (ROM), random access memory (RAM), a magnetic disk, or an optical disk.

[0072] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element. In addition, it should be noted that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the opposite order according to the functions involved. For example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0073] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a more preferred embodiment. Based on this understanding, the technical solution of this application, or the part that contributes to the existing technology, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of this application.

[0074] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

Claims

1. A method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph, characterized in that: include: Constructing a data graph, a computation graph, and a scenario graph, wherein the data graph is used to integrate wetland monitoring data entities, the computation graph is used to integrate wetland executable computation entities, and the scenario graph is used to represent different wetland application scenario entities; Construct a data graph based on wetland monitoring data entities, and transform the multi-source heterogeneous wetland monitoring dataset into a knowledge graph with entities as nodes and semantic relationships as edges. Specifically, this includes: establishing the wetland monitoring entity, and formally describing the wetland discrete entities, continuous entities, and monitoring information based on the existing ontology extension; constructing a hybrid ontology strategy, and realizing a hybrid ontology architecture by adopting a hierarchical semantic modeling method combining global and local ontologies; integrating multi-source heterogeneous wetland monitoring data, and realizing heterogeneous data integration through a hybrid integration of materialization and virtualization strategies; Construct a computational graph, integrate wetland executable computational entities into the knowledge graph, and enable the computational entities to perform computation and conversion operations on entities in the data graph based on algorithmic rules; specifically, the following steps are involved: designing an execution mode for the executable knowledge graph, which is divided into synchronous and asynchronous modes; constructing a knowledge graph interaction mediator for the computational entity API-P based on the API-P framework, and implementing job management and monitoring by persisting API-P jobs; developing a knowledge graph job manager, which is responsible for job management and monitoring of the computational entity, and which implements knowledge graph persistence for the job by sending SPARQL statements to endpoints of the knowledge graph; designing an input mediation algorithm and an output mediation algorithm, and implementing data interaction between the computational entity API-P and the knowledge graph, wherein the input mediation algorithm converts entities in the knowledge graph into a list of input parameters supported by the API-P, and the output mediation algorithm integrates the execution results of the API-P into the knowledge graph; A scenario graph is constructed to represent a knowledge graph composed of different wetland application scenario entities, where the application scenario entities are associated with data entities from the data graph and computational entities from the computational graph.

2. The method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph according to claim 1, characterized in that: In the step of implementing heterogeneous data integration through hybrid integration of materialization strategy and virtualization strategy, the materialization strategy includes: For data with a value greater than the preset query analysis threshold, a change frequency less than the preset frequency, and a cost less than the preset update cost, a materialization strategy is used. Materialized integration is accomplished through an ETL process that includes collecting data from input sources, mapping the data from the source schema to the target schema, and saving the transformed data to a triple store; In the conversion phase, the source data is mapped to RDF by constructing RML mapping rules; During the loading phase, a materialization engine is used to materialize the multi-source heterogeneous wetland monitoring data into RDF triples according to the RML mapping rules, and persist them in the graph database.

3. The method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph according to claim 1, characterized in that: In the step of realizing heterogeneous data integration through hybrid integration of materialization strategy and virtualization strategy, the virtualization strategy includes: For batch real-time monitoring data, virtual integration is performed using ontology-based data access technology. Provide a unified data view through an intermediary architecture, eliminating the need to move or copy data to a unified location; The query on the knowledge graph is translated into a query on the data source in real time online.

4. The method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph according to claim 1, characterized in that: The step of constructing the data graph further includes: Create a federated storage using the federated storage feature of the graph database; Configure multiple local materialized or remote virtualized SPARQL endpoints; A plurality of virtualized and materialized RDF graphs are integrated into one of the knowledge graphs.

5. The method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph according to claim 1, characterized in that: The steps of constructing the scene graph specifically include: Each wetland application scenario entity is represented as a named graph, where the named graph is a triple graph and has a uniform resource identifier; Use the external information semantic relationship set of wetland application scenario entities to define the basic attributes of each wetland application scenario entity and the semantic connections between entities; Use the internal structure description set of the wetland application scenario entity in the form of a four-tuple description. Based on the triple information containing the data graph and the computation graph, the URI of the named graph is specially specified to clarify the data entities, computation entities and their dependencies used by the wetland application scenario entity. The named graph is used to realize the separation and version control of complex wetland monitoring data and computing entities.

6. A wetland semantic virtual geographic environment construction system based on hierarchical knowledge graph, characterized by: include: Construct a data graph module for integrating wetland monitoring data entities, including converting multi-source heterogeneous wetland monitoring data sets into a knowledge graph with entities as nodes and semantic relationships as edges based on wetland monitoring data entities; specifically, establishing the wetland monitoring entity, formally describing wetland discrete entities, continuous entities and monitoring information based on the existing ontology extension; constructing a hybrid ontology strategy, and realizing a hybrid ontology architecture by adopting a hierarchical semantic modeling method combining global and local ontologies; integrating multi-source heterogeneous wetland monitoring data, and realizing heterogeneous data integration through a hybrid integration of materialization strategy and virtualization strategy; Construct a computational graph module for integrating wetland executable computational entities, including integrating wetland executable computational entities into the knowledge graph, and enabling the computational entity to perform computation and conversion operations on entities in the data graph based on algorithmic rules; specifically including: designing an execution mode of the executable knowledge graph, wherein the execution mode is divided into synchronous and asynchronous parts; constructing a knowledge graph interaction mediator for the computational entity API-P based on the API-P framework, and realizing job management and monitoring by persisting API-P jobs; developing a knowledge graph job manager, which is responsible for job management and monitoring of the computational entity, and which realizes knowledge graph persistence of the job by sending SPARQL statements to endpoints of the knowledge graph; designing an input mediation algorithm and an output mediation algorithm to realize data interaction between the computational entity API-P and the knowledge graph, wherein the input mediation algorithm converts entities of the knowledge graph into a list of input parameters supported by the API-P, and the output mediation algorithm integrates the execution results of the API-P into the knowledge graph; Construct a scene graph module for representing different wetland application scenario entities; including a knowledge graph representing the composition of different wetland application scenario entities, wherein the application scenario entities are associated with data entities from the data graph and computational entities from the computational graph.

7. An electronic device, characterized in that: It includes a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph as described in any one of claims 1 to 5 are implemented.

8. A readable storage medium, characterized in that: The readable storage medium stores a program or instruction, and when the program or instruction is executed by the processor, the steps of the method for constructing a wetland semantic virtual geographic environment based on a hierarchical knowledge graph as described in any one of claims 1 to 5 are implemented.

Images