Intelligent question answering system based on space entity information, construction method and question answering method
Through an intelligent question-and-answer system based on spatial entity information, combined with GIS analysis model, the spatial entity knowledge graph is constructed, and the problem of insufficient combination of GIS analysis model and intelligent question-and-answer system in the existing technology is solved, and the support for complex environment analysis and non-professional users are realized. ease of use.
Patent Information
- Application Number
- CN202410892221.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-04
- Publication Date
- 2024-11-15
AI Technical Summary
The existing intelligent question-and-answer system is less combined with professional GIS analysis models in the field of spatial information, which is difficult to support the application needs of complex environment analysis. Especially for users without GIS background, it is difficult to obtain complex geographic analysis through natural language questions. The result.
Using an intelligent question-and-answer system architecture based on spatial entity information, a spatial entity knowledge graph is constructed through spatial entity five-tuple
The organic combination of geographic data and GIS analysis model has been realized, which improves the availability of the system in a roadless environment or in a wild environment, lowers the threshold for users to choose data types, data scope and algorithm models, and breaks the non-professional field users and professional GIS The knowledge barriers between algorithms enhance environmental analysis capabilities.
Smart Images

Figure CN118964630A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of spatial information system data processing technology, and in particular to an intelligent question-answering system based on spatial entity information and a construction method, a question-answering method, a device and a medium thereof. Background Art
[0002] With the rapid development of spatial information technology and the popularization of the Internet, the demand for spatial information is growing, covering a wide range of fields from daily travel route planning to complex environment analysis. As an important form of human-computer interaction, intelligent question-answering systems have been widely used in various scenarios. However, in the application of spatial information, intelligent question-answering systems in related technologies are generally built based on knowledge graphs.
[0003] It first integrates spatial information data such as geographic location data, geographic feature data, map data and text data to build a domain knowledge graph, then uses semantic analysis algorithms such as intent analysis and named entity recognition to extract keywords for questions entered by users, and finally uses the extracted key information to find the best answer from the knowledge graph and feed it back to the user.
[0004] The intelligent question-answering system in the field of spatial information is mainly used in life scenarios such as route planning, geographic knowledge query, and weather query. It is rarely combined with professional GIS analysis models. Therefore, when facing users without GIS background, it is often difficult to obtain results based on complex geographic analysis by asking questions in natural language, and it is difficult to support the application needs of complex environmental analysis. Summary of the invention
[0005] The present invention provides an intelligent question-answering system based on spatial entity information and a construction method, a question-answering method, a device and a medium, which solves the problem in related technologies that the intelligent question-answering system is less integrated with professional GIS analysis models and is difficult to support application requirements of environmental analysis.
[0006] In order to achieve the above objectives, this application adopts the following technical solutions:
[0007] In a first aspect, an intelligent question-answering system based on spatial entity information is provided, wherein the intelligent question-answering system architecture includes a data layer, an engine layer, a service layer, and an application layer;
[0008] The data layer is used to provide data support for the system, including ontology concept data, spatial entity data and behavior feature description data; wherein the ontology concept data defines the top-level concepts, relationships, attributes and classifications in the system; the spatial entity data is an instance formed under the constraints of the ontology concept; the behavior feature description data stores the description information of the spatial entity behavior feature service;
[0009] The engine layer includes an ontology parsing engine, a graph retrieval engine, a semantic understanding engine, a Web service engine and a GIS computing engine, which are used to provide computing power support for the system; wherein the ontology parsing engine is used to parse the ontology file and provide ontology concept query and computing power; the graph retrieval engine is used to provide the retrieval capability for the ontology concept, the spatial entity data and the behavior feature description data; the semantic understanding engine is used to identify the intent of natural language questions, perform slot filling and entity linking; the Web service engine is used to provide the registration, publishing and management capabilities of Web services; the GIS computing engine is used to provide GIS computing and analysis capabilities;
[0010] The service layer is used to provide specific service interfaces and functions, including knowledge graph services, semantic analysis services, behavioral feature services, and spatial computing services;
[0011] The application layer is used to provide an interactive interface based on the data layer, the engine layer and the service layer to realize information interaction of spatial entity data query, behavior feature service query and intelligent question and answer in the spatial information field.
[0012] In a first possible implementation manner of the first aspect, the intelligent question-answering system constructs a spatial entity knowledge graph based on a spatial entity quintuple <entity code, spatial feature, relationship feature, attribute feature, behavior feature>, with spatial entities as nodes and spatial entity relationship features as edges;
[0013] Among them, the entity code is used as the unique identifier of the spatial entity; the spatial characteristics include the spatial reference and spatial morphological data of the spatial entity; the relationship characteristics include the association relationship between the spatial entities described from different aspects; the attribute characteristics are used to describe the inherent attributes and thematic attributes of the spatial entity; the behavioral characteristics are used to describe the dynamic behavioral capabilities of the spatial entity, and the association and connection between the GIS analysis model and the spatial entity are realized through the behavioral characteristics.
[0014] Based on the first possible implementation method of the first aspect, in a second possible implementation method of the first aspect, the Web service engine adopts the Consul architecture as the service registration center, and parses natural language questions through intent analysis and slot filling methods to obtain keyword information; uses the keyword information to retrieve spatial entity data and behavioral feature description data that meet the requirements from the knowledge graph through entity linking, service discovery, and service matching; obtains execution results through service execution, and visualizes the execution results as answers to questions and displays them.
[0015] Based on the second possible implementation of the first aspect, in a third possible implementation of the first aspect, the semantic understanding engine is configured with a predefined intent slot template, and uses rule definition and template matching methods to implement intent analysis and slot filling of natural language questions input by users.
[0016] Based on the second possible implementation of the first aspect, in a fourth possible implementation of the first aspect, natural language questions are vectorized and feature extracted based on machine learning or deep learning methods to complete intent analysis and slot filling.
[0017] In a second aspect, a method for constructing an intelligent question-answering system based on spatial entity information is provided, the method comprising:
[0018] Based on the spatial entity quintuple <entity code, spatial feature, relationship feature, attribute feature, behavior feature>, a spatial entity knowledge graph is constructed with spatial entities as nodes and spatial entity relationship features as edges; wherein the entity code is used as the unique identifier of the spatial entity; the spatial feature includes the spatial reference and spatial morphological data of the spatial entity; the relationship feature includes the association relationship between the spatial entities described from different aspects; the attribute feature is used to describe the inherent attributes and thematic attributes of the spatial entity; the behavior feature is used to describe the dynamic behavior capability of the spatial entity, and the association and connection between the GIS analysis model and the spatial entity is realized through the behavior feature;
[0019] The intelligent question-answering system is constructed based on the spatial entity knowledge graph.
[0020] In a third aspect, a question-answering method is provided, the question-answering method comprising:
[0021] Perform semantic analysis on the input natural language question to obtain structured key information, including question intent, slot information, and corresponding matching slot values;
[0022] Using the slot information and the slot value as keywords, using a graph query engine to query related spatial entity data in a knowledge graph, and linking the spatial entity with the slot value;
[0023] Performing Web service discovery according to the question intent, using the graph search engine to search for behavior feature service description information using the question intent as a keyword, and obtaining a list of candidate behavior feature services;
[0024] According to the question intent and the slot value, a service with the highest matching degree is selected from the candidate behavior feature service list according to a predefined matching rule, and the slot value is bound to an input parameter of the matching service;
[0025] Based on the Web service engine, the service is called according to the bound input parameters, and the service description, the actual input of the service and the execution result are returned to the front end;
[0026] Based on the returned service description and the actual input and execution results of the service, a visual display and dynamic effects are made and output.
[0027] In a first possible implementation manner of the third aspect, the question-and-answer method further includes:
[0028] If the slot value does not match the input parameter, the spatial entity is queried using the slot value as a keyword, and the attribute value or spatial feature data of the spatial entity is bound to the input parameter of the matching service.
[0029] In a fourth aspect, an electronic device is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the question-and-answer method described in the third aspect.
[0030] In a fifth aspect, a readable storage medium is provided, characterized in that a program or instruction is stored on the readable storage medium, and when the program or instruction is executed by a processor, the steps of the question-and-answer method described in the third aspect are implemented.
[0031] The present invention is driven by spatial entity data. Spatial entities are abstractions of geographic objects in the objective world. They are holistic, conform to human cognitive laws, and avoid the destruction of conceptual integrity. Spatial entity data integrates text data, multi-source geographic spatial data (such as terrain data, POI data, place name data, etc.), association relationship data, and GIS analysis model description information. Based on the spatial entity relationship characteristics, a spatial entity knowledge graph is constructed with spatial entities as nodes and association relationships as edges, and an intelligent question-and-answer system is built on this basis, realizing the organic combination of geographic data and GIS analysis models, making up for the shortcoming of the lack of association relationships in traditional GIS. At the same time, the intelligent question-and-answer system is realized based on the spatial entity knowledge graph, maintaining the conceptual integrity of geographic objects, improving the usability of the system in roadless or field environments, and providing a basis for the application of the system in environmental analysis scenarios. Capability support; GIS analysis models are linked to spatial entities through "behavioral characteristics" and published as Web services, which are uniformly managed by the service registration center. The GIS analysis models are encapsulated as Web services and bound to the specified type of spatial entities as the behavioral characteristics of spatial entities, enhancing the system's environmental analysis capabilities; the discovery, matching, selection and execution of Web services are completely driven by the system. In the question-and-answer process, the system is responsible for the automatic selection, parameter binding and execution of GIS analysis models, which simplifies the process of users selecting data types, data ranges, and algorithm models, lowers the threshold for using the system, and breaks the knowledge barriers between non-professional users and professional GIS algorithms. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] Figure 1 A schematic flowchart of a method for constructing an intelligent question-answering system based on spatial entity information provided in an embodiment of the present application;
[0033] Figure 2 A schematic diagram of a spatial entity quintuple feature composition provided in an embodiment of the present application;
[0034] Figure 3 A schematic diagram of the structure of an intelligent question-answering system based on spatial entity information provided in an embodiment of the present application;
[0035] Figure 4 A schematic flow chart of a question-and-answer method provided in an embodiment of the present application;
[0036] Figure 5 A schematic flow chart of another question-and-answer method provided in an embodiment of the present application;
[0037] Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0038] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined purpose, the technical solutions in the embodiments of the present application are clearly described. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present application.
[0039] 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 sequence. It should be understood that the terms used in this way are interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.
[0040] The description of the method flow in the specification of the present application and the steps of the flowchart in the drawings of the present specification do not have to be strictly executed according to the step numbers, and the method steps can be executed in a different order. Moreover, some steps can be omitted, multiple steps can be combined into one step, and / or one step can be decomposed into multiple steps.
[0041] In conjunction with the accompanying drawings and preferred embodiments, an intelligent question-and-answer system based on spatial entity information and its construction method, question-and-answer method, device and medium provided in the embodiments of the present application are described in detail below.
[0042] First, an intelligent question-and-answer system based on spatial entity information and its construction method, question-and-answer method, device and medium application scenarios of an embodiment of the present application are described in detail.
[0043] The intelligent question-answering system built on the knowledge graph constructs a domain knowledge graph by integrating spatial information data such as geographic location data, geographic feature data, map data and text data. It uses semantic analysis algorithms such as intent analysis and named entity recognition to extract keywords for questions input by users. Finally, it uses the extracted key information to find the best answer from the knowledge graph and feed it back to the user.
[0044] This type of intelligent question-answering system focuses on the retrieval of static data in the knowledge base or knowledge graph, and is less integrated with GIS analysis models. The application scenarios are limited by data and are mainly used in life scenarios, such as travel planning and weather inquiries. These applications are highly dependent on data, have poor flexibility, and lack dynamic analysis capabilities. Moreover, due to the fragmented characteristics of elements in geographic spatial data, the answers generated by the system are also fragmented and lack an overall concept. When searching with "Yangtze River" as the keyword, the results returned are all place names containing "Yangtze River", not the Yangtze River in the usual sense.
[0045] GIS (Geographic Information System) is a technical system that, with the support of computer hardware and software systems, collects, stores, manages, calculates, analyzes, displays and describes geographically distributed data in the entire or part of the earth's surface (including the atmosphere).
[0046] Professional GIS analysis models generally have precise requirements on data format and scope. To realize the combination of intelligent question-answering system and GIS analysis model, it is necessary not only to be able to select the appropriate model according to the user's intention, but also to select the data required for model calculation from massive spatial information data.
[0047] The intelligent question-answering system and its construction method, question-answering method, device and medium proposed in the present invention integrate geographic data, knowledge graphs and GIS analysis models into one, solve the limitations of existing intelligent question-answering system application scenarios, realize automatic GIS analysis model selection, parameter binding and execution, and break the knowledge barrier between non-professional field users and professional GIS algorithms.
[0048] See also Figure 1 The embodiment of the present application provides a method for constructing an intelligent question-answering system based on spatial entity information, the method comprising:
[0049] Step A1: Based on the spatial entity quintuple <entity code, spatial feature, relationship feature, attribute feature, behavior feature>, a spatial entity knowledge graph is constructed with spatial entities as nodes and spatial entity relationship features as edges;
[0050] Step A2: construct the intelligent question-answering system based on the spatial entity knowledge graph.
[0051] See also Figure 2, the spatial entity quintuple refers to <entity code, spatial features, relationship features, attribute features, behavioral features>, where the entity code is used as the unique identifier of the spatial entity; the spatial features include the spatial benchmark and spatial morphological data of the spatial entity; the relationship features include the association between spatial entities described from different aspects; the attribute features are used to describe the inherent attributes and thematic attributes of the spatial entity; the behavioral features are used to describe the dynamic behavioral capabilities of the spatial entity, and the GIS analysis model can be associated with the spatial entity through the behavioral features. Behavioral features are an abstraction of the behavior and response of geographic objects in the real world, and are used to describe the dynamic behavioral capabilities of spatial entities. Behavioral features include behavioral descriptions and operation methods. The behavioral description mainly describes the behavioral capabilities or functions of the spatial entity with the behavioral features; the operation method is the specific implementation of the behavioral features, which includes four elements: preconditions, input parameters, output parameters, and operation results. A behavioral feature can contain multiple operation methods.
[0052] Based on the quintuple features of spatial entities, a spatial entity knowledge graph is constructed with spatial entities as nodes and spatial entity relationship features as edges, and an intelligent question-answering system is constructed based on the spatial entity knowledge graph; the system parses the natural language questions input by users through intent analysis and slot filling methods, extracts keyword information according to predefined intent slot templates, and uses the extracted keyword information through entity linking, service discovery and service matching to retrieve spatial entity data and behavioral feature description data that meet user needs from the knowledge graph, and finally obtains the answers to questions and displays them to users through service execution and visualization of the execution results.
[0053] Among them, intent recognition aims to identify the intention or purpose expressed in the user's speech or text. Through intent recognition, the system can determine the user's intention and make corresponding responses, queries or operations.
[0054] Slot filling is used to fill in the key information slots involved in the user's speech. It extracts the required information by identifying the keywords or phrases in the user input and mapping them to the corresponding predefined slots. In the intent recognition task, slot filling technology can help the system understand the user's intent more accurately and complete related operations.
[0055] See also Figure 3 The embodiment of the present application provides an intelligent question-answering system based on spatial entity information. The intelligent question-answering system is based on the five-tuple feature of the spatial entity, and its system architecture is divided into a data layer, an engine layer, a service layer and an application layer from bottom to top.
[0056] The data layer includes ontology concept data, spatial entity data, and behavior feature description data. It provides data support for the intelligent question-answering system. The ontology concept data defines the top-level concepts, relationships, attributes, and classifications. The construction of spatial entity data is constrained by the ontology concept, and spatial entity data can be regarded as an instance of the ontology concept. The behavior feature service description data stores the description information of the spatial entity behavior feature service, that is, the service summary information.
[0057] The engine layer includes ontology parsing engine, graph retrieval engine, semantic understanding engine, Web service engine and GIS computing engine, which provide computing power support for the question-answering system. The ontology parsing engine provides ontology file parsing, ontology concept query and computing power; the graph retrieval engine establishes data query indexes for different data and provides efficient data query capabilities; the semantic understanding engine provides computing power related to semantic understanding, and the intention recognition, slot filling and entity linking of natural language questions are all completed by the semantic understanding engine; the Web service engine provides the registration, publishing and management capabilities of Web services, and Web service discovery, service matching and service execution are all completed in the Web service engine; the GIS computing engine provides GIS computing and analysis capabilities, and the GIS computing involved in the Web service execution process is undertaken by the GIS computing engine.
[0058] See the figure. Specifically, the ontology parsing engine provides the ability to parse ontology files, query and calculate semantic information of ontology concepts, relationships, attributes, etc. Ontology file parsing relies on the rdflib and json libraries in python, providing the ability to parse ontology concepts, relationships, and attribute files in RDF or json format. Neo4j's own query, calculation and reasoning capabilities are used to implement the ontology concept query and calculation function, which mainly realizes the query of ontology concepts and the calculation of semantic relationships between concepts, such as the hierarchical relationship and similarity of concepts.
[0059] The graph search engine provides the ability to search for data such as ontology concepts, spatial entity data, and behavioral feature description information. To improve search efficiency, the graph search engine builds indexes for ontology, spatial entity, and behavioral feature description data to generate index files. Then, the natural language query conditions entered by the user are analyzed through lexical and syntactic analysis to retrieve relevant documents, which are sorted and returned to the user. The key data in the entire search process can be stored as logs and evaluated for analysis to improve search results.
[0060] The semantic understanding engine mainly parses the natural language questions input by users and extracts key information. The rule matching method is used to identify the intent and fill in the slots of the natural language questions input by users. By matching and calculating the user questions with the predefined intent slot templates, the structured text element information is parsed, including the question intent, slot filling information elements, predefined question intent and question parsing results, see Table 1 and Table 2. After the user question is parsed, the entity referents in the parsed element information are linked to the spatial entities. The slot values in the output elements of the question semantic parsing are used as keywords, and the spatial entity retrieval capabilities provided by the graph retrieval engine are used to query, sort and select spatial entity data, generate a candidate spatial entity set, and then the candidate entities are scored and ranked through the entity disambiguation ranking model to achieve entity linking.
[0061] Table 1 Definition of question intention
[0062]
[0063] Table 2 Question semantic analysis
[0064]
[0065]
[0066] The Web service engine provides a series of capabilities such as Web service registration and publishing, service discovery, service semantic matching, and service execution. The Web service engine uses the Consul architecture as the service registration center to implement a series of methods such as service discovery and service management. Web service registration and publishing publishes the behavior feature description information in the service registration center and binds it with the execution of the Web service to achieve the decoupling association between service description and service implementation. After the Web service is published, the description information of the service is constructed into an index file through the graph retrieval engine, which can be retrieved and queried through the retrieval engine. Web service discovery uses the elements output by the semantic understanding engine after parsing the user's questions as keywords to search the knowledge base and discover the candidate Web service set in the knowledge base. Web service matching includes candidate service matching and service parameter binding. The candidate service matching matches the input parameters of the candidate service returned by the service discovery according to the slot information of the question and the weight is calculated according to the degree of grammatical or semantic matching. It is believed that the higher the weight, the higher the degree of matching. The candidate service with the highest weight is selected as the service matching the virtual service. Service parameter binding is performed based on the structured question element information and candidate spatial entity set generated by the semantic parsing engine. The sources of the input parameters include: structured question element information and spatial entity data. The spatial entity data can provide attribute feature data and spatial feature data to the service. After completing the service matching and parameter binding, the service can be executed, and the result of the service execution is returned to the front end for visualization of the result.
[0067] The GIS computing engine provides GIS computing support, including spatial analysis, path analysis, etc. When the Web service engine calls the Web service encapsulated by the GIS analysis model, the underlying calculations involved in the GIS analysis model are completed in the GIS computing engine.
[0068] Please continue to refer to the figure. The service layer includes knowledge graph service, semantic parsing service, behavior feature service and spatial computing service. The knowledge graph service provides access and query of ontology concepts, spatial entity data and behavior feature description data in the knowledge graph. The semantic parsing service provides intent recognition, slot filling and entity linking capabilities, and can parse the natural language questions input by users to extract user intent and key information and generate candidate entity sets. The behavior feature service provides the ability to register, publish, discover, match and execute behavior feature services. The spatial computing service provides spatial computing capabilities such as spatial analysis and path analysis.
[0069] The application layer provides an interactive interface to users based on the data layer, engine layer and service layer, and realizes information interaction of spatial entity data query, behavior feature service query and intelligent question and answer in the field of spatial information. Among them, intelligent question and answer in the field of spatial information can judge the user's intention based on the natural language questions input by the user, and query and match the spatial entity data and behavior feature services based on the key information in the user's question, and finally return the results of the service execution to the user.
[0070] The present invention proposes an intelligent question-answering system based on the quintuple feature of spatial entities. The system is driven by spatial entity data. Spatial entities are abstractions of geographical objects in the objective world. They are holistic, conform to human cognitive laws, and avoid the destruction of conceptual integrity. The spatial entity data integrates text data, multi-source geographic spatial data (such as terrain data, POI data, place name data, etc.), association relationship data, and GIS analysis model description information. Based on the spatial entity relationship characteristics, a spatial entity knowledge graph is constructed with spatial entities as nodes and association relationships as edges, and an intelligent question-answering system is built on this basis. This realizes the organic combination of geographic data and GIS analysis models, and makes up for the shortcoming of the lack of association relationships in traditional GIS. At the same time, the intelligent question-answering system is realized based on the spatial entity knowledge graph, which maintains the conceptual integrity of geographical objects, improves the usability of the system in roadless or field environments, and provides a basis for the application of the system in environmental analysis scenarios. Capability support; GIS analysis models are linked to spatial entities through "behavioral characteristics" and published as Web services, which are uniformly managed by the service registration center. The GIS analysis models are encapsulated as Web services and bound to the specified type of spatial entities as the behavioral characteristics of spatial entities, enhancing the system's environmental analysis capabilities; the discovery, matching, selection and execution of Web services are completely driven by the system. In the question-and-answer process, the system is responsible for the automatic selection, parameter binding and execution of GIS analysis models, which simplifies the process of users selecting data types, data ranges, and algorithm models, lowers the threshold for using the system, and breaks the knowledge barriers between non-professional users and professional GIS algorithms.
[0071] In some possible implementations, in addition to using rule definition and template matching methods to implement intent analysis and slot filling of natural language questions input by users, the system can also use machine learning or deep learning methods to vectorize and extract features from user questions when there is sufficient corpus data for the answers to the questions, thereby completing the tasks of intent analysis and slot filling, reducing the manual work of template design, and improving the generalization of the system.
[0072] See also Figure 4-5 The embodiment of the present application provides a question-answering method, which performs question-answering based on the intelligent question-answering system of the above embodiment, such as Figure 1 As shown, the question-answering method of the embodiment of the present application includes:
[0073] Step S1: semantically parse the input natural language question to obtain structured key information, including question intent, slot information, and corresponding matching slot values.
[0074] For natural language questions input by users, the system can parse the questions according to the predefined intent slot template and convert them into structured key information, namely semantic parsing key information. These key information include:
[0075] The intent of the question, e.g., booking movie tickets, delivering supplies, travel planning, etc.
[0076] Slot information, such as the slot for booking movie tickets is "movie name", "time", etc.
[0077] The corresponding slot values, such as the slot value of "movie name" may be the specific movie name "xxx", and the slot value of "time" may be "this afternoon".
[0078] Step S2: Using the slot information and the slot value as keywords, use the graph query engine to query the relevant spatial entity data in the knowledge graph and link the spatial entity with the slot value.
[0079] The slot information in the semantically parsed key information and the corresponding matching slot values are used as keywords, and the knowledge graph query capability provided by the graph query engine is utilized to query the relevant spatial entity data from the knowledge graph to complete entity linking.
[0080] Step S3: Perform Web service discovery according to the question intent, use the graph search engine to search for behavior feature service description information using the question intent as a keyword, and obtain a list of candidate behavior feature services.
[0081] The service registration center discovers Web services based on the obtained question intent, uses the graph retrieval engine to retrieve the behavior feature service description information with the question intent as the keyword, and obtains a list of candidate behavior feature services.
[0082] Step S4: According to the question intent and the slot value, the service with the highest matching degree is selected from the candidate behavior feature service list according to a predefined matching rule, and the slot value is bound to the input parameter of the matching service.
[0083] The service registration center selects the service with the highest matching degree from the candidate behavior feature service list based on the obtained question intent and slot value and the predefined matching rules, and binds the slot value to the input parameters of the matching service; if the slot value does not match the input parameters, the slot value is used as a keyword to query the spatial entity, and the attribute value or spatial feature data of the spatial entity is bound to the input parameters of the matching service.
[0084] Step S5: Based on the Web service engine, the service is called according to the bound input parameters, and the service description, actual input of the service and the execution result are returned to the front end.
[0085] The Web service engine calls the service based on the bound input parameters and returns the service description, actual input of the service and the execution result to the front end;
[0086] Step S6: Visually display and dynamically display the returned service description and the actual input and execution results of the service and output them.
[0087] The front end presents the returned service description, actual service input, and execution results to the user through visual display and dynamic effect output.
[0088] See also Figure 6 Corresponding to the above-mentioned question-and-answer method embodiment, an embodiment of the present application provides an electronic device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the above-mentioned question-and-answer method embodiment and the various processes of the embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
[0089] The memory 1009 can be used to store software programs and various data. The memory 1009 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, an application program or instructions required for at least one function (such as a sound playback function, an image playback function, etc.), etc. In addition, the memory 1009 may include a volatile memory or a non-volatile memory, or the memory 1009 may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDRSDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a synchronous link dynamic random access memory (SLDRAM) and a direct memory bus random access memory (DRRAM). The memory 1009 in the embodiment of the present application includes but is not limited to these and any other suitable types of memory.
[0090] The processor 1010 may include one or more processing units; optionally, the processor 1010 integrates an application processor and a modem processor, wherein the application processor mainly processes operations related to an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communication signals, such as a baseband processor. It is understandable that the modem processor may not be integrated into the processor 1010.
[0091] Corresponding to the above-mentioned question-and-answer method embodiment, the embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by the processor, the steps of the above-mentioned question-and-answer method embodiment and the various processes of the embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
[0092] The processor is the processor in the electronic device described in the above embodiment of the present application. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.
Claims
1. Intelligent question-answering system based on spatial entity information, characterized by: The intelligent question answering system architecture includes a data layer, an engine layer, a service layer, and an application layer; The data layer is used to provide data support for the system, including ontology concept data, spatial entity data and behavior feature description data; wherein the ontology concept data defines the top-level concepts, relationships, attributes and classifications in the system; the spatial entity data is an instance formed under the constraints of the ontology concept; the behavior feature description data stores the description information of the spatial entity behavior feature service; The engine layer includes an ontology parsing engine, a graph retrieval engine, a semantic understanding engine, a Web service engine and a GIS computing engine, which are used to provide computing power support for the system; wherein the ontology parsing engine is used to parse the ontology file and provide ontology concept query and computing power; the graph retrieval engine is used to provide the retrieval capability for the ontology concept, the spatial entity data and the behavior feature description data; the semantic understanding engine is used to identify the intent of natural language questions, perform slot filling and entity linking; the Web service engine is used to provide the registration, publishing and management capabilities of Web services; the GIS computing engine is used to provide GIS computing and analysis capabilities; The service layer is used to provide specific service interfaces and functions, including knowledge graph services, semantic analysis services, behavioral feature services, and spatial computing services; The application layer is used to provide an interactive interface based on the data layer, the engine layer and the service layer to realize information interaction of spatial entity data query, behavior feature service query and intelligent question and answer in the spatial information field.
2. The intelligent question-answering system according to claim 1, characterized in that: The intelligent question answering system is based on the spatial entity quintuple <entity encoding, spatial features, Relationship characteristics, Attribute features, behavior features>, build a spatial entity knowledge graph with spatial entities as nodes and spatial entity relationship features as edges; Among them, the entity code is used as the unique identifier of the spatial entity; the spatial characteristics include the spatial reference and spatial morphological data of the spatial entity; the relationship characteristics include the association relationship between the spatial entities described from different aspects; the attribute characteristics are used to describe the inherent attributes and thematic attributes of the spatial entity; the behavioral characteristics are used to describe the dynamic behavioral capabilities of the spatial entity, and the association and connection between the GIS analysis model and the spatial entity are realized through the behavioral characteristics.
3. The intelligent question-answering system according to claim 2, characterized in that: The Web service engine uses the Consul architecture as the service registration center, and uses intent analysis and slot filling methods to parse natural language questions to obtain keyword information; through entity linking, service discovery, and service matching, the keyword information is used to retrieve spatial entity data and behavior feature description data that meet the requirements from the knowledge graph; An execution result is obtained through service execution, and the execution result is visualized and displayed as a question and answer answer.
4. The intelligent question-answering system according to claim 3, characterized in that: The semantic understanding engine is configured with predefined intent slot templates, and uses rule definition and template matching methods to implement intent analysis and slot filling of natural language questions input by users.
5. The intelligent question-answering system according to claim 3, characterized in that: Vectorize and extract features of natural language questions based on machine learning or deep learning methods to complete intent analysis and slot filling.
6. A method for constructing an intelligent question-answering system based on spatial entity information, characterized in that: include: Based on spatial entity quintuple <entity encoding, spatial features, Relationship characteristics, Attribute characteristics, Behavior characteristics>, constructing a spatial entity knowledge graph with spatial entities as nodes and spatial entity relationship characteristics as edges; wherein the entity code is used as the unique identifier of the spatial entity; the spatial characteristics include the spatial reference and spatial morphological data of the spatial entity; the relationship characteristics include the association relationship between the spatial entities described from different aspects; the attribute characteristics are used to describe the inherent attributes and thematic attributes of the spatial entity; the behavior characteristics are used to describe the dynamic behavior capabilities of the spatial entity, and the association and connection between the GIS analysis model and the spatial entity is realized through the behavior characteristics; The intelligent question-answering system is constructed based on the spatial entity knowledge graph.
7. A question-answering method, characterized in that: include: Perform semantic analysis on the input natural language question to obtain structured key information, including question intent, slot information, and corresponding matching slot values; Using the slot information and the slot value as keywords, using a graph query engine to query related spatial entity data in a knowledge graph, and linking the spatial entity with the slot value; Perform Web service discovery according to the question intent, use the graph search engine to search for behavior feature service description information using the question intent as a keyword, and obtain a list of candidate behavior feature services; According to the question intent and the slot value, a service with the highest matching degree is selected from the candidate behavior feature service list according to a predefined matching rule, and the slot value is bound to an input parameter of the matching service; Based on the Web service engine, the service is called according to the bound input parameters, and the service description, the actual input of the service and the execution result are returned to the front end; Based on the returned service description and the actual input and execution results of the service, a visual display and dynamic effects are made and output.
8. The question-answering method according to claim 7, characterized in that: Also includes: If the slot value does not match the input parameter, the spatial entity is queried using the slot value as a keyword, and the attribute value or spatial feature data of the spatial entity is bound to the input parameter of the matching service.
9. An electronic device, characterized in that: The electronic device comprises: a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps of the question-answering method as claimed in any one of claims 7 to 8 when executed by the processor.
10. 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 a processor, the steps of the question-answering method as described in any one of claims 7 to 8 are implemented.
Citation Information
Cited By
Geographic knowledge complex question and answer method based on space-time agent architecture
CN120745840A
Aerospace big data intelligent data governance system and method based on four-database collaboration and knowledge generation
CN121834012A
