Voice interaction operation method and system for natural resource investigation monitoring system

Abstract

The invention provides a voice interaction operation method and system for a natural resource investigation monitoring system, and the method comprises the steps: obtaining a voice operation instruction received by a user side of the natural resource investigation monitoring system, and carrying out the recognition and extraction of the voice operation instruction through a pre-constructed improved natural language processing model, obtaining basic elements in the voice operation instruction; querying and replacing a pre-constructed natural resource investigation monitoring knowledge graph through the basic elements to obtain corresponding operation prompts; generating a system executable code for the operation prompt by using a large language model, and outputting a final operation result of the natural resource investigation monitoring system according to the system executable code. According to the automatic interaction operation based on the user voice, natural language processing and the knowledge graph technology are combined, and therefore usability and user friendliness of interaction operation of a natural resource investigation and monitoring professional software system are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of natural language processing, and in particular to a voice interaction operation method and system for a natural resource survey and monitoring system. Background Art

[0002] Building a natural resource survey and monitoring system is an important measure to enhance the modernization of national governance capabilities. In the field of natural resource survey and monitoring system construction, the publicly available software system implementation methods mainly include: "Design and Implementation of Natural Resource Survey and Monitoring Results Analysis and Evaluation Index Calculation System" (Ma Weichun, Kong Li, Surveying and Mapping and Spatial Geographic Information, Issue 6, 2024, pp. 102-105), "Research on the Architecture and Key Technologies of Natural Resource Survey and Monitoring Platform" (Hao Zhilang, Zhao Bo, Yunnan Geographic Environment Research, Issue 1, 2023, pp. 66-70), "Natural Resource Survey and Monitoring Management System ... "Design and Implementation of Natural Resources Survey and Monitoring Platform" (Shu Wenqiang, Xiao Deng, Duan Songjiang, Zhang Shaojia, Wu Fengmin, Qian Jin, Geospatial Information, Vol. 20, No. 12, 2022, pp. 130-136) and others have designed natural resources survey and monitoring system platforms. The architectural design of these system platforms starts from functional design, and has high requirements for system users and operators. Based on the system's query and statistics of natural resources survey and monitoring related results, it is necessary to master the data structure, layer division, page layout, etc. in the system. The user's learning cost is high, the system is not universal, and the interface is not friendly enough.

[0003] In order to improve the intelligence level of natural resource survey and monitoring fields and related professional systems, the disclosed invention patents related to natural resource software systems and knowledge graph industry applications mainly include four types of solutions:

[0004] The first is to improve the intelligent description of data processing. For example, "Method and system for intelligent assimilation of spatiotemporal information of multi-source heterogeneous data of natural resources" (authorization announcement number CN 111680025 B) proposed a data processing and spatial data recognition method based on artificial intelligence; "Knowledge graph geographic anti-fraud system, method and storage medium based on GEOHash algorithm" (application publication number CN 116303847 A) proposed a systematic method for anti-fraud risk control;

[0005] The second is to deepen the retrieval accuracy of knowledge graphs. For example, "A Knowledge Graph Retrieval Method and System Based on Rule Reasoning" (Application Publication No. CN 116484017 A) solves the problem of low accuracy of existing retrieval results; "A Method for Massive Knowledge Graph Retrieval" (Authorization Announcement No. CN 112015911 B) solves the problems of low matching accuracy and poor user experience of current related knowledge graph retrieval methods by calculating the correlation and similarity between knowledge graphs; "A Knowledge Graph Search Path Method Based on Deep Learning" (Authorization Announcement No. CN 112015914 B) proposes a knowledge graph search path method based on deep learning to solve the problem of too many and inaccurate search paths retrieved by existing knowledge graph search path acquisition methods; "A Knowledge Graph Retrieval Method and System Based on Pattern Matching" (Application Publication No. CN117763200 A) studies an intelligent and efficient subgraph matching query method to meet the needs of domain knowledge graphs.

[0006] The third is to expand the application areas of knowledge graphs. For example, "Methods, media and devices for outputting consultation suggestion strategies based on knowledge graph search" (application publication number CN 118245580 A), "Construction of judicial knowledge graph and search method and device based on judicial knowledge graph" (application publication number CN 112749249 A), and "Implementation method of enterprise-level knowledge graph search engine based on electric power professional vocabulary" (application publication number CN 107908738A) respectively introduced knowledge graphs into consultation suggestion strategies, judicial knowledge search, and electric power enterprise knowledge search.

[0007] Fourth, it is combined with artificial intelligence technologies such as natural language processing. For example, "A knowledge graph retrieval method and system integrating pre-trained language model" (application publication number CN 117555985 A) proposes a knowledge graph retrieval method based on Gremlin language integrating pre-trained language model; "Natural resource survey and monitoring method and system based on artificial intelligence" (application publication number CN 118230155 A) proposes an analysis method and system for remote sensing image information extraction in natural resource survey and monitoring; "A natural language interaction method, device, equipment and medium for geographic information system" (authorization announcement number CN 117892828 B) designs a natural language-based interaction method for location information in geographic information system interaction.

[0008] In summary, the disclosed natural resource survey and monitoring systems are usually developed based on geographic information technology and are highly professional. Users must undergo in-depth training on the system's data organization, layer structure, and user interface before they can become proficient in using the system. In addition, conventional natural language processing models cannot process and understand the professional knowledge of natural resource survey and monitoring. Knowledge graphs can provide semantic understanding, knowledge assistance, and knowledge reasoning for conventional natural language processing models, which can solve the problem of lack of factual knowledge in the field of natural resource survey and monitoring, and enhance the understanding and generation capabilities of natural language processing models. None of the above disclosed technologies or inventions involve voice interaction operations of natural resource survey and monitoring systems that couple knowledge graphs and natural language processing, nor do they solve the ease of operation of system platforms and the professionalism of content retrieval in the field of natural resource survey and monitoring. Summary of the invention

[0009] The purpose of the present invention is to provide a voice interaction operation method and system for a natural resource survey and monitoring system, aiming to solve the above-mentioned problems in the prior art.

[0010] The embodiment of the present invention provides a voice interaction operation method for a natural resource survey and monitoring system, comprising:

[0011] Acquire the voice operation instructions received by the user end of the natural resources survey and monitoring system, and use the pre-built improved natural language processing model to recognize and extract the voice operation instructions to obtain the basic elements in the voice operation instructions;

[0012] Query and replace the pre-built natural resource survey and monitoring knowledge graph through the basic elements to obtain corresponding operation prompts; and

[0013] A large language model is used to generate system executable codes for the operation prompts, and the final operation results of the natural resources survey and monitoring system are output according to the system executable codes.

[0014] The embodiment of the present invention provides a voice interaction operating system for a natural resource survey and monitoring system, comprising:

[0015] A speech recognition module is used to obtain the speech operation instructions received by the user end of the natural resources survey and monitoring system, and use the pre-built improved natural language processing model to recognize and extract the speech operation instructions to obtain the basic elements in the speech operation instructions;

[0016] A prompt construction module, used to query and replace the pre-constructed natural resource survey and monitoring knowledge graph through the basic elements to obtain corresponding operation prompts; and

[0017] The result output module is used to generate system executable code for the operation prompt using a large language model, and output the final operation result of the natural resources survey and monitoring system according to the system executable code.

[0018] An embodiment of the present invention also provides an electronic device, 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 above-mentioned voice interaction operation method for a natural resource survey and monitoring system.

[0019] An embodiment of the present invention also provides a computer-readable storage medium, on which a program for implementing information transmission is stored. When the program is executed by a processor, the steps of the above-mentioned voice interaction operation method for a natural resource survey and monitoring system are implemented.

[0020] The use of the embodiments of the present invention may include the following beneficial effects: The embodiments of the present invention combine natural language processing and knowledge graph technology to improve the usability and user-friendliness of the interactive operation of the natural resource survey and monitoring professional software system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate one or more embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0022] Figure 1 is a flow chart of a voice interaction operation method for a natural resources survey and monitoring system according to an embodiment of the present invention;

[0023] Figure 2 It is a flow chart of a voice interaction operation method of a natural resource survey and monitoring system coupled with knowledge graph and natural language processing according to an embodiment of the present invention;

[0024] Figure 3 It is a schematic diagram of the relationship between entities, entity relationships, and interactive operation events of a voice operation text according to an embodiment of the present invention;

[0025] Figure 4 Schematic diagram of the knowledge graph of natural resource survey and monitoring according to an embodiment of the present invention;

[0026] Figure 5 It is a flow chart of entity replacement in the knowledge graph of natural resource survey and monitoring according to an embodiment of the present invention;

[0027] Figure 6It is a schematic diagram of a voice interaction operating system for a natural resource survey and monitoring system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the following will be combined with the drawings in one or more embodiments of this specification to clearly and completely describe the technical solutions in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this document.

[0029] Method Embodiment

[0030] According to an embodiment of the present invention, a voice interaction operation method for a natural resource survey and monitoring system is provided. Figure 1 is a flow chart of a voice interaction operation method for a natural resource survey and monitoring system according to an embodiment of the present invention. Figure 1 As shown, the voice interaction operation method for a natural resource survey and monitoring system according to an embodiment of the present invention specifically includes:

[0031] Step S101, obtaining the voice operation instruction received by the user end of the natural resources survey and monitoring system, using the pre-built improved natural language processing model to identify and extract the voice operation instruction, and obtaining the basic elements in the voice operation instruction, specifically including:

[0032] Convert the voice operation instruction into a text instruction, and input the text instruction into a pre-built improved natural language processing model for intention recognition, entity recognition, entity relationship extraction, and interactive operation event extraction, so as to obtain the operation intention, operation object, operation condition, and operation result requirement in the text instruction;

[0033] The improved natural language processing model is an improved natural language processing model obtained by fine-tuning the natural language processing model through intent samples, entity samples, entity relationship samples and interactive operation event samples used in the natural resources survey and monitoring system;

[0034] Step S102, querying and replacing the pre-built natural resources survey and monitoring knowledge graph through the basic elements to obtain corresponding operation prompts, specifically including:

[0035] The relevant information in the pre-constructed natural resource survey and monitoring knowledge graph is queried through the basic elements, entity replacement, entity relationship replacement and spatial analysis event replacement are performed according to the query results, and corresponding operation prompts are obtained for the replaced information using prompt templates;

[0036] Step S103, generating a system executable code for the operation prompt using a large language model, and outputting a final operation result of the natural resources survey and monitoring system according to the system executable code;

[0037] The operation results are displayed in the form of layer operation display, view operation display, attribute display and statistical chart display.

[0038] The above technical solution of the embodiment of the present invention is described in detail below in combination with the specific situation of the voice interaction operation method for a natural resource survey and monitoring system in the embodiment of the present invention.

[0039] The embodiment of the present invention discloses a human-computer interactive voice operation method for a natural resource survey and monitoring system that couples knowledge graphs and natural language processing, and relates to spatiotemporal information retrieval and human-computer interactive voice operation in natural resource survey and monitoring.

[0040] The specific steps of the embodiment of the present invention include: the system receives voice information input for the operation requirements of the natural resource survey and monitoring system, and converts the voice into text, uses the fine-tuned natural language processing model to perform intent recognition, entity recognition, entity relationship extraction and interactive operation event extraction on the input operation requirements, and then queries the corresponding natural resource survey and monitoring knowledge graph, constructs the prompt words required by the large language model, and generates executable code that can operate the information in the natural resource survey and monitoring system based on the large language model, and finally applies the code to the client of the B / S system to realize the voice operation of the natural resource survey and monitoring system. Through this method, the existing natural resource survey and monitoring systems of various B / S architectures can realize automatic interactive operations based on natural language voice, and improve the ease of use and simplicity of the system.

[0041] In order to realize the natural language interactive operation of professional systems in the field of natural resource survey and monitoring, and at the same time improve the professionalism of natural language interactive operation in knowledge query and retrieval in the field of natural resource survey and monitoring, so that the professional system of natural resource survey and monitoring can directly perform interactive operation through natural language, and ensure that the user experience of the system is improved under the premise of providing professional operation results, the specific implementation method of the human-computer voice interactive operation method for natural resource survey and monitoring system coupled with knowledge graph and natural language proposed in the embodiment of the present invention is as follows:

[0042] The system receives voice operation instructions input in the form of natural language and converts the received voice into text;

[0043] Based on the fine-tuned natural language processing model, the intention of the operation requirements contained in the voice instructions converted into text is recognized;

[0044] Based on the fine-tuned natural language processing model, the recognized voice text is subjected to entity recognition, entity relationship extraction, and interactive operation event extraction to determine the object, conditions, and result requirements in the voice operation instructions;

[0045] According to the identified operation intention, the natural resource survey and monitoring knowledge graph is queried, and then entity replacement, entity relationship replacement, and spatial analysis event replacement are performed to construct a prompt, and the objects, conditions, and result requirements for the natural resource survey and monitoring system expressed in the voice operation instructions are replaced with computer executable commands;

[0046] Generate interactive operation code based on the big language model, input the constructed prompt into the big language model, and generate JavaScript code that can be called and executed by the B / S front end;

[0047] The B / S front end executes the generated JavaScript code, and ultimately feeds back the operation results of the voice input in the form of layer operations, view operations, attribute displays, and statistical charts.

[0048] Furthermore, in order to realize the voice interactive operation of the natural resource survey and monitoring system based on natural language, it is necessary to identify the operation intention of the input voice operation instructions based on the natural language processing model. The operation intention is divided into four categories: layer operation, view operation, attribute display, and statistical charts.

[0049] Furthermore, since conventional natural language processing models cannot directly recognize the strong operational requirements of professional terms input when using the natural resource survey and monitoring system, it is necessary to construct interactive operation intention samples, entity samples, and voice interactive operation event samples of the natural resource survey and monitoring system, and fine-tune or train the natural language processing model, as follows:

[0050] (1) The interactive operation intention samples are constructed by classifying the input operation statements for the natural resources survey and monitoring system according to the operation intention;

[0051] (2) The physical sample contents in the natural resources survey and monitoring system include:

[0052] Layer name, the Chinese name of each vector layer;

[0053] Field name, the Chinese name of the attribute field of all vector layers;

[0054] Contains the feature attribute values ​​of the name, and the attribute values ​​of Chinese attributes in all vector layers.

[0055] (3) Samples of voice interaction operation events used in the natural resources survey and monitoring system include Chinese words used for spatial analysis, such as “open”, “close”, “zoom”, “intersect”, “do not intersect”, “include”, “be included”, “cross”, “overlap”, “boundary connection”, “envelope intersection”, and “index intersection”.

[0056] Furthermore, in order to realize the semantic understanding, knowledge assistance and knowledge reasoning of conventional large language models in the field of natural resources, it is necessary to construct a natural resources survey and monitoring knowledge graph.

[0057] 1. The entities in the natural resources survey and monitoring knowledge graph include:

[0058] Map patches are vector graphics formed by conducting basic surveys of natural resources and special surveys of cultivated land resources, forest resources, grassland resources, wetland resources, water resources, marine resources, underground resources, and surface matrix;

[0059] Administrative divisions, namely, cities, districts and counties (functional areas), towns and streets, and villages;

[0060] Three zones and three lines, namely urban space, agricultural space, and ecological space, as well as the corresponding urban development boundary, permanent basic farmland protection red line, and ecological protection red line;

[0061] Imagery, i.e. remote sensing imagery used for natural resource survey and monitoring;

[0062] An iron tower is an iron tower installed with video equipment used for natural resource survey and monitoring.

[0063] Operation events are requirements for interactive operations.

[0064] 2. The entity relationships in the natural resources survey and monitoring knowledge graph include:

[0065] Interaction-object, used to describe the relationship between interaction actions and interaction objects.

[0066] Interoperation-conditions, used to describe the interoperation requirements.

[0067] Attribute-association, used to describe the restriction or classification conditions of the interactive operation objects, including:

[0068] (1) Pre-phase / post-phase, describing the entity relationship between patches and patches, which can be expressed by a triple: <pattern, pre-phase / post-phase, patch>;

[0069] (2) Include / Located in, describing the entity relationship between the map and administrative division, the map and the three zones and three lines, and the tower and administrative division, which can be represented by triples: <map, located in / included, administrative division>, <map, located in / included, urban space>, <map, located in / included, agricultural space>, <map, located in / included, ecological space>, <map, located in / included, urban development boundary>, <map, located in / included, permanent basic farmland protection red line>, <map, located in / included, ecological protection red line>, <tower, located in / included, administrative division>;

[0070] (3) Identification / identified, which describes the entity relationship between the patch and the image, and can be expressed as a triple: <patch, identification / identified, image>;

[0071] (4) Covering / covered, describing the entity relationship between the patch and the tower, can be represented by a triple: <patch, covered / covering, tower>.

[0072] Furthermore, the voice interaction operation method for the natural resource survey and monitoring system coupled with the knowledge graph and natural language can complete at least four types of interaction operations based on natural language:

[0073] (1) Layer operation, that is, the natural resource survey and monitoring layers can be automatically opened or closed based on natural language operation requirements without the need for human-computer interaction such as a mouse;

[0074] (2) View operation, that is, the positioning, movement, and rotation of the map view can be automatically realized based on natural language operation requirements without the need for mouse or other human-computer interaction;

[0075] (3) Attribute display, that is, the query based on the attribute value of the natural resource survey and monitoring layer and the spatial relationship can be automatically realized based on the operation requirements of natural language without the need for human-computer interaction such as a mouse, and the attributes of the query elements can be displayed;

[0076] (4) Statistical charts, that is, the automatic production of thematic charts such as pie charts, bar charts, and line graphs based on natural language operation requirements without the need for human-computer interaction such as a mouse.

[0077] Traditional natural resource survey and monitoring systems are usually developed based on geographic information technology and are highly professional. Users need in-depth training on the system's data organization, layer structure, and user interface before they can use the system proficiently. In addition, conventional natural language processing models cannot process and understand the professional knowledge of natural resource survey and monitoring. Knowledge graphs can provide semantic understanding, knowledge assistance, and knowledge reasoning for conventional natural language processing models, which can solve the problem of lack of factual knowledge in the field of natural resource survey and monitoring and enhance the understanding and generation capabilities of natural language processing models.

[0078] In order to solve the problems of ease of operation of system platform and professional content retrieval in the field of natural resource survey and monitoring, the embodiment of the present invention combines natural language processing and knowledge graph technology to propose a voice interaction operation method for natural resource survey and monitoring system, and performs human-computer interaction operation of the system based on natural language, which is more in line with people's usage habits, so as to improve the ease of use and user-friendliness of the interactive operation of the natural resource survey and monitoring professional software system.

[0079] like Figure 2 As shown, the embodiment of the present invention discloses a voice interaction operation method for a natural resource survey and monitoring system coupled with a knowledge graph and natural language processing, the steps comprising:

[0080] Step S1, voice input, without keyboard text input or mouse operation, directly voice input the operation content to be performed on the natural resources survey and monitoring system, for example, voice input can be used to input "open three areas and three lines", "zoom to the southern mountainous area", "look at the map spots in the agricultural and rural areas", "statistics of the map spots in 2023 by department";

[0081] Step S2, converting the voice operation instruction input in step S1 into text;

[0082] Step S3, based on the fine-tuned natural language processing model, the voice operation instructions are recognized for their intent, which is divided into four categories:

[0083] "Layer Operation" class, such as "Open three zones and three lines". The interactive operation of the "Layer Operation" class is to open or close various natural resource survey and monitoring layers.

[0084] The "view operation" class, for example, "zoom to the southern mountain area". The interactive operations of the "view operation" class are to realize the zooming in, zooming out, and panning of the map view.

[0085] "Attribute display" class, for example, "Look at the map spots in the agricultural and rural areas". The interactive operation of the "Attribute display" class is to realize the attribute query or spatial relationship query of relevant spatiotemporal information in the natural resources survey and monitoring system, and display the query results.

[0086] "Statistical chart" category, for example, "Spot situation in 2023 under departmental statistics". The interactive operation of the "Statistical chart" category is to realize the summary statistics of spatiotemporal information in the natural resources survey and monitoring system based on classification requirements, which can realize the summary statistics of area and quantity and display the statistical results in the form of pie charts, bar charts, and line charts.

[0087] Since the existing natural language processing model cannot directly identify the intention of highly professional interactive operations regarding the natural resource survey and monitoring system, before performing step S3, it is necessary to generate four interactive operation category samples and fine-tune or train the natural language processing model.

[0088] Taking JSON format as an example, the interactive operation category sample can be expressed as:

[0089]

[0090] Step S4, based on the fine-tuned natural language processing model, entity recognition, entity relationship extraction and voice interaction operation event extraction are performed on the voice interaction operation text used for the natural resource survey and monitoring system. Since the existing natural language processing model cannot directly identify entities, entity relationships and interaction operation events in the field of natural resource survey and monitoring, it is necessary to generate entity, entity relationship and interaction operation event samples first, and fine-tune or train the natural language processing model.

[0091] The natural resources survey and monitoring element entity samples include three categories: layer name, that is, the Chinese name of each vector layer; field name, that is, the Chinese name of the attribute field of all vector layers; element attribute value containing the name, that is, the attribute value containing Chinese attributes in all vector layers. For example, a natural resources survey, monitoring and analysis system contains 4 layers. Using the conventional geographic information system method, the 4 layers are displayed in an attribute table. The attribute structure and content examples are as follows:

[0092] Table 1 Attribute table of each layer

[0093]

[0094]

[0095]

[0096]

[0097]

[0098] Taking JSON format as an example, the entity sample to be constructed can be expressed as:

[0099]

[0100]

[0101]

[0102]

[0103]

[0104] The interactive operation event samples include spatial analysis operation types of attribute query and spatial query. Taking JSON format as an example, they can be expressed as follows:

[0105]

[0106] Among them, "[]" represents an entity, and a line with an arrow represents an entity relationship. The relationship between entities, entity relationships, and interactive operation events in the voice interactive operation text of the natural resource survey and monitoring system can be represented by Figure 3 Express.

[0107] Step S5, based on the recognized entity recognition, entity relationship extraction and interactive operation events, query the natural resources survey and monitoring knowledge graph to perform entity replacement, entity relationship replacement, and spatial analysis event replacement, thereby constructing a prompt, and replacing the objects, conditions and result requirements of the natural resources survey and monitoring system voice operation expressed in natural language with computer executable commands.

[0108] In order to meet the semantic understanding, knowledge assistance and knowledge reasoning of large language models in the field of natural resources, it is necessary to build a natural resources survey and monitoring knowledge graph, such as Figure 4 As shown in the figure, it is an example of the knowledge graph constructed. Taking the voice interaction operation requirement of the natural resources survey and monitoring system "the situation of the spots in 2023 under the statistics of different departments" as an example, the entity replacement steps are as follows: Figure 5 As shown, including:

[0109] (1) Query the knowledge graph and replace the layer entity with the layer name that has the same name (or is equivalent to) the layer name stored in the database. Assume that the layer name after replacement is layerVector;

[0110] (2) Replace the date entity with the "YYYYMM" format. In this example, "2023" is replaced with "2023";

[0111] (3) Query the knowledge graph and replace the attribute condition with "Date == 2023";

[0112] (4) Query the knowledge graph and replace the interactive operation conditions with the field name associated with the layerVector layer's attributes. The replaced field name is QXMC.

[0113] After entity replacement, a prompt is constructed for the natural language processing model to call. In this example, the prompt construction steps include:

[0114] prompt1, "Condition 1: The layer name is layerVector",

[0115] prompt2, "Condition 2: The layer contains the field QXMC",

[0116] prompt3, "Condition 3: The layer contains the field DATE",

[0117] prompt4, "Condition 4: The query condition is DATE==2023",

[0118] prompt5, "Condition 5: Search the layer according to condition 1, and summarize the number of elements that meet condition 4 by QXMC classification",

[0119] prompt6, "Please generate ArcGIS API for JavaScript code according to the above conditions."

[0120] In order to construct prompts more accurately, four different prompt templates can be set according to four interaction operation categories.

[0121] Step S6, generating interactive operation code based on the large language model, inputting the constructed prompt into the large language model, and generating JavaScript code that can be called and executed by the B / S front end. Taking the 6 prompts generated in step S5 as an example, after inputting the large model, the generated JavaScript code is:

[0122]

[0123] Step S7, the voice interactive operation feedback of the natural resource survey and monitoring system, the B / S front end executes the generated JavaScript code, and expresses the interactive operation requirements for the natural resource survey and monitoring system of the feedback voice input in the form of layer operation, view operation, attribute display, and statistical charts. Taking the interactive operation requirements of the natural resource survey and monitoring system in step S6 as an example, the interactive operation results are shown in Table 2:

[0124] Table 2 Statistical chart of interactive operation results

[0125] District and County Name Number of patches Lixia District 23 Licheng District 104 Southern Mountains 56

[0126] System Example

[0127] According to an embodiment of the present invention, a voice interaction operating system for a natural resource survey and monitoring system is provided. Figure 6 Schematic diagram of a voice interactive operating system for a natural resource survey and monitoring system according to an embodiment of the present invention. Figure 6 As shown, the voice interaction operating system for the natural resource survey and monitoring system according to the embodiment of the present invention specifically includes:

[0128] The speech recognition module 60 is used to obtain the speech operation instructions received by the user end of the natural resources survey and monitoring system, and use the pre-built improved natural language processing model to recognize and extract the speech operation instructions to obtain the basic elements in the speech operation instructions, which are specifically used for:

[0129] Convert the voice operation instruction into a text instruction, and input the text instruction into a pre-built improved natural language processing model for intention recognition, entity recognition, entity relationship extraction, and interactive operation event extraction, so as to obtain the operation intention, operation object, operation condition, and operation result requirement in the text instruction;

[0130] The prompt construction module 62 is used to query and replace the pre-constructed natural resource survey and monitoring knowledge graph through the basic elements to obtain corresponding operation prompts, specifically for:

[0131] The relevant information in the pre-constructed natural resource survey and monitoring knowledge graph is queried through the basic elements, entity replacement, entity relationship replacement and spatial analysis event replacement are performed according to the query results, and corresponding operation prompts are obtained for the replaced information using prompt templates;

[0132] A result output module 64 is used to generate a system executable code for the operation prompt using a large language model, and output the final operation result of the natural resources survey and monitoring system according to the system executable code;

[0133] The operation results are displayed in the form of layer operation display, view operation display, attribute display and statistical chart display.

[0134] The above technical solution of the embodiment of the present invention is described in detail below in combination with the specific situation of the voice interactive operating system for the natural resources survey and monitoring system of the embodiment of the present invention.

[0135] Furthermore, the voice interaction operating system for the natural resource survey and monitoring system coupled with the knowledge graph and natural language also includes:

[0136] (1) a voice input module, which is used for the system to receive voice input and convert the voice into text;

[0137] (2) a command execution module, which is used to automatically convert the converted text into executable JavaScript code through a large language model and display the generated code on the front-end page;

[0138] (3) An interactive operation result display module is used to execute the generated commands on the front-end page and automatically realize at least four types of operations, such as opening or closing layers, map view operations, querying attribute values ​​(querying features or specifying layer attributes), and generating thematic charts (automatic generation of pie charts, bar charts, and line charts) without human-computer interaction such as mouse.

[0139] In summary, the embodiments of the present invention have the following beneficial effects:

[0140] 1. The embodiment of the present invention designs a voice interaction operation method for a natural resource survey and monitoring system that couples knowledge graphs and natural language processing, thereby realizing voice retrieval and other interactive operations on spatiotemporal information in the natural resource survey and monitoring system.

[0141] 2. The embodiment of the present invention proposes a fine-tuned natural language processing model, which realizes entity recognition, entity relationship extraction and interactive operation event extraction of voice operation text used in natural resource survey and monitoring systems.

[0142] 3. The embodiment of the present invention proposes a natural resource survey and monitoring knowledge graph, which can assist conventional large language models in achieving semantic understanding, knowledge assistance and knowledge reasoning in the field of natural resources.

[0143] 4. The embodiments of the present invention can realize automatic interactive operation based on natural language voice, and improve the geographic information system (GIS) such as natural resource survey and monitoring. ) The ease of use and simplicity of such special business systems.

[0144] Device Example 1

[0145] An embodiment of the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps described in the method embodiment when executed by the processor.

[0146] Device Example 2

[0147] An embodiment of the present invention provides a computer-readable storage medium, on which a program for implementing information transmission is stored. When the program is executed by a processor, the steps described in the method embodiment are implemented.

[0148] The computer-readable storage medium in this embodiment includes, but is not limited to, ROM, RAM, magnetic disk or optical disk, etc.

[0149] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A voice interactive operation method for a natural resource survey and monitoring system, characterized in that include: Acquire the voice operation instructions received by the user end of the natural resources survey and monitoring system, and use the pre-built improved natural language processing model to recognize and extract the voice operation instructions to obtain the basic elements in the voice operation instructions; The pre-built natural resource survey and monitoring knowledge graph is queried and replaced through the basic elements to obtain corresponding operation prompts; as well as A large language model is used to generate system executable codes for the operation prompts, and the final operation results of the natural resources survey and monitoring system are output according to the system executable codes.

2. The method according to claim 1, characterized in that The pre-built improved natural language processing model is used to identify and extract the voice operation instructions, and the basic elements in the voice operation instructions include: Convert the voice operation instruction into a text instruction, and input the text instruction into a pre-built improved natural language processing model for intention recognition, entity recognition, entity relationship extraction, and interactive operation event extraction, so as to obtain the operation intention, operation object, operation condition, and operation result requirement in the text instruction; Among them, the improved natural language processing model is an improved natural language processing model obtained by fine-tuning the natural language processing model through intent samples, entity samples, entity relationship samples and interactive operation event samples used in the natural resources survey and monitoring system.

3. The method according to claim 1, characterized in that The pre-built natural resources survey and monitoring knowledge graph is queried and replaced through the basic elements, and the corresponding operation prompts are obtained, including: The basic elements are used to query relevant information in the pre-constructed natural resource survey and monitoring knowledge graph, and entity replacement, entity relationship replacement and spatial analysis event replacement are performed according to the query results, and the corresponding operation prompts are obtained for the replaced information using prompt templates.

4. The method according to claim 1, characterized in that: The operation results are displayed in the form of layer operation display, view operation display, attribute display and statistical chart display.

5. A voice interactive operating system for a natural resource survey and monitoring system, characterized in that include: A speech recognition module is used to obtain the speech operation instructions received by the user end of the natural resources survey and monitoring system, and use the pre-built improved natural language processing model to recognize and extract the speech operation instructions to obtain the basic elements in the speech operation instructions; A prompt construction module is used to query and replace the pre-constructed natural resource survey and monitoring knowledge graph through the basic elements to obtain corresponding operation prompts; as well as The result output module is used to generate system executable code for the operation prompt using a large language model, and output the final operation result of the natural resources survey and monitoring system according to the system executable code.

6. The system according to claim 5, characterized in that The speech recognition module is specifically used for: The voice operation instructions are converted into text instructions, and the text instructions are input into a pre-built improved natural language processing model for intent recognition, entity recognition, entity relationship extraction and interactive operation event extraction to obtain the operation intention, operation object, operation condition and operation result requirements in the text instructions.

7. The system according to claim 5, characterized in that The prompt building module is specifically used for: The basic elements are used to query relevant information in the pre-constructed natural resource survey and monitoring knowledge graph, and entity replacement, entity relationship replacement and spatial analysis event replacement are performed according to the query results, and the corresponding operation prompts are obtained for the replaced information using prompt templates.

8. The system according to claim 5, characterized in that The operation results are displayed in the form of layer operation display, view operation display, attribute display and statistical chart display.

9. An electronic device, characterized in that: include: 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 voice interaction operation method for a natural resource survey and monitoring system as described in any one of claims 1 to 4.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an implementation program for information transmission, and when the program is executed by the processor, the steps of the voice interaction operation method for a natural resource survey and monitoring system as described in any one of claims 1 to 4 are implemented.

Images