GIS buffer generation method and system based on fuzzy semantics and geographic constraints

Abstract

The invention provides a GIS buffer generation method and system based on fuzzy semantics and geographic constraints, and belongs to the technical field of geographic information systems.The method comprises the steps that fuzzy descriptors, membership values and geographic entity types are input into a pre-trained language-space mapping model, and basic distance parameters are obtained; querying a spatial database based on the geographic entity keywords, obtaining geometric features associated with the geographic entities, and calculating feature lengths corresponding to the geometric features; correcting the basic distance parameter based on the intention direction, the membership value, the geographic entity type and the feature length expressed by the fuzzy description word to obtain a geographic distance threshold; generating an initial two-dimensional buffer polygon based on the geographic distance threshold and the spatial geometric features of the geographic entity; and querying a digital elevation model to obtain topographic data based on the initial two-dimensional buffer region polygon, and generating a three-dimensional buffer region curved surface through a gradient sensitive interpolation algorithm. According to the invention, the accuracy of generating the GIS buffer area is improved.

Description

Technical Field

[0001] The present application relates to the field of geographic information system technology, and in particular to a GIS buffer generation method and system based on fuzzy semantics and geographic constraints. Background Art

[0002] Buffer generation technology is crucial in GIS spatial analysis, effectively defining the influence of geographic entities and widely used in numerous fields, including urban planning and resource management. Existing buffer generation methods struggle with fuzzy semantics expressed in natural language, such as "nearby" and "surrounding." Because these methods rely heavily on precise distance parameters, they struggle to translate fuzzy semantics into precise spatial parameters. This creates a significant semantic gap between user needs and system execution, severely hindering efficient human-computer interaction.

[0003] At the same time, existing methods often overlook key factors such as the scale of geographic entities and topographical characteristics. Different geographic entities, such as roads of varying lengths and cities of varying sizes, have significant scale variations. Diverse terrain, such as mountainous areas and plains, significantly impacts the actual coverage of buffer zones. However, existing methods fail to fully account for these factors when generating buffer zones, resulting in significant deviations from the actual geographic landscape. With the growing demand for intelligent interaction, accurately converting fuzzy semantics into precise spatial parameters and fully integrating geographic constraints to generate more realistic buffer zones has become a core and critical issue in improving the effectiveness of GIS applications. Therefore, a GIS buffer generation method and system based on fuzzy semantics and geographic constraints is needed to better understand user natural language commands and improve the accuracy of GIS buffer generation. Summary of the Invention

[0004] In order to solve the above technical problems, the present application provides a GIS buffer generation method and system based on fuzzy semantics and geographic constraints.

[0005] A first aspect of an embodiment of the present application provides a GIS buffer generation method based on fuzzy semantics and geographic constraints, comprising: Extracting geographic entity keywords, geographic entity types, and fuzzy descriptive words from the user's natural language instructions through a semantic parsing engine, and parsing the fuzzy descriptive words to obtain membership values ​​corresponding to the fuzzy descriptive words; Inputting the fuzzy descriptor, the membership value, and the geographic entity type into a pre-trained language-space mapping model to obtain a basic distance parameter, wherein the basic distance parameter is a reference value with a distance dimension or a scaling factor relative to a typical scale of an entity; Querying a spatial database based on the geographic entity keyword to obtain geometric features associated with the geographic entity, and calculating feature lengths corresponding to the geometric features; The basic distance parameter is modified based on the intention direction, membership value, geographic entity type and feature length expressed by the fuzzy descriptor to obtain a geographic distance threshold; generating an initial two-dimensional buffer polygon based on the geographic distance threshold and the spatial geometric characteristics of the geographic entity; Based on the initial two-dimensional buffer polygon, a digital elevation model is queried to obtain terrain data, and a three-dimensional buffer surface is generated through a slope-sensitive interpolation algorithm.

[0006] A second aspect of the embodiments of the present application provides a GIS buffer generation system based on fuzzy semantics and geographic constraints, including: A semantic parsing module is used to extract geographic entity keywords, geographic entity types and fuzzy descriptive words from the user's natural language instructions through a semantic parsing engine, and parse the fuzzy descriptive words to obtain membership values ​​corresponding to the fuzzy descriptive words; a data mapping module, configured to input the fuzzy descriptor, the membership value, and the geographic entity type into a pre-trained language-space mapping model to obtain a basic distance parameter, wherein the basic distance parameter is a reference value having a distance dimension or a scaling factor relative to a typical scale of an entity; A data calculation module, configured to query a spatial database based on the geographic entity keywords, obtain geometric features associated with the geographic entity, and calculate feature lengths corresponding to the geometric features; A data correction module is used to correct the basic distance parameter based on the intention direction, membership value, geographic entity type and feature length expressed by the fuzzy descriptor to obtain a geographic distance threshold; A first generating module is configured to generate an initial two-dimensional buffer polygon based on the geographic distance threshold and the spatial geometric characteristics of the geographic entity; The second generating module is used to query the digital elevation model to obtain terrain data based on the initial two-dimensional buffer polygon, and generate a three-dimensional buffer surface through a slope-sensitive interpolation algorithm.

[0007] A third aspect of an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of the above-mentioned GIS buffer generation method based on fuzzy semantics and geographic constraints are implemented.

[0008] In a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned GIS buffer generation system based on fuzzy semantics and geographic constraints are implemented.

[0009] The beneficial effects of a GIS buffer generation method and system based on fuzzy semantics and geographic constraints provided by an embodiment of the present application are as follows: the present application extracts fuzzy descriptive words in natural language and parses membership values ​​through a semantic parsing engine, and obtains basic distance parameters through a pre-trained language-space mapping model, effectively filling the gap between fuzzy semantics and precise spatial parameters, and improving the accuracy and efficiency of human-computer interaction. Secondly, the characteristic length is calculated based on the geometric characteristics of the geographic entity, and the scale correction coefficient is obtained through a scale correction function, so that the geographic distance threshold is fully adapted to the entity's own scale, reducing the result deviation caused by ignoring entity differences in traditional methods. A three-dimensional buffer surface is generated based on a digital elevation model and a terrain adaptive algorithm, taking into account terrain factors, so that the buffer is more in line with the real geographic scene, significantly improving the application efficiency of GIS in urban planning, resource management and other fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A flowchart of a GIS buffer generation method based on fuzzy semantics and geographic constraints provided in one embodiment of the present application; Figure 2 A structural block diagram of a GIS buffer generation system based on fuzzy semantics and geographic constraints provided in one embodiment of the present application; Figure 3 A schematic block diagram of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0011] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0012] To make the purpose, technical solutions and advantages of this application clearer, Figure 1-3 The following description will be given through specific examples.

[0013] Please refer to Figure 1 , Figure 1 A flowchart of a method for generating a GIS buffer zone based on fuzzy semantics and geographic constraints provided in one embodiment of the present application is provided. The method includes: S101: extracting geographic entity keywords, geographic entity types, and fuzzy descriptive words from a user's natural language instruction through a semantic parsing engine, and parsing the fuzzy descriptive words to obtain membership values ​​corresponding to the fuzzy descriptive words; In this embodiment, the semantic parsing engine includes a bidirectional long short-term memory network and a conditional random field model, which can extract geographic entity keywords, geographic entity types and fuzzy descriptors; among them, geographic entity keywords, such as XX river and YY mountain, etc.; geographic entity types, such as linear water bodies and surface terrain, etc.; fuzzy descriptors, such as nearby, surrounding, distant areas, etc.

[0014] Based on a preset fuzzy semantic dictionary, a membership function is used to calculate the membership value corresponding to the fuzzy descriptor, where the membership value ranges from 0 to 1, for example, "very close" corresponds to 0.9 and "far" corresponds to 0.3. The preset fuzzy semantic dictionary includes common fuzzy spatial terms in real life.

[0015] S102: Inputting the fuzzy descriptor, the membership value, and the geographic entity type into a pre-trained language-space mapping model to obtain a basic distance parameter, which is a reference value with a distance dimension or a scaling factor relative to a typical scale of the entity; In this embodiment, the language-space mapping model adopts a neural network model based on the Transformer architecture, wherein the input layer is a fusion module for receiving fuzzy descriptors, membership values, and geographic entity types; the output layer is a dual-channel structure, the first channel outputs the absolute distance value of the basic distance parameter, and the second channel outputs the scaling factor of the basic distance parameter.

[0016] The language-space mapping model of this embodiment is trained using a large-scale geographic text-space correspondence dataset during the pre-training phase. The training samples include: fuzzy spatial relationship phrases described in natural language, corresponding geographic entity types, and annotated real basic distance parameters (base distance values ​​or base scale factors determined based on the context and geographic background).

[0017] The training process adopts a weighted loss function: for samples predicted as absolute distance values, the mean square error between the predicted distance and the true distance is calculated; for samples predicted as scale factors, the mean square error between the predicted scale factor and the true scale factor is calculated, and the feature length of the geographic entity corresponding to the sample is additionally considered as a weight factor to amplify the impact of scale error at larger entity scales.

[0018] In this embodiment, geographic entity types are mapped into high-dimensional embedding vectors before being input into the language-space mapping model; membership values ​​are normalized to the interval [0, 1]; and feature lengths are logarithmically transformed or standardized before being input into the scale correction function to improve the stability and generalization ability of model training.

[0019] During reasoning applications, the language-space mapping model of this embodiment selects to activate either the first channel (outputting absolute distance) or the second channel (outputting scaling factors) as the final base distance parameter based on predefined rules or the model's internal gating mechanism. This selection logic is based on the inherent characteristics of the geographic entity type (e.g., point entities tend to use absolute distances, while line or area entities tend to use scaling factors) or implicit preferences expressed in user instructions.

[0020] In this embodiment, the fusion module uses a multi-head attention mechanism to perform weighted fusion of fuzzy descriptors, membership values, and geographic entity types, including: The fuzzy descriptors, membership scalars, and geographic entity types are concatenated into a joint input vector. The association weights of the fuzzy descriptors, membership scalars, and geographic entity types are calculated separately through a multi-head self-attention layer. The original joint input vector is superimposed after the output of the attention layer to form a residual structure to retain the original feature information. A learnable gated linear unit (GLU) is used to dynamically adjust the contribution ratio of word embedding and category embedding to the final fusion feature.

[0021] S103: querying a spatial database based on geographic entity keywords, obtaining geometric features associated with the geographic entity, and calculating feature lengths corresponding to the geometric features; In this embodiment, the spatial database utilizes a spatiotemporal index structure (supporting spatial data engines such as PostGIS and Oracle Spatial), enabling millisecond-level geometric feature retrieval. Geometric feature extraction includes a topological relationship verification module that automatically corrects abnormal nodes. This embodiment calculates feature lengths using a spatial analysis algorithm: for point entities, the equivalent radius is calculated within the coordinate accuracy range; for line entities, the arc length or total length of the polyline is calculated; and for face entities, the feature value of the longest axis length or boundary perimeter is extracted. The spatial reference system of the entity is also recorded to ensure consistency in subsequent distance calculations.

[0022] S104: Modifying the basic distance parameter based on the intention direction, membership value, geographic entity type, and feature length expressed by the fuzzy descriptor to obtain a geographic distance threshold; In this embodiment, the intended direction is confirmed through directional word vector encoding, and the distance weight is adjusted using a directional attenuation factor. The scale correction function of this embodiment can be designed as a piecewise function. For example, when the feature length is less than 100 meters, linear correction is used; when the feature length is between 100-1000 meters, exponential correction is used; when the feature length is greater than 1000 meters, logarithmic correction is used. A dynamic weight matrix is ​​added when weighting the feature length, and the proportion of the base distance and the correction factor is allocated according to the entity type.

[0023] S105: generating an initial two-dimensional buffer polygon based on the geographic distance threshold and the spatial geometric characteristics of the geographic entity; In this embodiment, an offset algorithm is used to address sharp corners in line entity buffers, generating an initial two-dimensional buffer and improving offset accuracy. The surface entity buffer supports bidirectional expansion, both inward and outward, and a buffer fusion algorithm is used to handle overlapping areas (for example, Boolean operations on adjacent polygon buffers). The point entity buffer offers a variety of shapes, including circular, rectangular, and sector-shaped options. The sector-shaped buffer can be configured with an intention direction parameter.

[0024] S106: Based on the initial two-dimensional buffer polygon, query the digital elevation model to obtain terrain data, and generate a three-dimensional buffer surface through a slope-sensitive interpolation algorithm.

[0025] In this embodiment, the digital elevation model uses 10-meter resolution DEM data (terrain elevation data) and implements multi-scale terrain loading through a tile pyramid structure; This embodiment uses a slope-sensitive interpolation algorithm to expand a two-dimensional buffer zone into three dimensions based on terrain data. In areas with steeper slopes, the elevation changes on the buffer surface are more pronounced; in flatter areas, the elevation changes are more gradual, resulting in a three-dimensional buffer zone surface that reflects the actual terrain. This ensures that the buffer zone not only meets the required planar extent but also truly reflects the influence of terrain factors, providing more comprehensive and accurate spatial information support for geographic analysis, planning decisions, and other aspects. In another possible embodiment, this embodiment queries a digital elevation model to obtain terrain data based on the initial two-dimensional buffer polygon, and generates a three-dimensional buffer surface through a terrain adaptive algorithm.

[0026] In this embodiment, the terrain adaptation algorithm includes three core modules: slope threshold judgment (elevation correction is initiated in areas with a slope >25°), slope attenuation factor (increases the distance in the direction of the slope by 20% and decreases it in the direction of the slope by 30%), and elevation scaling factor (adjusted based on the density of contour lines). 3D surface generation is output in the form of irregular triangulation and / or grids, and triangulation nodes are encrypted in steep terrain areas (slope >45°) to ensure that the surface accuracy reaches an error range of 1 meter.

[0027] From the above, it can be concluded that this application extracts fuzzy descriptive words in natural language and parses membership values ​​through a semantic parsing engine, and obtains basic distance parameters through a pre-trained language-space mapping model, effectively filling the gap between fuzzy semantics and precise spatial parameters, and improving the accuracy and efficiency of human-computer interaction. Secondly, the feature length is calculated based on the geometric features of the geographic entity, and the scale correction coefficient is obtained through the scale correction function, so that the geographic distance threshold is fully adapted to the scale of the entity itself, reducing the result deviation caused by ignoring the entity difference caused by the traditional method. Based on the digital elevation model and the terrain adaptive algorithm, a three-dimensional buffer surface is generated, taking into account the terrain factors, so that the buffer is more in line with the real geographic scene, which significantly improves the application efficiency of GIS in urban planning, resource management and other fields.

[0028] In one embodiment of the present application, parsing the membership value corresponding to the fuzzy description word includes: Based on the fuzzy description word, a preset fuzzy semantic dictionary is queried to obtain the corresponding membership value, or a semantic understanding model based on deep learning is used to calculate the membership value corresponding to the fuzzy description word according to the context of the fuzzy description word in the natural language instruction; The fuzzy semantic dictionary stores a plurality of fuzzy descriptive words and their corresponding membership functions or membership value intervals.

[0029] In this embodiment, the fuzzy semantic dictionary is constructed by collating and analyzing a large amount of natural language data related to geographic space. It stores multiple common fuzzy descriptors and the membership function or membership value range corresponding to each fuzzy descriptor. For example, for the fuzzy descriptor "very close," the dictionary stores its corresponding membership function, which can characterize the mapping relationship between distance and membership. When the distance is 0-100 meters, the membership gradually decreases from 1.0 to 0.5. A membership value range, such as 0.8-1.0, can also be directly given. After extracting the fuzzy descriptor in the user instruction, the corresponding membership information can be quickly obtained by searching the fuzzy semantic dictionary, and the specific membership value can be determined.

[0030] If a fuzzy descriptor is not included in the preset fuzzy semantic dictionary, or its semantics are greatly affected by the context, a semantic understanding model based on deep learning can be used for calculation. This semantic understanding model is trained on a large amount of natural language text and can deeply understand the meaning of words in different contexts. After the natural language instructions containing fuzzy descriptors are input into the semantic understanding model, the semantic understanding model will analyze the degree of fuzziness expressed by the fuzzy descriptor in combination with the context, and then calculate the corresponding membership value. For example, in the instruction "walk down from the top of the mountain to a place not far away", the meaning of "not far" will be affected by the geographical entity "top of the mountain" and the action of "walking down". The deep learning model can combine this contextual information to more accurately calculate the membership value corresponding to "not far". Specifically, when using a deep learning-based semantic understanding model, the model uses a BERT pre-trained model. The input layer contains a window of text consisting of the fuzzy descriptor itself and three contextual words before and after it. Contextual features are extracted using a 12-layer Transformer encoder, and the output layer uses a fully connected network to map membership values ​​in the range [0, 1]. The model training dataset includes a large number of annotated samples, each of which is annotated by geographic information experts with a reference membership value for the fuzzy descriptor in a specific context. The training process uses the Mean Separation (MSE) loss function, and the average error on the validation set is kept within ±0.05.

[0031] In practical applications, this embodiment can adopt a hybrid parsing strategy of dictionary priority + model completion: when the matching degree of the fuzzy descriptive word in the dictionary is ≥90%, the membership function in the dictionary is directly called to calculate the value; when the matching degree is <90% or it is an unregistered word, the semantic understanding model is automatically triggered to perform context calculation. For example, when processing complex descriptions such as a park not far from a school, the model can correct the membership value not far away based on the typical scale of the school. In summary, this embodiment provides two ways to parse the membership values ​​of fuzzy descriptive words. Querying the fuzzy semantic dictionary can quickly obtain the membership value by using the existing knowledge reserves, and the semantic understanding model based on deep learning can more accurately calculate the membership value according to the context. The combination of the two improves the accuracy and adaptability of the membership value parsing, and provides more reliable basic data for subsequent buffer generation.

[0032] In one embodiment of the present application, a preset fuzzy semantic dictionary is queried based on the fuzzy description word to obtain the corresponding membership value, including: A preset fuzzy semantic dictionary is queried. If the fuzzy description word exists in the dictionary, a membership value corresponding to the fuzzy description word is calculated using a semantic understanding model based on deep learning as a first membership value. Based on the fuzzy description word, a preset fuzzy semantic dictionary is queried to obtain a corresponding membership value as a second membership value; determining a final membership value based on the first membership value, the second membership value, and a difference between the first membership value and the second membership value; If the fuzzy descriptor is not included in the dictionary, the membership value corresponding to the fuzzy descriptor is calculated using a semantic understanding model based on deep learning as the final membership value.

[0033] Specifically, in this embodiment, when a fuzzy descriptor exists in a preset fuzzy semantic dictionary, the calculation process for the first membership value is consistent with the aforementioned deep learning-based semantic understanding model, i.e., derived from the context. The second membership value is obtained from the corresponding membership function or membership value interval in the fuzzy semantic dictionary. If it is a membership function, the relevant parameters of the geographic entity are substituted into the calculation to obtain the specific value; if it is a membership value interval, the median of the interval is used as the second membership value.

[0034] Specifically, when querying the preset fuzzy semantic dictionary, first determine whether the fuzzy descriptive word exists in the dictionary. If the fuzzy descriptive word exists in the dictionary, two calculation methods will be started at the same time. On the one hand, the fuzzy descriptive word is processed using a semantic understanding model based on deep learning, and its contextual context in the natural language instruction is combined to calculate the corresponding membership value, which is used as the first membership value. This process can fully consider the impact of the context on fuzzy semantics, so that the result is more in line with the user's true intention. On the other hand, the preset fuzzy semantic dictionary is directly queried based on the fuzzy descriptive word, and the corresponding membership value is obtained from the membership function or membership value interval stored in the dictionary as the second membership value. It relies on the standardized data already in the dictionary and has a certain stability. The final membership value is then determined based on the first and second membership values, as well as the difference between them. For example, if the difference is small, it indicates a high degree of consistency between the two methods, and the average of the two can be used as the final value. If the difference is large, further analysis is needed to determine the cause. Weights should be assigned based on factors such as the usage scenario of the fuzzy descriptors and the type of geographic entity, assigning a higher weight to the one that is more consistent with the actual situation. The final membership value is then determined through a weighted calculation, ensuring both flexibility and stability. If a fuzzy descriptor isn't included in the pre-defined fuzzy semantic dictionary, lacking the reference data provided by the dictionary, the deep learning-based semantic understanding model directly calculates the membership value corresponding to the fuzzy descriptor and uses this as the final membership value. In this case, the model, leveraging its ability to understand complex contexts, can more flexibly handle unincluded fuzzy descriptors and ensure the effective acquisition of membership values.

[0035] In summary, this embodiment uses both deep learning model-calculated values ​​and dictionary query values ​​for fuzzy descriptors in the dictionary, and determines the final value based on the relationship between the two. This approach combines the advantages of both approaches. When the difference between the two is small, a weighted average is taken to integrate the information from both, improving accuracy. When the difference is large, the final value is further determined based on the complexity of the natural language instructions, making the result more consistent with the actual situation and enhancing the robustness of the method.

[0036] In one embodiment of the present application, determining a final membership value based on the first membership value, the second membership value, and the difference between the first membership value and the second membership value includes: If the difference between the first membership value and the second membership value is less than a preset difference threshold, calculating a weighted average of the first membership value and the second membership value as a final membership value; If the difference between the first membership value and the second membership value is greater than or equal to a preset difference threshold, complexity analysis is performed on the natural language instruction to obtain a complexity value; and a final membership value is determined based on the complexity value.

[0037] In this embodiment, when determining the final membership value, the difference between the first membership value and the second membership value is first calculated. When the difference is less than or equal to the first preset difference threshold, it means that the difference between the calculation results of the two methods is small, and the final membership value is taken as the arithmetic mean of the two. When the difference is greater than the first preset difference threshold and less than the second preset difference threshold, the difference is medium, and a weighted average method is adopted, where the weight of the first membership value belongs to the first range value, and the weight of the second membership value belongs to the second range value; wherein the value in the second range value is less than the value in the first range value; the second preset difference threshold is the preset difference threshold; the third preset difference threshold is greater than the second preset difference threshold, and the second preset difference threshold is greater than the first preset difference threshold.

[0038] When the difference is greater than or equal to the second preset difference threshold and less than the third threshold, the difference is large. At this time, complexity analysis is performed on the natural language instruction to obtain a complexity value; and the final membership value is determined based on the complexity value.

[0039] When the difference is greater than or equal to the third preset difference threshold, the difference is very large, and the manual verification mechanism is activated. Experts determine the final membership value based on the specific context and geographic entity characteristics, and feed the result back to the system. At the same time, the training sample library of the fuzzy semantic dictionary and semantic understanding model is updated to improve the accuracy of subsequent calculations.

[0040] In this embodiment, when the difference is greater than or equal to the second preset difference threshold and less than the third threshold, the natural language instruction complexity analysis is carried out from three dimensions: one is syntactic complexity, which calculates the average sentence length and the number of nested clauses through dependency syntactic analysis; the second is semantic complexity, which counts the number of geographical entities and the number of fuzzy descriptive words; the third is contextual relevance, which calculates the association strength between fuzzy descriptive words and the entity words in the context through word vector cosine similarity. For example, if the difference between the first membership value and the second membership value is less than the preset difference threshold, it means that the difference between the results calculated by the two methods is small and has a high consistency. At this time, the weighted average of the first membership value and the second membership value is calculated as the final membership value. The weights here can be set according to the actual application scenario and the trust in the two calculation methods. For example, if the deep learning model's understanding of the context is more inclined, the first membership value can be given a higher weight of 0.6, and the second membership value can be given a weight of 0.4. The final result is obtained by weighted calculation, which not only combines the advantages of the two methods, but also reflects the emphasis on a specific method. If the difference between the first and second membership values ​​is greater than or equal to a preset difference threshold, it indicates a significant discrepancy between the two calculated results and requires further processing. At this point, a complexity analysis is performed on the natural language instruction to determine a complexity value. Complexity analysis can be performed based on factors such as the number of geographic entities contained in the instruction, the number and degree of nesting of fuzzy descriptors, and the complexity of the sentence's grammatical structure. For example, a long sentence containing multiple geographic entities and nested fuzzy descriptors will have a higher complexity value, while a short sentence involving only a single geographic entity and simple fuzzy descriptors will have a lower complexity value. The final membership value is determined based on the resulting complexity value. If the complexity value is low, indicating relatively clear instruction semantics, the first membership value calculated by the deep learning model can be preferred, as it better integrates the simple context. If the complexity value is high, indicating complex instruction semantics, the standardized data in the dictionary is more relevant. In this case, the second membership value can be used, or the final value can be determined by combining both values ​​with additional auxiliary information.

[0041] In summary, this embodiment employs different processing methods based on the difference between the first and second membership values. When the difference is small, the two values ​​are rationally combined through weighted averaging. When the difference is large, the complexity of the instruction is analyzed, allowing for more flexible response to different situations. For complex instructions, the first membership value calculated by the deep learning model is more relied upon because it has a deeper understanding of the context and can improve the accuracy of the membership value.

[0042] In one embodiment of the present application, determining a final membership value according to the complexity value includes: If the complexity value is greater than or equal to the complexity threshold, the first membership value is used as the final membership value; If the complexity value is less than the complexity threshold, the first membership value is used as the final membership value.

[0043] In this embodiment, the complexity threshold is determined based on a large number of natural language instruction processing cases, aiming to unify decision-making criteria and improve processing efficiency. This setting is based on the fact that the first membership value is calculated through a deep learning model combined with the specific context, which better reflects the user's true intention in a specific instruction. Whether it is a simple low-complexity instruction or a high-complexity instruction with multiple semantics, the first membership value can more accurately reflect the degree of fuzzy descriptors in the current situation by relying on the model's deep understanding of the context.

[0044] Specifically, the complexity value is calculated on a percentage basis, with the three dimensions accounting for 40%, 40%, and 20% respectively. The calculation method is: complexity value = 0.4 × syntactic score + 0.4 × semantic score + 0.2 × relevance score, where each score is mapped to the range of 0-100 through normalization.

[0045] If the complexity value is greater than or equal to the complexity threshold, the first membership value is used as the final membership value. This is because when the complexity of a natural language instruction reaches or exceeds the set threshold, it means that the information contained in the instruction is relatively complex, with multiple interrelated geographic entities and the nested use of fuzzy descriptors. The semantic understanding model based on deep learning can better capture the true meaning of fuzzy descriptors in such complex contexts. Through in-depth analysis of the context, it can obtain a first membership value that is more consistent with the user's intention, so this is selected as the final result. If the complexity value is less than the complexity threshold, the first membership value is also used as the final membership value. Even if the natural language instruction is less complex and the semantics are relatively simple and clear, the deep learning-based semantic understanding model can still accurately understand the meaning of fuzzy descriptors in a simple context. The first membership value it calculates can accurately reflect the user's vague description of the spatial range. In this case, using the first membership value as the final result can ensure the consistency and accuracy of the results when processing instructions of different complexities.

[0046] For example, when the complexity value is less than 30, the first membership value is used as the final result because the model calculation is more reliable in a simple context; when 30≤complexity value<70, the secondary context analysis is started, and the geographic entity encyclopedia knowledge is added to the original model for recalculation, and the average of the recalculated value and the second membership value is taken; when the complexity value is ≥70, the human-computer interaction mechanism is triggered, and the system shows the user a preview of the buffer corresponding to the two membership values. The user selects the result that is more in line with the expectations, and the user's selection preference is recorded for model optimization.

[0047] In summary, for instructions with high complexity, this embodiment uses the first membership value calculated by the deep learning model as the basis, which can better handle complex semantic situations; for instructions with low complexity, the first membership value is also used, which ensures the consistency and simplicity of the method and helps to improve the rationality and reliability of the determination of the membership value.

[0048] In one embodiment of the present application, calculating the characteristic length corresponding to the geometric feature includes: According to the geographic entity type and the intention direction expressed by the fuzzy descriptor, the corresponding calculation strategy is selected from the preset feature length calculation rule library; Among them, the preset feature length calculation rule library defines the feature length calculation method corresponding to different entity types or intention directions; The characteristic length corresponding to the geometric feature is calculated based on the calculation strategy.

[0049] In this embodiment, a preset feature length calculation rule library is constructed based on a large number of geographic entity features and actual application scenarios, which clearly defines the feature length calculation methods corresponding to different entity types or intended directions. Specifically, a two-dimensional index structure is adopted. The horizontal dimension is divided into three categories according to geographic entity type: point entities, line entities, and surface entities, and each category is further divided into multiple subcategories. The vertical dimension is divided into three categories according to intended direction: omnidirectional, unidirectional, and multidirectional. The feature length calculation rule library includes multiple calculation rules, each of which includes four elements: applicable entity type, direction range, calculation algorithm, and accuracy parameter. For example, when the geographic entity type is a linear river and the intended direction expressed by the fuzzy descriptor is along the upstream and downstream sides of the river, the corresponding calculation strategy is to calculate the total length of the river. If the intended direction is on both sides of a specific river section, the calculation strategy is to calculate the length of that specific section. For a planar urban park, if the intended direction is the entire perimeter of the park, the calculation strategy is to calculate the park's perimeter. If the intended direction is near the park entrance, the average distance from the entrance point to the park boundary is used as the basis for calculating the characteristic length. Then, based on the selected calculation strategy, the characteristic length corresponding to the geometric feature is calculated. After obtaining the geometric feature data of the geographic entity, the calculation is performed according to the steps and methods specified in the calculation strategy. For example, when using the strategy to calculate the total length of the river, the coordinates of the linear geometric feature of the river are extracted from the spatial database, and the total length of the river is calculated by accumulating the distances between the coordinate points. This length is the corresponding characteristic length. When using the strategy to calculate the perimeter of the park, the boundary coordinates of the planar geometric feature are used to sequentially calculate the distances between each adjacent boundary point and accumulate them to obtain the park's perimeter as the characteristic length. In this way, the calculated feature length can be ensured to match the geographic entity type and user intention direction, providing accurate data support for subsequent geographic distance threshold correction.

[0050] In summary, this embodiment selects a feature length calculation strategy based on the geographic entity type and the intended direction of the fuzzy descriptor, ensuring that feature length calculation is more consistent with the actual geographic entity and user needs. Different calculation methods correspond to different geographic entity types and different intended directions, enabling more accurate reflection of the spatial characteristics of the geographic entity, providing more precise parameters for subsequent buffer generation, and improving buffer generation accuracy.

[0051] In one embodiment of the present application, a corresponding calculation strategy is selected from a preset feature length calculation rule library based on the geographic entity type and the intended direction expressed by the fuzzy descriptor, including: According to the geographic entity type, a candidate calculation strategy set applicable to the geographic entity type is screened out from a preset characteristic length calculation rule library; Analyze the intention direction of the fuzzy description words and generate the corresponding intention direction semantic vector; Calculate the cosine similarity between the intention direction semantic vector and all standard strategy semantic vectors in the preset feature length calculation rule library; the standard strategy semantic vector is generated based on the standard intention description label associated with each calculation strategy defined in the preset feature length calculation rule library; The calculation strategy corresponding to the standard strategy semantic vector with the highest cosine similarity with the user intention direction semantic vector in the candidate calculation strategy set is selected as the final calculation strategy.

[0052] In this embodiment, the preset characteristic length calculation rule library contains corresponding calculation strategies for different geographic entity types. For example, when the geographic entity type is a linear road, the calculation strategies applicable to the road in the rule library include calculating the total length of the road, the length of a specific road section, and the extended length corresponding to the widest point of the road. These strategies collectively constitute the candidate calculation strategy set.

[0053] This embodiment uses natural language processing technology to deeply analyze the directional information expressed by fuzzy descriptors, extracting semantic features. These features are then converted into a computer-readable vector form, namely the intended direction semantic vector. For example, for the intended direction of "both sides of a highway," the semantic vector generated after analysis would include relevant semantic information such as "highway," "both sides," and "around the highway and toward both sides."

[0054] This embodiment calculates the cosine similarity between the intention direction semantic vector and all standard strategy semantic vectors in the preset feature length calculation rule library. Among them, the standard strategy semantic vector is generated based on the standard intention description label associated with each calculation strategy defined in the preset feature length calculation rule library. Each calculation strategy has a corresponding standard intention description label, such as calculating the total length and calculating the length of a specific road section, etc., and a standard strategy semantic vector is generated based on these labels. By calculating the cosine similarity, the degree of similarity between the intention direction semantic vector and each standard strategy semantic vector can be quantified. The higher the similarity, the more consistent the intention directions expressed by the two are. This embodiment selects the calculation strategy corresponding to the standard strategy semantic vector with the highest cosine similarity to the user's intended direction semantic vector in the candidate calculation strategy set as the final calculation strategy. Since the candidate calculation strategy set already limits its applicability to the current geographic entity type, selecting the calculation strategy with the highest similarity ensures that the strategy meets the requirements of the geographic entity type and maximizes the match to the intended direction expressed by the user's fuzzy descriptor, thereby providing a precise and tailored method for calculating feature length.

[0055] Specifically, when filtering the candidate computing strategy set based on the geographic entity type, the geographic entity type is first hierarchically matched. For example, when the entity type is a highway, the rule library will automatically filter out all computing strategies marked as line entity-traffic road class, forming a candidate computing strategy set including 5-8 strategies. The screening process is implemented through hash mapping of type labels. BERT word vector encoding is used to generate semantic vectors for parsing intent directions. The directional description text is converted into a 768-dimensional dense vector, where directional, distance, and range words are reinforced using attention weights. For descriptions without explicit directions, a default omnidirectional semantic vector is generated. The generation of standard strategy semantic vectors is based on the 3-5 standard intent description labels bound to each strategy in the rule base. For example, the strategy-associated labels for "line entity-one-way projection length" include "side", "unilateral", and "extending in a certain direction". The standard vector is obtained by calculating the average word vector of these labels. Cosine similarity is calculated using vector normalization, using the formula: cosθ = (A・B) / (|A|・|B|), where A is the semantic vector of the intent direction and B is the semantic vector of the standard strategy. After the calculation is complete, the strategy with the highest similarity is selected from the candidate strategy set as the final solution. If the highest similarity falls below a preset similarity threshold, the strategy fusion mechanism is triggered, resulting in a weighted combination of the top two most similar strategies, with weights allocated based on the similarity ratio.

[0056] In summary, this embodiment selects a calculation strategy by calculating the cosine similarity between the semantic vector of the intent direction and the semantic vector of the standard strategy. This quantifies the degree of match between the user's intent and the strategies in the rule base, thereby selecting the calculation strategy that best matches the user's intent. This method, based on semantic vector similarity, is highly scientific and accurate, and can improve the rationality and intelligence of feature length calculation strategy selection.

[0057] In one embodiment of the present application, the GIS buffer generation method based on fuzzy semantics and geographic constraints further includes: if the highest similarity is lower than a preset similarity threshold, Parse the intended direction into preset basic direction categories; Select the corresponding calculation strategy according to the entity type and basic direction category; If the basic direction category cannot be resolved, the default strategy is selected according to the entity type The default strategies include: for linear features, select the average width strategy; for area features, select the equivalent radius strategy; for point features, select the fixed radius strategy.

[0058] In this embodiment, the intended direction is parsed into a preset basic direction category. The preset basic direction category is set according to common spatial direction requirements, such as omnidirectional, unilateral, bilateral, specific area, etc. When the highest similarity between the intention direction semantic vector and the standard strategy semantic vector does not reach the preset threshold, it means that the user's intention direction expression is relatively vague or special, and it is difficult to directly match the standard strategy in the rule base. At this time, through further semantic parsing, the user's intention direction is classified into the preset basic direction category to simplify the subsequent strategy selection process. For example, if the user's fuzzy description word is not far from the surrounding area of ​​the highway, it will be classified into the omnidirectional basic direction category after parsing. The preset characteristic length calculation rule library of this embodiment also stores calculation strategies corresponding to the combination of entity type and basic direction category. For example, for the combination of linear features and bilateral basic direction categories, the corresponding calculation strategy is to calculate the average width within 50 meters on both sides of the road as the characteristic length; for the combination of surface features and basic direction categories near the entrance, the corresponding calculation strategy is to calculate the average length of the park boundary within a radius of 200 meters with the park entrance as the center as the characteristic length. Through the dual limitation of entity type and basic direction category, this embodiment can filter out a calculation strategy that better suits current needs from the preset characteristic length calculation rule library. If the basic direction category cannot be resolved, that is, after semantic analysis, the user's intended direction still cannot be classified into any preset basic direction category, the default strategy is selected according to the entity type. The default strategy is a pre-set general calculation strategy for different entity types to ensure that the characteristic length calculation can be completed even in extreme cases. Specifically, for linear features, the average width strategy is selected, that is, the average width of the entire linear feature is calculated as the characteristic length, such as calculating the average width of a river; for surface features, the equivalent radius strategy is selected, that is, the radius of an equivalent circle is calculated based on the area of ​​the surface feature as the characteristic length, such as converting the area of ​​an irregular lake into the radius of a circle of equal area; for point features, the fixed radius strategy is selected, that is, a fixed radius value is preset as the characteristic length, such as fixing the characteristic length of a monument to 50 meters. Through these default strategies, the characteristic length calculation work can be carried out smoothly in various complex situations, providing basic data for subsequent geographic distance threshold correction.

[0059] In summary, this embodiment provides a further approach when the similarity falls below a preset threshold. This approach resolves the intended direction into a basic direction category and selects a strategy based on the entity type. If resolution is unsuccessful, a default strategy is used. This allows for the selection of a relatively reasonable calculation strategy based on the general characteristics of the geographic entity type when an accurate matching strategy is unavailable, ensuring the method's proper operation in a variety of situations and enhancing its versatility and adaptability.

[0060] In one embodiment of the present application, the geographic entity types include: linear features, surface features, and point features; The mapping rules of the characteristic length calculation strategy include: When the geographic entity type is a linear feature, if the intended direction includes a description of the along-line type, the length segmentation strategy is adopted, and the percentage of the total length of the entity is used as the characteristic length; if the intended direction includes a description of the horizontal type, the average width strategy is adopted, and the length of the minor axis of the entity's geometric enclosing rectangle is calculated as the characteristic length; When the geographic entity type is a surface feature, if the intention direction includes a boundary class description, the perimeter strategy is used to calculate the total length of the geometric boundary as the characteristic length; if the intention direction includes an internal class description, the equivalent radius strategy is used to calculate the radius of the area equivalent circle as the characteristic length; When the geographic entity type is a point feature, a fixed radius strategy is adopted, and the characteristic length is the preset typical influence radius.

[0061] In one embodiment of the present application, the basic distance parameter is modified based on the intention direction, membership value, geographic entity type, and feature length expressed by the fuzzy descriptor to obtain the geographic distance threshold, including: Based on the intention direction, membership value, geographic entity type and feature length expressed by the fuzzy descriptor, the scale correction coefficient is calculated through the preset scale correction function; The basic distance parameter and the scale correction coefficient are weighted to obtain the geographic distance threshold.

[0062] In this embodiment, a scale correction coefficient is first calculated using a preset scale correction function based on the intended direction, membership value, geographic entity type, and feature length expressed by the fuzzy descriptor. This preset scale correction function is a mathematical model constructed by integrating multiple factors. Its input is the four key parameters mentioned above, and its output is a scale correction coefficient that reflects the combined influence of these parameters.

[0063] From the perspective of the intended direction, different directions will affect the correction factor. For example, if the intended direction is one side of a linear feature, the correction factor will be lower than if it is all-around, because the one-side range is usually smaller than the all-around range. If the intended direction is towards the key area of ​​the feature, the correction factor will be smaller to narrow the distance threshold range. The membership value is also closely related to the correction factor. A higher membership value indicates greater certainty about the fuzzy range, and the correction factor is closer to 1, reducing adjustments to the base distance parameter. Conversely, a lower membership value causes the correction factor to expand or contract the base distance parameter to accommodate a more ambiguous range.

[0064] The type of geographic entity is a significant factor influencing the correction factor. Different geographic entities have different spatial impacts and characteristics. For example, for densely populated residential areas, the correction factor for the surrounding distance threshold will be relatively small to avoid errors caused by excessive coverage; while for open nature reserves, the correction factor will be larger to include a wider range of surrounding areas. The characteristic length provides a quantitative basis for the correction coefficient. Geographic entities with larger characteristic lengths will have larger correction coefficients, because larger entities usually require a wider surrounding range; conversely, entities with smaller characteristic lengths will have smaller correction coefficients. After calculating the scale correction coefficient using the scale correction function, which integrates the above factors, the base distance parameter and the scale correction coefficient are weighted to obtain the geographic distance threshold. The weights in this weighted calculation can be set based on the importance of each parameter in the actual application scenario, ensuring that the resulting geographic distance threshold accurately reflects the user's fuzzy semantic needs while also meeting the spatial characteristics and constraints of geographic entities.

[0065] This embodiment calculates a scale correction coefficient using a preset scale correction function, then weights it with the base distance parameter to determine the geographic distance threshold. This approach can correct the distance parameter based on multiple factors, including the intended direction of the fuzzy descriptor, the degree of membership, the geographic entity type, and the feature length. This approach makes the determination of the geographic distance threshold more scientific and reasonable, more accurately reflecting the user's buffer range requirements, and thus generating a GIS buffer that better meets practical needs.

[0066] In one embodiment of the present application, a semantic parsing engine is used to extract geographic entity keywords and fuzzy descriptive words from a user's natural language instruction, and parse the membership values ​​corresponding to the fuzzy descriptive words, including: The predefined dictionary includes geographic entity type dictionary, fuzzy description word dictionary and associated grammar rules. The fuzzy description word dictionary stores the mapping relationship between vocabulary and basic membership intervals. Identify the core predicate structure in the instruction through dependency parsing, combined with part-of-speech tagging and named entity recognition: Extract geographic entity keywords from noun phrases; extract vague descriptive words from adverbial / adjective phrases; The membership of fuzzy descriptors is scalarized. If it is a single descriptor, the median of the basic membership interval is taken as the scalar membership value. If it is a combination descriptor, the scalar membership value is calculated using the membership aggregation function. Among them, the progressive combination uses multiplication synthesis, and the parallel combination uses maximum synthesis. The metadata table of the spatial database is queried based on the geographic entity keywords to determine the geographic entity type.

[0067] In one embodiment of the present application, when a geographic entity has a time attribute or is a dynamic entity: Characteristic length is calculated based on the geometric shape of the geographic entity at a specific point in time or within a specific time period; Alternatively, the characteristic length is calculated based on the expected maximum extent of the geographic entity or its geometry in a typical state; Alternatively, the characteristic length is calculated based on the length or spatial distribution range of the geographic entity's movement trajectory.

[0068] In this embodiment, the characteristic length is calculated based on the geometric shape of the geographic entity at a specific time point or a specific time period. For geographic entities with clear time attributes, their geometric shape changes over time, so it is necessary to lock a specific time range to determine the characteristic length. For example, a tidal flat under the influence of tides is a geographic entity with a time attribute, and its area and boundaries will be significantly different at high tide and low tide. If the user is interested in the buffer zone around a beach at high tide, it is necessary to extract the geometric shape of the beach at this specific time point of high tide and calculate the characteristic length based on this, such as the maximum width of the tidal flat or the length of the coastline at high tide; for seasonal rivers, if the surrounding area during the rainy season is analyzed, it is necessary to use the average width, length and other geometric features of the river during the specific time period of the rainy season as the basis for calculating the characteristic length. In this embodiment, the characteristic length is calculated based on the expected maximum range of the geographic entity or the geometric form under the typical state. For dynamic entities, if there is regularity or predictability in the range of its variation, the expected maximum range or the geometric form under the typical state can be used to determine the characteristic length to meet the universal requirements of buffer zone generation. For example, migrating bird populations are dynamic entities, and their activity range will change along the migration route. If a buffer zone is to be generated around the activity area of ​​the bird population, the characteristic length can be calculated based on the expected maximum activity range predicted by its historical migration data, such as the diameter of the maximum activity area; for temporary gathering areas in cities, the geometric form under the typical state (such as the area occupied by regular gatherings and the distribution range) can be used as the basis for calculating the characteristic length, such as the perimeter of the gathering area under the typical state. In this embodiment, the characteristic length is calculated based on the length of the geographic entity's trajectory or its spatial distribution range. For dynamic entities with continuous motion, the spatial distribution characteristics of their motion are key to calculating the characteristic length. For example, a moving vehicle fleet is a dynamic entity, and its characteristic length can be determined based on the length of its trajectory. For a spreading pollutant cloud, its spatial distribution range (such as the maximum radius of the diffusion area or the equivalent diameter corresponding to the covered area) can be used as the characteristic length. If the cloud forms an elliptical distribution during its diffusion process, the length of its major axis can be calculated as the characteristic length to reflect the scale of its spatial impact range. This embodiment, through the above-mentioned characteristic length calculation method for temporal or dynamic entities, ensures that the characteristic length value remains consistent with the actual state or changing patterns of these special geographic entities when processing them. This provides accurate basic data for subsequent correction of geographic distance thresholds, thereby generating a GIS buffer that meets practical needs.

[0069] Corresponding to the GIS buffer generation method based on fuzzy semantics and geographic constraints in the above embodiment, Figure 2This is a structural block diagram of a GIS buffer generation system based on fuzzy semantics and geographic constraints provided in one embodiment of the present application. For ease of explanation, only the parts related to the embodiment of the present application are shown. Figure 2 The GIS buffer generation system 20 based on fuzzy semantics and geographic constraints includes: a semantic parsing module 21, a data mapping module 22, a data calculation module 23, a data correction module 24, a first generation module 25 and a second generation module 26.

[0070] The semantic parsing module 21 is used to extract geographic entity keywords, geographic entity types and fuzzy descriptive words from the user's natural language instructions through a semantic parsing engine, and parse the fuzzy descriptive words to obtain the membership value of the corresponding fuzzy descriptive words; The data mapping module 22 is used to input the fuzzy descriptors, membership values ​​and geographic entity types into a pre-trained language-space mapping model to obtain a basic distance parameter, which is a reference value with a distance dimension or a scaling factor relative to a typical scale of the entity; The data calculation module 23 is used to query the spatial database based on the geographic entity keywords, obtain the geometric features associated with the geographic entity, and calculate the feature length corresponding to the geometric features; A data correction module 24 is used to correct the basic distance parameter based on the intention direction, membership value, geographic entity type and feature length expressed by the fuzzy descriptor to obtain a geographic distance threshold; A first generating module 25 is used to generate an initial two-dimensional buffer polygon based on a geographic distance threshold and spatial geometric features of a geographic entity; The second generating module 26 is configured to query the digital elevation model to obtain terrain data based on the initial two-dimensional buffer polygon, and generate a three-dimensional buffer surface using a slope-sensitive interpolation algorithm.

[0071] See also Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided in one embodiment of the present application. Figure 3 The electronic device 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of the modules in the above-mentioned device embodiments, such as Figure 2The functions of the semantic parsing module 21, the data mapping module 22, the data calculation module 23, the data correction module 24, the first generation module 25 and the second generation module 26 are shown.

[0072] It should be understood that in the embodiment of the present application, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0073] The input device 302 may include a touchpad, a fingerprint collection sensor (for collecting user fingerprint information and fingerprint direction information), a microphone, etc. The output device 303 may include a display (LCD, etc.), a speaker, etc.

[0074] The memory 304 may include a read-only memory and a random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store device type information.

[0075] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiments of the present application can execute the implementation method described in any embodiment of the GIS buffer generation method based on fuzzy semantics and geographic constraints provided in the embodiments of the present application, and can also execute the implementation method of the electronic device described in the embodiments of the present application, which will not be repeated here.

[0076] In another embodiment of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program. The computer program includes program instructions. When the program instructions are executed by a processor, all or part of the process of the method in the above embodiment is implemented. The computer program can also be used to instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of each of the above method embodiments are implemented. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium.

[0077] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the aforementioned embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. Furthermore, the computer-readable storage medium can include both an internal storage unit of the electronic device and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or is about to be output.

[0078] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0079] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the electronic devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0080] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or can be an electrical, mechanical or other form of connection.

[0081] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0082] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0083] The above are only specific embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A GIS buffer generation method based on fuzzy semantics and geographic constraints, characterized by: include: Extracting geographic entity keywords, geographic entity types, and fuzzy descriptive words from the user's natural language instructions through a semantic parsing engine, and parsing the fuzzy descriptive words to obtain membership values ​​corresponding to the fuzzy descriptive words; Inputting the fuzzy descriptor, the membership value and the geographic entity type into a pre-trained language-space mapping model to obtain a basic distance parameter; Querying a spatial database based on the geographic entity keyword to obtain geometric features associated with the geographic entity, and calculating feature lengths corresponding to the geometric features; The basic distance parameter is modified based on the intention direction, membership value, geographic entity type and feature length expressed by the fuzzy descriptor to obtain a geographic distance threshold; generating an initial two-dimensional buffer polygon based on the geographic distance threshold and the spatial geometric characteristics of the geographic entity; Based on the initial two-dimensional buffer polygon, a digital elevation model is queried to obtain terrain data, and a three-dimensional buffer surface is generated through a slope-sensitive interpolation algorithm.

2. A GIS buffer generation method based on fuzzy semantics and geographic constraints according to claim 1, characterized in that: The step of parsing the fuzzy description word to obtain a membership value corresponding to the fuzzy description word includes: Based on the fuzzy description word, a preset fuzzy semantic dictionary is queried to obtain a corresponding membership value, or a semantic understanding model based on deep learning is used to calculate the membership value corresponding to the fuzzy description word according to the context of the fuzzy description word in the natural language instruction; The fuzzy semantic dictionary stores a plurality of fuzzy description words and their corresponding membership functions or membership value intervals.

3. A GIS buffer generation method based on fuzzy semantics and geographic constraints according to claim 2, characterized in that: The querying of a preset fuzzy semantic dictionary based on the fuzzy description word to obtain a corresponding membership value includes: Querying the preset fuzzy semantic dictionary, and if the fuzzy description word exists in the dictionary, calculating a membership value corresponding to the fuzzy description word as a first membership value using a semantic understanding model based on deep learning; Based on the fuzzy description word, a preset fuzzy semantic dictionary is queried to obtain a corresponding membership value as a second membership value; determining a final membership value based on the first membership value, the second membership value, and a difference between the first membership value and the second membership value; If the fuzzy description word is not included in the dictionary, the membership value corresponding to the fuzzy description word is calculated using a semantic understanding model based on deep learning as the final membership value.

4. A GIS buffer generation method based on fuzzy semantics and geographic constraints according to claim 3, characterized in that: The determining a final membership value based on the first membership value, the second membership value, and a difference between the first membership value and the second membership value includes: If the difference between the first membership value and the second membership value is less than a preset difference threshold, calculating a weighted average of the first membership value and the second membership value as a final membership value; If the difference between the first membership value and the second membership value is greater than or equal to a preset difference threshold, complexity analysis is performed on the natural language instruction to obtain a complexity value; and a final membership value is determined based on the complexity value.

5. A GIS buffer generation method based on fuzzy semantics and geographic constraints according to claim 4, characterized in that: The determining of a final membership value according to the complexity value includes: If the complexity value is greater than or equal to the complexity threshold, taking the first membership value as the final membership value; If the complexity value is less than the complexity threshold, the first membership value is used as the final membership value.

6. A GIS buffer generation method based on fuzzy semantics and geographic constraints according to claim 1, characterized in that: The calculating the characteristic length corresponding to the geometric feature includes: Selecting a corresponding calculation strategy from a preset feature length calculation rule library according to the geographic entity type and the intention direction expressed by the fuzzy descriptor; The preset feature length calculation rule library defines the feature length calculation methods corresponding to different entity types or intention directions; The characteristic length corresponding to the geometric feature is calculated based on the calculation strategy.

7. A GIS buffer generation method based on fuzzy semantics and geographic constraints according to claim 6, characterized in that: The selecting a corresponding calculation strategy from a preset feature length calculation rule library according to the geographic entity type and the intention direction expressed by the fuzzy descriptor includes: According to the geographic entity type, a candidate calculation strategy set applicable to the geographic entity type is screened out from the preset characteristic length calculation rule library; Analyzing the intention direction of the fuzzy description word and generating a corresponding intention direction semantic vector; Calculating the cosine similarity between the intention direction semantic vector and all standard strategy semantic vectors in the preset feature length calculation rule library; the standard strategy semantic vector is generated based on the standard intention description label associated with each calculation strategy defined in the preset feature length calculation rule library; The calculation strategy corresponding to the standard strategy semantic vector having the highest cosine similarity with the user intention direction semantic vector in the candidate calculation strategy set is selected as the final calculation strategy.

8. A GIS buffer generation system based on fuzzy semantics and geographic constraints, characterized by: include: A semantic parsing module is used to extract geographic entity keywords, geographic entity types and fuzzy descriptive words from the user's natural language instructions through a semantic parsing engine, and parse the fuzzy descriptive words to obtain membership values ​​corresponding to the fuzzy descriptive words; a data mapping module, configured to input the fuzzy descriptor, the membership value, and the geographic entity type into a pre-trained language-space mapping model to obtain a basic distance parameter; A data calculation module, configured to query a spatial database based on the geographic entity keywords, obtain geometric features associated with the geographic entity, and calculate feature lengths corresponding to the geometric features; A data correction module is used to correct the basic distance parameter based on the intention direction, membership value, geographic entity type and feature length expressed by the fuzzy descriptor to obtain a geographic distance threshold; A first generating module is configured to generate an initial two-dimensional buffer polygon based on the geographic distance threshold and the spatial geometric characteristics of the geographic entity; The second generating module is used to query the digital elevation model to obtain terrain data based on the initial two-dimensional buffer polygon, and generate a three-dimensional buffer surface through a slope-sensitive interpolation algorithm.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images