Ecotourism path simulation route selection method and system based on natural resource spatial data

Abstract

The application discloses an ecological recreation path simulation route selection method based on natural resource data, and comprises the following steps: obtaining the name and spatial coordinate information of a recreation node in a target region, and counting the common occurrence frequency of the recreation node as a recreation activity connection frequency observation value; a multiple linear regression model is constructed, the recreation activity connection frequency observation value is taken as a dependent variable, and the spatial resource background suitability, ecological landscape visibility and traffic connection convenience spatial indexes of the recreation node pair are taken as independent variables, so that a recreation activity connection frequency prediction model is fitted; for a region to be predicted, the index values are calculated based on the spatial data of the region to be predicted and input into the model, and the recreation activity connection frequency observation value between the nodes is predicted; and the node pairs with the recreation activity connection frequency observation value higher than the average value are screened out, so as to generate a recreation path route selection scheme. The application integrates multi-source spatial data and user behavior data, and realizes scientific and generalizable recreation path planning support.

Description

Technical Field

[0001] This invention relates to the field of smart tourism and recreation planning technology, specifically to a method and system for simulating and selecting ecological recreational routes based on natural resource data. Background Technology

[0002] With the acceleration of urbanization and the increasing demand for leisure among residents, parks and scenic spots, as important recreational spaces, directly impact the quality of urban life and the utilization of tourism resources through their planning and connectivity efficiency. Traditional recreational path planning often relies on planners' experience and judgment, existing road network structures, or simple spatial distance analysis (such as the shortest path method). While these methods are highly practical, they often fail to fully incorporate tourists' actual behavioral preferences and complex environmental factors, leading to planned paths that may be out of touch with the public's actual recreational activity patterns and reducing the overall effectiveness and attractiveness of the recreational system.

[0003] Specifically, existing technologies have the following limitations: First, when analyzing the recreational connections between two attractions, there is a lack of objective, large-scale behavioral observation data to support the analysis, making it difficult to accurately quantify the actual strength of the connection. Second, route selection decisions often focus on a single factor (such as geographical distance or existing roads), failing to systematically integrate the diverse spatial factors that influence tourists' travel choices, such as the attractiveness of the resources themselves (resource background), the quality of the visual landscape along the route (ecological landscape visibility), and the ease of access (convenience of transportation connections). Finally, existing methods typically lack good regional migration and prediction capabilities, making them difficult to directly apply to undeveloped areas with weak data foundations or those in the planning stage. Summary of the Invention

[0004] To overcome the above-mentioned technical deficiencies, the present invention provides a method and system for simulating and selecting ecological recreational routes based on natural resource data.

[0005] To solve the above problems, the present invention is implemented according to the following technical solution:

[0006] In a first aspect, the present invention provides a method for simulating and selecting ecological recreational routes based on natural resource data, comprising the following steps:

[0007] S1: Determine the target area and obtain the names and spatial coordinates of all recreational nodes within the target area, wherein the recreational nodes are any one of urban parks, green squares, natural landscapes, cultural heritage sites, museums, and commercial and leisure areas;

[0008] S2: Based on social media data, count the frequency of each pair of recreational nodes appearing together in the same user's itinerary or published content, generate a data table of recreational node pairs and their corresponding frequency of co-occurrence, and use the frequency of co-occurrence as the recreational activity connection frequency observation value of the corresponding recreational node pair;

[0009] S3: Construct a multiple linear regression model, wherein the multiple linear regression model uses the observed frequency of recreational activity connections as the dependent variable and the combination of spatial data indicators representing the resource suitability, ecological landscape visibility and transportation connectivity of the corresponding recreational node pair as independent variables.

[0010] S4: Based on the spatial data of all recreational nodes in the target area, calculate the independent variable index values ​​for all recreational node pairs;

[0011] S5: Fit the observed values ​​of recreational activity connection frequency with the corresponding independent variable index values, solve the undetermined coefficients of the multiple linear regression model, and obtain the trained recreational activity connection frequency prediction model.

[0012] S6: For the area to be predicted, based on the spatial data of all recreational nodes in the area to be predicted, calculate the independent variable index values ​​of all recreational node pairs in the area to be predicted, and substitute them into the trained recreational activity connection frequency prediction model to predict the observed values ​​of recreational activity connection frequency between all recreational node pairs in the area to be predicted.

[0013] S7: Based on the observed frequency of recreational activity connections among all recreational node pairs, select all recreational node pairs whose observed frequency of recreational activity connections is higher than the average value, and generate a recreational path selection scheme for the recreational node pairs in the area to be predicted.

[0014] In conjunction with the first aspect, the present invention provides a first specific implementation of the first aspect, specifically, the step of determining the target area and obtaining the names and spatial coordinate information of all recreational nodes within the target area specifically includes:

[0015] S11: Obtain the administrative division vector boundary data of the target area;

[0016] S12: Collect a list of names of all recreational nodes in the target area from tourism information platforms and / or local data;

[0017] S13: Perform geographic parsing on the recreation nodes in the name list to obtain their geographic coordinates;

[0018] S14: Correct and spatially locate the geographic coordinates to generate recreation node vector data with spatial location information.

[0019] In conjunction with the first aspect, the present invention provides a second specific implementation of the first aspect. Specifically, based on social media data, the frequency of co-occurrence of each pair of recreational nodes in the same user's itinerary or published content is statistically analyzed to generate a data table of recreational node pairs and their corresponding co-occurrence frequencies. The co-occurrence frequency is used as the recreational activity connection frequency observation value for the corresponding recreational node pair. Specifically, this includes:

[0020] S21: Construct query conditions from at least one social media platform using the name of the target area and keywords related to recreational activities, collect post data published within a preset time period through a web crawler, and store the collection results as a structured text file;

[0021] S22: Clean the post data in the text file and filter out records whose geographical location belongs to the target area and whose text content contains one or more recreation node names as valid recreation records;

[0022] S23: Construct a co-occurrence relation matrix for recreational nodes. For each pair of recreational nodes, count the frequency of the two recreational node names appearing simultaneously in the same post or the same user's trip in all valid recreational records. Use this as the recreational activity connection frequency observation value for the recreational node pair and organize it into a data table with recreational node pairs as the index and recreational activity connection frequency observation values ​​as the data items.

[0023] In conjunction with the first aspect, the present invention provides a third specific implementation of the first aspect, specifically characterized in that the multiple linear regression model is constructed based on a gravity model, and its expression is: In the above formula, The frequency of recreational activity connections between recreational node i and recreational node j is represented by C, S, and L, which represent relevant indicators in terms of resource suitability, ecological landscape visibility, and transportation connectivity convenience, respectively. For elevation indicators, For slope index, As an aspect indicator, The vegetation coverage rate is an indicator. This is an indicator of the visibility of mountains. As an indicator of water body visibility; It represents the minimum actual path length between any pair of resting nodes.

[0024] In conjunction with the first aspect, the present invention provides a fourth specific implementation of the first aspect. Specifically, the calculation of the independent variable index values ​​for all recreational node pairs based on the spatial data of all recreational nodes in the target area specifically includes:

[0025] S41: Acquire and integrate spatial data of all recreational nodes in the target area, the spatial data including: vector boundary data of recreational nodes, digital elevation model data, land use / cover data, point of interest data and road network data;

[0026] S42: For each pair of recreational nodes, calculate the values ​​of three types of independent variable indicators based on the spatial data. The three types of independent variable indicators include the resource background suitability indicator, the ecological landscape visibility indicator, and the transportation connection convenience indicator.

[0027] S43: Standardize the calculated values ​​of the three types of independent variables to form a dataset for fitting the multiple linear regression model.

[0028] In conjunction with the first aspect, the present invention provides a fifth specific implementation of the first aspect, specifically, the calculation of three types of independent variable index values ​​based on the spatial data for each pair of recreational nodes specifically includes:

[0029] S421: The resource background suitability index value is one or more of the following, calculated based on the vector boundary data, land use / cover data and point of interest data: area, internal facility density, and natural landscape area ratio of each recreation node pair;

[0030] S422: The ecological landscape visibility index value is calculated based on the digital elevation model data, which calculates the visual visibility probability or cumulative visible area between recreation node pairs.

[0031] S423: The traffic connectivity convenience index is calculated based on the road network data, using the shortest path algorithm to determine the network distance, travel time cost, or number of topological connections between two points.

[0032] In conjunction with the first aspect, the present invention provides a sixth specific implementation of the first aspect. Specifically, for the region to be predicted, based on the spatial data of all recreational nodes in the region to be predicted, the independent variable index values ​​of all recreational node pairs in the region to be predicted are calculated, and substituted into the trained recreational activity connection frequency prediction model to predict the observed recreational activity connection frequency values ​​between all recreational node pairs in the region to be predicted. This specifically includes:

[0033] S61: Determine the boundary range of the area to be predicted;

[0034] S62: Obtain a list of standardized names for all recreational nodes within the stated boundary area through publicly available tourism information service systems, geographic information databases, or field surveys;

[0035] S63: Using geocoding services, convert the list of standardized names into corresponding spatial coordinate data;

[0036] S64: Perform coordinate correction and spatial registration processing on the spatial coordinate data to generate spatial vector data of recreational nodes in the area to be predicted;

[0037] S65: Calculate the independent variable index value corresponding to each pair of recreational nodes in the area to be predicted based on the spatial vector data;

[0038] S66: Substitute the independent variable index value corresponding to each pair of recreational nodes in the area to be predicted into the trained recreational activity connection frequency prediction model, and simulate and calculate the recreational activity connection frequency prediction value between each pair of recreational nodes.

[0039] In conjunction with the first aspect, the present invention provides a seventh specific implementation of the first aspect. Specifically, the step of selecting all recreational node pairs with recreational activity connection frequency observations higher than the average value based on the observed values ​​of recreational activity connection frequency observations among all recreational node pairs, and generating a recreational path selection scheme for the recreational node pairs in the area to be predicted, specifically includes:

[0040] S71: Based on the predicted frequency of recreational activity connections between all pairs of recreational nodes obtained from simulation prediction, calculate the arithmetic mean of all predicted frequencies of recreational activity connections, and use it as the screening threshold;

[0041] S72: Select recreational node pairs from all recreational node pairs whose predicted recreational activity connection frequency is higher than the selection threshold to form a set of node pairs with high connection strength.

[0042] S73: Using all the recreational node pairs in the set of high-connection-strength node pairs as basic connection units, perform network connection and topology merging among the recreational nodes in the region to be predicted;

[0043] S74: Based on the merged network structure, extract the main activity connection paths and generate a recreational path selection scheme for the area to be predicted.

[0044] Secondly, the present invention also provides an eco-recreation path simulation and route selection system based on natural resource data, comprising:

[0045] The data acquisition module is used to determine the target area and obtain the names and spatial coordinates of all recreational nodes within the target area. The recreational nodes can be any one of urban parks, green squares, natural landscapes, cultural heritage sites, museums, and commercial and leisure areas. The module is also used to calculate the frequency of co-occurrence of each pair of recreational nodes in the same user's itinerary or published content based on social media data, generate a data table of recreational node pairs and their corresponding co-occurrence frequencies, and use the co-occurrence frequency as the recreational activity connection frequency observation value of the corresponding recreational node pair.

[0046] The model building module is used to build a multiple linear regression model, wherein the multiple linear regression model uses the observed frequency of recreational activity connections as the dependent variable and the combination of spatial data indicators that characterize the resource background suitability, ecological landscape visibility and transportation connectivity of the corresponding recreational node pair as independent variables.

[0047] The indicator calculation module is used to calculate the independent variable indicator values ​​of all recreational node pairs based on the spatial data of all recreational nodes in the target area.

[0048] The model fitting module is used to fit the observed values ​​of recreational activity connection frequency with the corresponding independent variable index values, solve the undetermined coefficients of the multiple linear regression model, and obtain the trained recreational activity connection frequency prediction model.

[0049] The prediction application module is used to calculate the independent variable index values ​​of all recreational node pairs in the area to be predicted based on the spatial data of all recreational nodes in the area to be predicted, and substitute them into the trained recreational activity connection frequency prediction model to predict the observed values ​​of recreational activity connection frequency between all recreational node pairs in the area to be predicted.

[0050] The route selection scheme generation module is used to filter out all recreational node pairs whose recreational activity connection frequency observation values ​​are higher than the average value based on the observed values ​​of recreational activity connection frequency among all recreational node pairs, and generate a recreational path selection scheme for the recreational node pairs in the area to be predicted.

[0051] In conjunction with the second aspect, the present invention provides a first specific implementation of the second aspect, wherein the data acquisition module performs the following steps:

[0052] Obtain the administrative division vector boundary data of the target area;

[0053] Collect a list of names of all recreational nodes within the target area from tourism information platforms and / or local data;

[0054] Geographically analyze the recreation nodes in the name list to obtain their geographic coordinates;

[0055] The geographic coordinates are corrected and spatially positioned to generate vector data of recreational nodes with spatial location information.

[0056] Compared with the prior art, the beneficial effects of the present invention are:

[0057] By leveraging the frequency of shared recreational nodes on social media to replace traditional empirical judgments, the observation and prediction of recreational connection strength can be grounded in actual behavioral data, enhancing the objectivity and reliability of the analysis. The constructed multiple linear regression model systematically integrates three spatial indicators: resource background, ecological landscape, and transportation connectivity. It can analyze key environmental factors influencing recreational activities from multiple dimensions, shifting route selection criteria from single-distance judgments to comprehensive spatial suitability assessments, resulting in a more scientific and comprehensive analysis. This method possesses strong extrapolation and prediction capabilities. After calibration in the target area, the model can be directly applied to areas lacking data for prediction, simulating and generating recreational connection networks to support new area planning or current status assessments. The recreational connection frequencies predicted by the model can be used to screen high-connection-strength node pairs, thereby generating feasible recreational route selection schemes. This transforms complex spatial analysis into intuitive planning guidelines, enhancing the practicality of the results in land spatial planning, tourism route design, and green space system optimization. Attached Figure Description

[0058] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, wherein:

[0059] Figure 1 This is a flowchart of the eco-recreation path simulation and route selection method based on natural resource data of the present invention. Detailed Implementation

[0060] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0061] like Figure 1 The flowchart shown is a process for simulating and selecting eco-recreational routes based on natural resource data, as described in this invention.

[0062] Example 1

[0063] like Figure 1As shown, the method for simulating and selecting ecological recreational paths based on natural resource data includes the following steps: determining the target area and obtaining the names and spatial coordinates of all recreational nodes within the target area, wherein the recreational nodes can be any one of urban parks, green squares, natural landscapes, cultural heritage sites, museums, and commercial and leisure areas; based on social media data, calculating the frequency of co-occurrence of each pair of recreational nodes in the same user's itinerary or published content, generating a data table of recreational node pairs and their corresponding co-occurrence frequencies, and using the co-occurrence frequency as the observed value of recreational activity connection frequency for the corresponding recreational node pair; constructing a multiple linear regression model, wherein the multiple linear regression model uses the observed value of recreational activity connection frequency as the dependent variable, and uses a combination of spatial data indicators characterizing the resource background suitability, ecological landscape visibility, and transportation connectivity convenience of the corresponding recreational node pair as the dependent variable. Independent variables: Based on the spatial data of all recreational nodes in the target area, calculate the independent variable index values ​​for all recreational node pairs; fit the observed values ​​of recreational activity connection frequency with the corresponding independent variable index values, solve for the undetermined coefficients of the multiple linear regression model, and obtain the trained recreational activity connection frequency prediction model; for the area to be predicted, based on the spatial data of all recreational nodes in the area to be predicted, calculate the independent variable index values ​​for all recreational node pairs in the area to be predicted, and substitute them into the trained recreational activity connection frequency prediction model to predict the observed values ​​of recreational activity connection frequency between all recreational node pairs in the area to be predicted; based on the observed values ​​of recreational activity connection frequency between all recreational node pairs, select all recreational node pairs whose observed values ​​of recreational activity connection frequency are higher than the average value, and generate a recreational path selection scheme for the recreational node pairs in the area to be predicted.

[0064] Specifically, by utilizing the frequency of recreational node pairs appearing together on social media platforms to replace traditional experience-based judgments, the observation and prediction of recreational connection strength are grounded in objective, massive amounts of actual behavioral data, significantly improving the objectivity and reliability of the analysis. The constructed multiple linear regression model systematically integrates three major categories of spatial indicators: resource background, ecological landscape, and transportation connectivity. It can analyze and quantify key environmental factors affecting recreational activities from multiple dimensions, deepening the route selection basis from a single distance judgment to a comprehensive spatial suitability assessment, making the analysis more scientific and comprehensive. This method has a powerful extrapolation and prediction function. Once the model is calibrated in the target area, it can be directly applied to areas with scarce data to simulate and generate recreational connection networks, thereby effectively supporting new area planning or current status assessment. The predicted frequency values ​​of recreational activity connections output by the model can be directly used to screen out node pairs with high connection strength, and use these as the core to generate feasible recreational route selection schemes, transforming complex spatial analysis into intuitive planning guidelines. This greatly enhances the practicality and decision support value of the research results in land spatial planning, tourism route design, and green space system optimization.

[0065] In a preferred embodiment, S1: Determine the target area and obtain the names and spatial coordinates of all recreational nodes within the target area, wherein the recreational nodes are any one of urban parks, plaza green spaces, natural landscapes, cultural heritage sites, museums, and commercial and leisure areas.

[0066] In a preferred embodiment, determining the target area and obtaining the names and spatial coordinates of all recreational nodes within the target area specifically includes:

[0067] S11: Obtain the administrative division vector boundary data of the target area;

[0068] Specifically, the vector boundary data of the target area is loaded and determined. The vector boundary data is usually the polygonal data of the administrative division of a city or district, which serves as the benchmark range for all subsequent spatial analysis and data collection.

[0069] S12: Collect a list of names of all recreational nodes in the target area from tourism information platforms and / or local data.

[0070] Specifically, the data collection sources include official documents such as government-released green space system plans and cultural tourism directories, as well as popular location information from mainstream travel social media platforms such as Ctrip, Xiaohongshu, and Douyin. Through web data scraping, public API interface calls, or supplementary on-site surveys, the data collection sources are used to obtain and integrate a list of recreational nodes, including all urban parks, forest parks, wetland parks, scenic spots, historical sites, and museums within the target area.

[0071] S13: Perform geographic parsing on the recreation nodes in the name list to obtain their geographic coordinates.

[0072] Specifically, geocoding technology is used to perform geographic parsing on each recreation node in the above list. This can be achieved by calling APIs from open platforms such as Amap and Baidu Maps, or by using a local address database, to convert text names into precise geographic coordinates (such as latitude and longitude in the WGS84 or CGCS2000 coordinate systems), thus completing the initial spatial location.

[0073] S14: Correct and spatially locate the geographic coordinates to generate recreation node vector data with spatial location information.

[0074] Specifically, the obtained coordinate point data is spatially matched and verified against the vector boundary of the target area. First, coordinate correction is performed (converting to a unified national geodetic coordinate system if necessary). Second, using GIS software, the spatial location of the coordinate points is verified against the administrative boundary of S11. Anomalies clearly falling outside the target area are filtered and eliminated through a spatial query of "points within the area," and points on the edge or questionable are manually reviewed and corrected. Finally, a vector layer of recreational node points with accurate spatial location and complete attribute information is output.

[0075] S2: Based on social media data, count the frequency of each pair of recreational nodes appearing together in the same user's itinerary or published content, generate a data table of recreational node pairs and their corresponding frequency of co-occurrence, and use the frequency of co-occurrence as the recreational activity connection frequency observation value of the corresponding recreational node pair.

[0076] Specifically, one or more social media platforms (such as Weibo, Xiaohongshu, Mafengwo, TripAdvisor, and Ctrip) containing rich geotags and text content are selected as data sources. User-generated content with spatial location information published in the target area within a specific historical period (such as the past year) is collected using their provided open APIs or authorized, compliant web crawling tools. Each data entry should include at least the publisher's identifier, publication time, text content (title, body, tags), and associated geographic coordinates or geotag information. The collected raw data is cleaned by deduplication, removal of advertisements, and irrelevant noise. Next, the geographic coordinates of each data entry are spatially matched with the vector boundary of the target area, retaining only data whose coordinates fall within the target area to form a valid data sample set. Natural language processing techniques are used to extract keywords and identify names of parks, attractions, and other recreational nodes mentioned in the cleaned text content (such as titles, body, and tags). Simultaneously, spatial proximity analysis is performed between the geographic coordinates of the data and the geographic coordinates of the recreational nodes generated in step S1. If the coordinates fall within a preset buffer zone (e.g., 500 meters) of a node, the recreational node is considered relevant. Based on the association results, all pairs of recreational nodes are traversed. For any two different recreational nodes A and B, the number of times they co-occur under any of the following conditions is counted:

[0077] Both are mentioned in the same post;

[0078] It is mentioned or associated separately in multiple related posts published by the same user within a preset time interval (such as 24 hours).

[0079] The above number is defined as the frequency observation value of recreational activity association between A and B. By traversing and counting all pairs of resting nodes, a symmetric system is ultimately generated. Connect the frequency observation matrix (where N is the total number of recreational nodes), and the off-diagonal elements of the matrix. This represents the observed frequency of recreational activity connections between the i-th and j-th recreational nodes. This matrix can be directly stored as a table showing the node pair-frequency correspondence. Based on the aforementioned statistical results, it is organized into a structured table. The table structure must contain at least three columns: "Node A Name," "Node B Name," and "Frequency of Co-occurrence," used to record all non-repeating node pairs and their observed frequencies.

[0080] In a preferred embodiment, based on social media data, the frequency of co-occurrence of each pair of recreational nodes in the same user's itinerary or published content is statistically analyzed to generate a data table of recreational node pairs and their corresponding co-occurrence frequencies. The co-occurrence frequency is used as the recreational activity connection frequency observation value for the corresponding recreational node pair, specifically including:

[0081] S21: Construct query conditions from at least one social media platform using the name of the target area and keywords related to recreational activities, collect post data published within a preset time period using a web crawler, and store the collection results as a structured text file.

[0082] Specifically, based on the geographical scope and recreational activity characteristics of the target area, multiple sets of query keywords (such as "[City Name] Park Hiking" and "[Region Name] Scenic Spot Tour Route") are constructed, including the area name, attraction type, and activity type. User-generated content published within a preset historical time period (the past year, typically referring to the 12 months preceding the collection date) is collected through official API interfaces or compliant web crawling technologies provided by social media platforms (such as Weibo, Xiaohongshu, and Douyin). Each piece of data collected should include text content, publisher ID, publication time, and geographical coordinates (such as location markers), and be saved as a structured text file.

[0083] S22: Clean the post data in the text file and filter out records whose geographical location belongs to the target area and whose text content contains one or more recreation node names as valid recreation records.

[0084] Specifically, the acquired post data undergoes data cleaning, including but not limited to deleting completely duplicate records, filtering advertising or marketing content, removing noise information irrelevant to recreational activities, and standardizing or removing non-textual content (such as emojis, links, and meaningless characters). Geographic information processing methods are used to spatially match the geographic coordinates of each data record with the vector boundaries of the target area, retaining only records whose coordinates fall within that area. A standard name thesaurus is constructed based on a list of recreational node names. Keyword matching or named entity recognition technology is used to scan the text content of each record (including title, body, and tags) to identify whether it contains any recreational node name from the thesaurus. If a record mentions at least one node name, the node and its context information are recorded. Text recognition is further performed on the location-filtered data to identify the recreational nodes mentioned in the text. Simultaneously, for data relying solely on textual mentions but with unclear spatial coordinates, location aliases and proximity descriptions are used for auxiliary judgment and association to improve the completeness and accuracy of valid records. Records confirmed through both spatial and textual filtering are merged and organized to form a valid recreational record dataset. Each valid record should include the following information: original post ID or link, posting time, geographic coordinates (if available), a list of mentioned recreational node names, and a relevant text summary.

[0085] S23: Construct a co-occurrence relation matrix for recreational nodes. For each pair of recreational nodes, count the frequency of the two recreational node names appearing simultaneously in the same post or the same user's trip in all valid recreational records. Use this as the recreational activity connection frequency observation value for the recreational node pair and organize it into a data table with recreational node pairs as the index and recreational activity connection frequency observation values ​​as the data items.

[0086] Specifically, based on the obtained effective recreational records, a system of size [size missing] is constructed. Connection frequency observation matrix (Where N is the total number of recreational nodes within the target area). Off-diagonal elements in the matrix. This represents the frequency observation of recreational activity connections between the i-th and j-th recreational nodes. For any two distinct recreational nodes... and ( Iterate through all valid recreational records and count their co-occurrence frequency. Co-occurrence includes two determination cases:

[0087] Single Post Co-occurrence: Recreation Nodes and The names appear simultaneously in the text content or tags of the same post;

[0088] Co-occurrence of Trips: Recreation Nodes and These appear in multiple posts by the same user within a preset time window (usually set to 24 hours) that have a clear connection to a specific trip.

[0089] Statistical results of recreation node pairs The total number of co-occurrences is filled into the matrix. The corresponding position The matrix satisfies symmetry, that is... Diagonal elements Setting it to 0 indicates that a node does not co-occur with itself. This relates to the symmetric matrix. Convert the data into a long table format structured table, which should contain at least three columns: "Node A Name", "Node B Name", and "Frequency of Co-occurrence". Each row represents a unique pair of nodes and its observation frequency.

[0090] In a more specific embodiment, the step of "acquiring social media data and counting the frequency of each pair of recreational nodes appearing together in the same user's itinerary or published content" can be implemented based on a specific platform and tools according to the following detailed process:

[0091] Taking the Xiaohongshu platform as an example, using a Python programming environment: First, run a pre-written web crawler script (e.g., getPOSTList.py) in the IDLE tool. In the script, set search keywords related to the target area and recreational activities (e.g., "Beijing park tour routes," "Shanghai attraction walking guide"), and configure it to crawl posts published within the past year. After the script runs, it will automatically obtain and save the core information of the posts (e.g., post ID, publisher, text, tags, publication time), and store it as a structured CSV file. Import the CSV file generated in the previous step into Excel for preliminary processing. First, anonymize the data and use Excel's "remove duplicates" function to remove completely duplicate records based on post ID or text content, ensuring data uniqueness. Second, to filter posts strongly related to recreational routes, construct a keyword library containing two dimensions: "recreational route" and "recreational location." Route dimension keywords include "route recommendations," "walking trails," "sightseeing routes," and "connections"; location dimension keywords include the specific names of recreational nodes. Two complementary methods can be used for filtering:

[0092] Rule matching: Using Excel's COUNTIF and other functions, count the frequency of keyword matching in the post content and set a threshold for initial screening.

[0093] Model selection: The text content is imported into a machine learning model (such as TextCNN or BERT) for binary classification training and prediction, automatically identifying posts highly relevant to "recreational route planning". Through cleaning and selection, the final set of effective recreational records is obtained.

[0094] Based on the valid recreational record set, the co-occurrence frequency of recreational node pairs is statistically analyzed. For each different pair of recreational nodes (e.g., "Summer Palace" and "Yuanmingyuan"), Excel's COUNTIFS function is used for traversal and statistical analysis: for each post record, it is determined whether the two names appear simultaneously in the post's body or tags. The total number of times this combination appears in all posts is counted, and this total number is defined as the recreational activity connection frequency observation value between the two locations. All obtained recreational node pairs and their corresponding connection frequency observation values ​​are summarized. In Excel, the connection pair names are in the form of "Recreational Node A_Recreational Node B", and the recreational activity connection frequency observation values ​​are used as data, forming a structured data table.

[0095] S3: Construct a multiple linear regression model, wherein the multiple linear regression model uses the observed frequency of recreational activity connections as the dependent variable and the combination of spatial data indicators representing the resource suitability, ecological landscape visibility, and transportation connectivity convenience of the corresponding recreational node pair as independent variables.

[0096] Specifically, this method draws on the basic idea of ​​the "gravity model" in spatial interaction theory, which states that the intensity of interaction between two locations is directly proportional to their respective attractive forces and inversely proportional to their resistance forces. Based on this, a multiple linear regression model is constructed as the core prediction tool. This model uses the observed frequency of recreational activity connections as the dependent variable (Y), and a set of spatial data indicators (X) corresponding to the resource suitability, ecological landscape visibility, and transportation connectivity of recreational node pairs as independent variables. An indicator system is constructed, selecting various resource-related spatial data such as resource suitability (elevation, slope, aspect, vegetation coverage), ecological landscape visibility (mountain visibility, water visibility), and transportation connectivity convenience (minimum actual path length between any two recreational node pairs) as characteristic independent variables, which are then used as analytical parameters of the gravity model. Substituting these characteristic independent variables into the multiple linear regression equation, a mathematical relationship is constructed between the observed frequency of recreational activity connections and the corresponding independent variable indicators of the recreational node pairs. The specific formula is shown below: Its expression is: In the above formula, The frequency of recreational activity connections between recreational node i and recreational node j is represented by C, S, and L, which represent relevant indicators in terms of resource suitability, ecological landscape visibility, and transportation connectivity convenience, respectively. For elevation indicators, For slope index, As an aspect indicator, The vegetation coverage rate is an indicator. This is an indicator of the visibility of mountains. As an indicator of water body visibility; It represents the minimum actual path length between any pair of resting nodes.

[0097] S4: Based on the spatial data of all recreational nodes in the target area, calculate the independent variable index values ​​for all recreational node pairs.

[0098] In a preferred embodiment, calculating the independent variable index values ​​for all recreational node pairs based on the spatial data of all recreational nodes in the target area specifically includes:

[0099] S41: Acquire and integrate spatial data of all recreational nodes in the target area, the spatial data including: vector boundary data of recreational nodes, digital elevation model data, land use / cover data, point of interest data and road network data;

[0100] Specifically, multi-source geospatial data within the target area is collected and preprocessed to provide a foundation for indicator calculation. The required core data includes:

[0101] Vector boundary data of recreational nodes: namely, the point vector data and its service range surface data generated in step S14.

[0102] Digital elevation model data: used to extract topographic factors such as elevation, slope, and aspect.

[0103] Land use / land cover data: used to extract information on land cover such as vegetation, water bodies, and built-up areas.

[0104] Road network data: Vector line data containing road classifications and traffic attributes, used to analyze connectivity and calculate path distances.

[0105] Points of interest data: used as supplementary data to verify and supplement attraction information.

[0106] S42: For each pair of recreational nodes, calculate the values ​​of three types of independent variable indicators based on the spatial data. The three types of independent variable indicators include the resource background suitability indicator, the ecological landscape visibility indicator, and the transportation connection convenience indicator.

[0107] Specifically, the resource background suitability index is an indicator calculated based on the vector boundary data, land use / cover data, and point-of-interest data, representing the recreational service capacity of the recreation node itself, including:

[0108] Area index: Calculate the actual area occupied by each recreation node;

[0109] Facility density index: The ratio of the number of points of interest within or near all recreational nodes to the area of ​​the node, reflecting the degree of completeness of service facilities;

[0110] Natural landscape area ratio: Based on land use type data, calculate the area ratio of natural landscapes such as green spaces and water bodies within all recreational nodes.

[0111] The ecological landscape visibility index is based on the digital elevation model data and uses the view domain analysis method to determine the visual connectivity between nodes. Specific indicators are as follows:

[0112] Visibility probability: Calculates the probability of observing recreational node B from recreational node A;

[0113] Cumulative visible area: Calculates the total area or proportion of the visible area between recreational node pairs.

[0114] The transportation connectivity convenience index is based on the road network data and is calculated using a shortest path algorithm to determine the spatial accessibility between recreational nodes. The specific index is as follows:

[0115] Network shortest distance: Calculate the shortest path length along the road from recreation node A to recreation node B;

[0116] Travel time cost: Calculate the shortest travel time between recreation node pairs by combining road type and speed limit;

[0117] Topology connection count: The number of key intersections or crossroads along the shortest path connecting pairs of recreational nodes.

[0118] S43: Standardize the calculated values ​​of the three types of independent variables to form a dataset for fitting the multiple linear regression model.

[0119] Specifically, due to differences in the dimensions and numerical ranges of the various indicators, a standardization method is used to render the original indicator values ​​dimensionless. The Z-score standardization method can be used, and the specific calculation formula is as follows: ;in, This represents the original value of a certain independent variable indicator. This is the average value of the index across all pairs of recreational nodes. Standard deviation The standardized index values ​​are then used. All standardized independent variable index values ​​are matched and integrated with the recreational activity connection frequency observations obtained in step S23 for the corresponding recreational node pairs to construct a structured dataset. This dataset consists of one record for each node pair, including the following fields: Node A name, Node B name, Resource background suitability index 1 value (standardized), Resource background suitability index 2 value (standardized)... Ecological landscape visibility index value (standardized), Transportation connectivity convenience index 1 value (standardized), Transportation connectivity convenience index 2 value (standardized)... and recreational activity connection frequency observations.

[0120] Before modeling, the dataset should be validated, including: checking for missing and outlier values; truncating or imputing extreme values; and verifying multicollinearity among independent variables (e.g., calculating the variance inflation factor, VIF). The final processed dataset should be exported as a common format file (e.g., .csv or .xlsx) for import into statistical analysis software or programming environments for subsequent model fitting.

[0121] S5: Fit the observed values ​​of recreational activity connection frequency with the corresponding independent variable index values, solve the undetermined coefficients of the multiple linear regression model, and obtain the trained recreational activity connection frequency prediction model.

[0122] Specifically, ensure that each recreational node is suitable for... Each row corresponds to a complete line of observation data, which contains a dependent variable (related to the frequency observations). ) and multiple independent variables (standardized spatial index values) This process creates a final structured data table for model fitting. Using statistical analysis software (such as SPSS, R, or Python's statsmodels library), the integrated data is input into the established initial equation framework for multiple linear regression. The least squares method is used for model fitting, and mathematical optimization algorithms are employed to solve for the undetermined coefficients in the equations. This process will determine the specific weight of each independent variable on the frequency of recreational interactions, thus obtaining a preliminary regression equation with clearly defined coefficient parameters. The fitted preliminary regression model will then undergo statistical testing and performance evaluation to ensure its validity and reliability. Key evaluation steps include:

[0123] Significance testing: The F-test is used to determine whether the model as a whole is statistically significant; the t-test is used to determine whether the coefficients of each independent variable are significantly different from zero. Insignificant variables are considered for elimination or re-examination based on domain knowledge.

[0124] Goodness-of-fit assessment: Calculation of coefficient of determination With adjustment To evaluate the model's ability to explain the variation in observed data.

[0125] Multicollinearity diagnosis: Calculate the variance inflation factor to check for high correlations among independent variables in order to avoid model distortion.

[0126] Residual analysis: Check whether the residuals meet the assumptions of normality, independence, etc., to verify the applicability of the model.

[0127] Based on the test and evaluation results, the preliminary regression equation was optimized as necessary (e.g., removing insignificant variables and handling outliers). Finally, a validated multiple linear regression equation with determined coefficients was obtained, which is a quantitative model for predicting the frequency of recreational activity connections. This model can be used to predict the strength of potential recreational activity connections between recreational nodes given their spatial index values. Its final mathematical expression is as follows: ;in, This represents the frequency observation of recreational activity connections between recreational nodes i and j. To obtain the specific coefficient values, Input values ​​for the standardized independent variables.

[0128] S6: For the area to be predicted, based on the spatial data of all recreational nodes in the area to be predicted, calculate the independent variable index values ​​of all recreational node pairs in the area to be predicted, and substitute them into the trained recreational activity connection frequency prediction model to predict the observed values ​​of recreational activity connection frequencies between all recreational node pairs in the area to be predicted.

[0129] In a preferred embodiment, for the area to be predicted, based on the spatial data of all recreational nodes in the area to be predicted, the independent variable index values ​​of all recreational node pairs in the area to be predicted are calculated, and substituted into the trained recreational activity connection frequency prediction model to predict the observed recreational activity connection frequency values ​​between all recreational node pairs in the area to be predicted, specifically including:

[0130] S61: Determine the boundary of the area to be predicted; specifically, identify the area where recreational route simulation will be conducted, and obtain the administrative division vector boundary data of this area as a geographic benchmark for data processing and spatial analysis. This data can be downloaded from national or provincial geographic information public service platforms.

[0131] S62: Obtain a standardized list of names for all recreational nodes within the stated boundary area through publicly available tourism information service systems, geographic information databases, or field surveys; specifically, obtain an authoritative list of names for all recreational nodes within the boundary area through various means. The main means include: querying official directories and planning documents from local natural resources and cultural tourism departments; filtering from the point-of-interest databases of map service providers such as Gaode and Baidu; and using web crawling technology to scrape popular location names in the area from tourism booking platforms and social media.

[0132] S63: Using geocoding services, convert the list of standardized names into corresponding spatial coordinate data; specifically, call the geocoding API service provided by Amap or Baidu Maps, taking the list of standardized names obtained in step S62 as input, and batch request and obtain the latitude and longitude coordinates corresponding to each name. The output result is a preliminary point data table containing the fields "name-longitude-latitude".

[0133] S64: Perform coordinate correction and spatial registration processing on the spatial coordinate data to generate spatial vector data of recreational nodes in the area to be predicted. Specifically, to ensure consistency with the spatial reference used during model training, the coordinates obtained in step S63 need to be uniformly converted to the same coordinate system as the target area data. Subsequently, use GIS software to perform spatial overlay analysis on the converted coordinate points and the boundary of the area to be predicted, verify and remove outliers that obviously fall outside the area, and generate the final accurate spatial vector point layer of recreational nodes.

[0134] S65: Calculate the independent variable index values ​​corresponding to each pair of recreational nodes within the area to be predicted based on the spatial vector data. Specifically, for the area to be predicted, repeat the process of step S4. That is, acquire multi-source spatial data such as the digital elevation model, land use, and road network of the area, and calculate the corresponding resource background suitability, ecological landscape visibility, and transportation connectivity convenience index values ​​for each pair of recreational nodes in the vector point layer generated in step S64. Crucially, the index values ​​here need to be standardized using the mean and standard deviation determined based on the target area data during model fitting in step S5 to ensure consistency with the input scale of the prediction model.

[0135] S66: Substitute the independent variable index values ​​corresponding to each pair of recreational nodes in the area to be predicted into the trained recreational activity connection frequency prediction model to simulate and calculate the predicted recreational activity connection frequency value between each pair of recreational nodes. Specifically, calculate and standardize the independent variable index values ​​of each pair of recreational nodes, and use them as input data, substituting them one by one into the "trained recreational activity connection frequency prediction model" determined in step S5. Through model calculation, obtain the predicted recreational activity connection frequency value between each pair of recreational nodes in the area to be predicted. By summarizing all recreational node pairs and their recreational activity connection frequency prediction values, a complete simulated network data table reflecting the potential recreational connection strength of all recreational node pairs in the area to be predicted can be generated.

[0136] S7: Based on the observed frequency of recreational activity connections between all recreational node pairs, select all recreational node pairs whose observed frequency of recreational activity connections is higher than the average value, and generate a recreational path selection scheme for the recreational node pairs in the area to be predicted.

[0137] In a preferred embodiment, the step of filtering out all recreational node pairs whose recreational activity connection frequency observations are higher than the average based on the observed values ​​of recreational activity connection frequency observations among all recreational node pairs, and generating a recreational path selection scheme for the recreational node pairs in the area to be predicted, specifically includes:

[0138] S71: Based on the predicted frequency values ​​of recreational activity connections between all pairs of recreational nodes obtained from simulation prediction, calculate the arithmetic mean of all predicted frequencies of recreational activity connections, and use it as the screening threshold; specifically, let the area to be predicted have a total of If there are ____ recreational nodes, then the number of corresponding recreational node pairs is ____. The predicted frequency values ​​of recreational activity connections for all pairs of recreational nodes constitute a set. Filtering threshold The calculation method is as follows: ;in, This represents the total number of recreational node pairs. This threshold reflects the overall average level of recreational connection strength within the area to be predicted and is used to identify recreational node pairs with high connection strength.

[0139] S72: From all recreational node pairs, select recreational node pairs whose predicted recreational activity connection frequency is higher than the selection threshold to form a set of node pairs with high connection strength. Specifically, construct the set: ;in, This refers to the set of node pairs with high connection strength. Node pairs in this set are considered key node pairs with significant potential recreational connections and can serve as basic connection units for subsequent path construction and network merging.

[0140] S73: Using all the recreational node pairs in the set of high-connection-strength node pairs as basic connection units, network connections and topology merging are performed among the recreational nodes in the region to be predicted.

[0141] Specifically, taking all recreational nodes as vertices, the set Each node pair in the network is spatially connected to form an initial set of undirected connections. The topological relationships of these connections reflect the potential pathways between nodes based on the predicted connection strength. The predicted connection frequency values ​​of the corresponding node pairs obtained from the simulation in step S66 are stored and associated as important attributes of the connections. Based on the initial network, topology merging is performed on the following cases:

[0142] If multiple connecting lines form a continuous chain structure (such as A—B, B—C) and satisfy certain directional continuity and intensity consistency, they can be merged into a single multi-segment connecting line representing the main recreational route.

[0143] If a node has only a single connection line and the connection strength is low, connection pruning can be performed based on set rules to highlight the main connection structure in the network.

[0144] Visualizing the connecting lines and their corresponding connection strength values ​​can be achieved using a hierarchical symbolization method, such as using line thickness or color intensity to represent the frequency of connections, thereby enhancing the readability of the network structure.

[0145] S74: Based on the merged network structure, extract the main activity connection paths and generate a recreational path selection scheme for the area to be predicted.

[0146] Specifically, topological and statistical analysis is performed on the merged networks to identify:

[0147] Key hub nodes: Nodes with high connectivity in the network that are located at the intersection of multiple paths;

[0148] Main connecting corridor: A path segment consisting of a series of frequently connected lines;

[0149] Cluster structure: a tightly connected, relatively independent subgroup of a local network.

[0150] Core connection paths are extracted based on the following principles: connection frequency higher than the regional average; backbone paths forming a ring or chain structure; and corridors with spatial continuity and functional connectivity.

[0151] The extracted paths undergo topology optimization processes such as spatial smoothing and breakpoint connection to ensure their feasibility and connectivity in real-world spaces. The finalized core recreational connection paths are output in a standard geographic data format (e.g., Shapefile, GeoJSON) with accompanying attribute information, including: path number and name; sequence of nodes along the route; predicted connection frequency statistics (e.g., average, maximum); and suggested path types (e.g., walking path, cycling path, scenic route). This output file constitutes a recreational path selection scheme based on simulation predictions, which can be used for planning support or spatial optimization design.

[0152] Example 2

[0153] An eco-recreation route simulation and selection system based on natural resource data includes:

[0154] The data acquisition module is used to determine the target area and obtain the names and spatial coordinates of all recreational nodes within the target area. The recreational nodes can be any one of urban parks, green squares, natural landscapes, cultural heritage sites, museums, and commercial and leisure areas. The module is also used to calculate the frequency of co-occurrence of each pair of recreational nodes in the same user's itinerary or published content based on social media data, generate a data table of recreational node pairs and their corresponding co-occurrence frequencies, and use the co-occurrence frequency as the recreational activity connection frequency observation value of the corresponding recreational node pair.

[0155] The model building module is used to build a multiple linear regression model, wherein the multiple linear regression model uses the observed frequency of recreational activity connections as the dependent variable and the combination of spatial data indicators that characterize the resource background suitability, ecological landscape visibility and transportation connectivity of the corresponding recreational node pair as independent variables.

[0156] The indicator calculation module is used to calculate the independent variable indicator values ​​of all recreational node pairs based on the spatial data of all recreational nodes in the target area.

[0157] The model fitting module is used to fit the observed values ​​of recreational activity connection frequency with the corresponding independent variable index values, solve the undetermined coefficients of the multiple linear regression model, and obtain the trained recreational activity connection frequency prediction model.

[0158] The prediction application module is used to calculate the independent variable index values ​​of all recreational node pairs in the area to be predicted based on the spatial data of all recreational nodes in the area to be predicted, and substitute them into the trained recreational activity connection frequency prediction model to predict the observed values ​​of recreational activity connection frequency between all recreational node pairs in the area to be predicted.

[0159] The route selection scheme generation module is used to filter out all recreational node pairs whose recreational activity connection frequency observation values ​​are higher than the average value based on the observed values ​​of recreational activity connection frequency among all recreational node pairs, and generate a recreational path selection scheme for the recreational node pairs in the area to be predicted.

[0160] In a preferred embodiment, the data acquisition module performs the following steps:

[0161] Obtain the administrative division vector boundary data of the target area;

[0162] Collect a list of names of all recreational nodes within the target area from tourism information platforms and / or local data;

[0163] Geographically analyze the recreation nodes in the name list to obtain their geographic coordinates;

[0164] The geographic coordinates are corrected and spatially positioned to generate vector data of recreational nodes with spatial location information.

[0165] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Therefore, any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for simulating and selecting eco-recreational routes based on natural resource data, characterized in that, Includes the following steps: S1: Determine the target area and obtain the names and spatial coordinates of all recreational nodes within the target area, wherein the recreational nodes are any one of the following: urban parks, green squares, natural landscapes, cultural heritage sites, museums, and commercial and leisure areas; S2: Based on social media data, count the frequency of each pair of recreational nodes appearing together in the same user's itinerary or published content, generate a data table of recreational node pairs and their corresponding frequency of co-occurrence, and use the frequency of co-occurrence as the recreational activity connection frequency observation value of the corresponding recreational node pair; S3: Construct a multiple linear regression model, wherein the multiple linear regression model uses the observed frequency of recreational activity connections as the dependent variable and the combination of spatial data indicators representing the resource suitability, ecological landscape visibility and transportation connectivity of the corresponding recreational node pair as independent variables. S4: Based on the spatial data of all recreational nodes in the target area, calculate the independent variable index values ​​for all recreational node pairs; S5: Fit the observed values ​​of recreational activity connection frequency with the corresponding independent variable index values, solve the undetermined coefficients of the multiple linear regression model, and obtain the trained recreational activity connection frequency prediction model. S6: For the area to be predicted, based on the spatial data of all recreational nodes in the area to be predicted, calculate the independent variable index values ​​of all recreational node pairs in the area to be predicted, and substitute them into the trained recreational activity connection frequency prediction model to predict the observed values ​​of recreational activity connection frequency between all recreational node pairs in the area to be predicted. S7: Based on the observed frequency of recreational activity connections among all recreational node pairs, select all recreational node pairs whose observed frequency of recreational activity connections is higher than the average value, and generate a recreational path selection scheme for the recreational node pairs in the area to be predicted.

2. The method for simulating and selecting eco-recreational routes based on natural resource data according to claim 1, characterized in that, The process of determining the target area and obtaining the names and spatial coordinates of all recreational nodes within the target area specifically includes: S11: Obtain the administrative division vector boundary data of the target area; S12: Collect a list of names of all recreational nodes in the target area from tourism information platforms and / or local data; S13: Perform geographic parsing on the recreation nodes in the name list to obtain their geographic coordinates; S14: Correct and spatially locate the geographic coordinates to generate recreation node vector data with spatial location information.

3. The method for simulating and selecting eco-recreational routes based on natural resource data according to claim 1, characterized in that, Based on social media data, the frequency of co-occurrence of each pair of recreational nodes in the same user's itinerary or published content is statistically analyzed, generating a data table of recreational node pairs and their corresponding co-occurrence frequencies. The co-occurrence frequencies are used as the recreational activity connection frequency observation values ​​for the corresponding recreational node pairs, specifically including: S21: Construct query conditions from at least one social media platform using the name of the target area and keywords related to recreational activities, collect post data published within a preset time period through a web crawler, and store the collection results as a structured text file; S22: Clean the post data in the text file and filter out records whose geographical location belongs to the target area and whose text content contains one or more recreation node names as valid recreation records; S23: Construct a co-occurrence matrix of recreation nodes. For each pair of recreation nodes, count the frequency of the two recreation node names appearing simultaneously in the same post or the same user's trip in all valid recreation records. Use this as the recreation activity connection frequency observation value of the recreation node pair and organize it into a data table with recreation node pairs as the index and recreation activity connection frequency observation values ​​as data items.

4. The method for simulating and selecting eco-recreational routes based on natural resource data according to claim 1, characterized in that, The multiple linear regression model is constructed based on the gravity model, and its expression is as follows: In the above formula, The frequency of recreational activity connections between recreational node i and recreational node j is represented by C, S, and L, which represent relevant indicators in terms of resource suitability, ecological landscape visibility, and transportation connectivity convenience, respectively. For elevation indicators, For slope index, As an aspect indicator, The vegetation coverage rate is an indicator. This is an indicator of the visibility of mountains. As an indicator of water body visibility; It represents the minimum actual path length between any pair of resting nodes.

5. The method for simulating and selecting eco-recreational routes based on natural resource data according to claim 1, characterized in that, The calculation of independent variable index values ​​for all recreational node pairs based on spatial data of all recreational nodes in the target area specifically includes: S41: Acquire and integrate spatial data of all recreational nodes in the target area, the spatial data including: vector boundary data of recreational nodes, digital elevation model data, land use / cover data, point of interest data and road network data; S42: For each pair of recreational nodes, calculate the values ​​of three types of independent variable indicators based on the spatial data. The three types of independent variable indicators include the resource background suitability indicator, the ecological landscape visibility indicator, and the transportation connection convenience indicator. S43: Standardize the calculated values ​​of the three types of independent variables to form a dataset for fitting the multiple linear regression model.

6. The method for simulating and selecting ecological recreational routes based on natural resource data according to claim 5, characterized in that, For each pair of recreational nodes, three types of independent variable index values ​​are calculated based on the spatial data, specifically including: S421: The resource background suitability index value is one or more of the following, calculated based on the vector boundary data, land use / cover data and point of interest data: area, internal facility density, and natural landscape area ratio of each recreation node pair; S422: The ecological landscape visibility index value is calculated based on the digital elevation model data, which calculates the visual visibility probability or cumulative visible area between recreation node pairs. S423: The traffic connectivity convenience index is calculated based on the road network data, using the shortest path algorithm to determine the network distance, travel time cost, or number of topological connections between two points.

7. The method for simulating and selecting eco-recreational routes based on natural resource data according to claim 1, characterized in that, For the region to be predicted, based on the spatial data of all recreational nodes in the region, the independent variable index values ​​of all recreational node pairs in the region are calculated and substituted into the trained recreational activity connection frequency prediction model to predict the observed recreational activity connection frequency values ​​between all recreational node pairs in the region, specifically including: S61: Determine the boundary range of the area to be predicted; S62: Obtain a list of standardized names for all recreational nodes within the stated boundary area through publicly available tourism information service systems, geographic information databases, or field surveys; S63: Using geocoding services, convert the list of standardized names into corresponding spatial coordinate data; S64: Perform coordinate correction and spatial registration processing on the spatial coordinate data to generate spatial vector data of recreational nodes in the area to be predicted; S65: Calculate the independent variable index value corresponding to each pair of recreational nodes in the area to be predicted based on the spatial vector data; S66: Substitute the independent variable index value corresponding to each pair of recreational nodes in the area to be predicted into the trained recreational activity connection frequency prediction model, and simulate and calculate the recreational activity connection frequency prediction value between each pair of recreational nodes.

8. The method for simulating and selecting eco-recreational routes based on natural resource data according to claim 1, characterized in that, The process involves selecting all recreational node pairs with activity frequency observations higher than the average value based on the observed activity frequency observations among all recreational node pairs, and generating a recreational path selection scheme for the area to be predicted. Specifically, this includes: S71: Based on the predicted frequency of recreational activity connections between all pairs of recreational nodes obtained from simulation prediction, calculate the arithmetic mean of all predicted frequencies of recreational activity connections, and use it as the screening threshold; S72: Select recreational node pairs from all recreational node pairs whose predicted recreational activity connection frequency is higher than the selection threshold to form a set of node pairs with high connection strength. S73: Using all the recreational node pairs in the set of high-connection-strength node pairs as basic connection units, perform network connection and topology merging among the recreational nodes in the region to be predicted; S74: Based on the merged network structure, extract the main activity connection paths and generate a recreational path selection scheme for the area to be predicted.

9. An eco-recreation path simulation and route selection system based on natural resource data, characterized in that, include: The data acquisition module is used to determine the target area and obtain the names and spatial coordinates of all recreational nodes within the target area. The recreational nodes can be any one of urban parks, green squares, natural landscapes, cultural heritage sites, museums, and commercial and leisure areas. The module is also used to calculate the frequency of co-occurrence of each pair of recreational nodes in the same user's itinerary or published content based on social media data, generate a data table of recreational node pairs and their corresponding co-occurrence frequencies, and use the co-occurrence frequency as the recreational activity connection frequency observation value of the corresponding recreational node pair. The model building module is used to build a multiple linear regression model, wherein the multiple linear regression model uses the observed frequency of recreational activity connections as the dependent variable and the combination of spatial data indicators that characterize the resource background suitability, ecological landscape visibility and transportation connectivity of the corresponding recreational node pair as independent variables. The indicator calculation module is used to calculate the independent variable indicator values ​​of all recreational node pairs based on the spatial data of all recreational nodes in the target area. The model fitting module is used to fit the observed values ​​of recreational activity connection frequency with the corresponding independent variable index values, solve the undetermined coefficients of the multiple linear regression model, and obtain the trained recreational activity connection frequency prediction model. The prediction application module is used to calculate the independent variable index values ​​of all recreational node pairs in the area to be predicted based on the spatial data of all recreational nodes in the area to be predicted, and substitute them into the trained recreational activity connection frequency prediction model to predict the observed values ​​of recreational activity connection frequency between all recreational node pairs in the area to be predicted. The route selection scheme generation module is used to filter out all recreational node pairs whose recreational activity connection frequency observation values ​​are higher than the average value based on the observed values ​​of recreational activity connection frequency among all recreational node pairs, and generate a recreational path selection scheme for the recreational node pairs in the area to be predicted.

10. The ecological recreation path simulation and route selection system based on natural resource data according to claim 9, characterized in that, The data acquisition module performs the following steps: Obtain the administrative division vector boundary data of the target area; Collect a list of names of all recreational nodes within the target area from tourism information platforms and / or local data; Geographically analyze the recreation nodes in the name list to obtain their geographic coordinates; The geographic coordinates are corrected and spatially positioned to generate vector data of recreational nodes with spatial location information.

Images