Scenic spot route recommendation method and system

Abstract

The invention discloses a scenic spot route recommendation method and system, and relates to the technical field of route planning, and the method comprises the steps of tourism basic data acquisition, tourism city data acquisition and analysis, initial route acquisition, initial scenic spot route planning, dynamic adjustment and terminal display. And triggering an interaction pop-up window to prompt a user whether to change a city, thereby solving the limitation problem in the current tourist attraction route recommendation development feasibility analysis process, generating an initial route based on road network modeling and an improved force-oriented algorithm, and optimizing a scenic spot sequence in combination with a dynamic planning algorithm. According to the method, hotel positions are matched according to morning or night market distribution, hotel density and user budget, scenic spot lines are dynamically allocated to different time periods, after the lines are interactively adjusted by users, a recommendation model is updated through Q-learning reinforcement learning, and finally an optimization result is displayed on a map terminal, so that the comprehensiveness of the final scenic spot lines is guaranteed.

Description

Technical Field

[0001] The present application relates to the technical field of route planning, and in particular to a method and system for recommending routes to tourist attractions. Background Art

[0002] With the advent of the big data era, tourism planning is no longer confined to paper formats. In the competition among similar tourism planning methods, analyzing user demand data and travel data is the key and core of tourism planning.

[0003] The existing technology, such as the invention application patent with announcement number: CN118536688A, discloses an artificial intelligence-based tourism route planning method and system, which belongs to the field of route planning. The method includes: establishing a user data set; reading input data, matching projects in the target scenic area with user characteristics, planning routes for the projects, and generating accommodation selection plans corresponding to the route planning, eliminating projects based on the route planning results, project matching values and input data, and reconstructing the route planning results; feeding back the reconstructed route planning results to the user to complete the tourism route planning management.

[0004] In response to the above scheme, the inventors of this application found that the above technology has at least the following technical problems: 1. There is currently a lack of quantified city recommendation degrees, and a lack of triggering interactive pop-up windows to prompt users whether to change cities. It is unable to solve the limitations of the current feasibility analysis process of tourist attraction route recommendation development, and does not combine real-time weather and crowd forecast data to improve evaluation accuracy. There is a lack of improved force-directed algorithms to introduce business hours overlap coefficients and user feedback data, and it cannot break through the limitations of traditional models that only consider spatial distances; at the same time, there is a lack of cross-routes in the initial attraction routes, which reduces traffic intersections during travel and saves time and energy.

[0005] 2. Currently, hotels are not planned based on the location and price matching of morning or night markets, which cannot enhance itinerary convenience and budget adaptability. There is a lack of a dynamic allocation strategy, and the remaining attraction routes cannot be allocated based on weather and distance, resulting in an inability to improve itinerary flexibility and rationality. There is also no learning feedback loop, and users adjust their behavior through Q-learning to update the model, which cannot continuously optimize recommendation results and satisfaction. Summary of the Invention

[0006] In view of the above-mentioned technical deficiencies, the purpose of this application is to provide a method and system for recommending tourist attraction routes.

[0007] In order to solve the above technical problems, the present application adopts the following technical solutions: In the first aspect, the present application provides a method for recommending tourist attraction routes, which includes the following steps: Step 1, obtaining basic tourism data: obtaining the user's basic tourism data, wherein the basic tourism data includes tourist cities, attraction lists and itinerary time periods.

[0008] Step 2: Acquisition and analysis of tourist city data: Based on the user's basic tourism data, obtain the tourist data of the tourist city, the weather data of the tourist city during the travel period, and the passenger flow data, and then analyze and derive the tourist score of the tourist city, and then analyze and derive the recommendation evaluation coefficient of the tourist city, triggering an interactive pop-up window to prompt the user to choose whether to change the city.

[0009] Step 3: Initial route acquisition: Obtain the target city's travel data, perform road network modeling, and then use the improved force-directed algorithm to generate the initial routes to each attraction.

[0010] Step 4: Initial tourist attraction route planning: Extract the recommended accommodation routes from each initial tourist attraction route, and then plan the location of each hotel in each recommended accommodation route, so as to analyze the attraction route arrangement for each route time period within the travel time period and obtain the initial tourist attraction route.

[0011] Step 5: Dynamic adjustment: Send the initial tourist attraction route to the user end, and obtain the initial tourist attraction route adjusted by the user, thereby deriving the tourist attraction route.

[0012] Step 6: Display terminal: Display the tourist attraction route on the map of the corresponding tourist city.

[0013] Preferably, the analysis to obtain the recommendation evaluation coefficient of a tourist city includes: calculating the coefficient according to the formula The analysis yields the recommendation evaluation coefficient β of a tourist city, where C and C 1 They are respectively represented as the tourism score of the tourist city and the standard score of the tourist city, R and R 1 They represent the passenger flow data and the standard value of passenger flow data of the tourist city during the travel period, τ 1 They are weight factors corresponding to the user's historical itinerary satisfaction scores of tourist cities, τ 2 It is represented as the weight factor corresponding to the passenger flow data of the tourist city during the travel period, and γ is represented as the weather data of the tourist city during the travel period.

[0014] Preferably, the interactive pop-up window is triggered, and the user chooses whether to change the city: the recommendation evaluation coefficient of the tourist city is compared with the recommendation evaluation coefficient threshold of the tourist city. When the recommendation evaluation coefficient of the tourist city is less than or equal to the recommendation evaluation coefficient threshold of the tourist city, the interactive pop-up window is triggered to prompt the user that the recommendation evaluation coefficient of the tourist city is too low and whether to change the tourist city; when the user chooses yes, step one is performed; when the user chooses no, the tourist city selected by the user is recorded as the target city, and step three is performed.

[0015] Preferably, the analysis obtains the arrangement of scenic spot routes for each route time period within the travel time period, including: planning each initial scenic spot route containing the morning market in the morning route time period within the travel time period, planning each initial scenic spot route containing the night market in the evening route time period within the travel time period, and allocating the remaining initial scenic spot routes according to the distance from the initial scenic spot routes of each route time period within the arranged travel time period and the real-time weather, thereby obtaining the initial tourist attraction routes.

[0016] In a second aspect, the present application provides a tourist attraction route recommendation system, including: a tourist basic data acquisition module, used to obtain the user's tourist basic data, wherein the tourist basic data includes tourist cities, attraction lists and travel time periods.

[0017] The tourist city data acquisition and analysis module is used to obtain the tourist data of the tourist city, the weather data and the passenger flow data of the tourist city during the travel period based on the user's basic tourism data, and then analyze the tourist score of the tourist city, and thus analyze the recommendation evaluation coefficient of the tourist city, triggering an interactive pop-up window to prompt the user to choose whether to change the city.

[0018] The initial route acquisition module is used to obtain the travel data of the target city, perform road network modeling, and then use the improved force-directed algorithm to generate the initial scenic spot routes.

[0019] The initial tourist attraction route planning module is used to extract the recommended accommodation attraction routes from each initial attraction route, and then plan the location of each hotel in each recommended accommodation attraction route, so as to analyze the attraction route arrangement for each route time period within the travel time period and obtain the initial tourist attraction route.

[0020] The dynamic adjustment module is used to send the initial tourist attraction route to the user terminal and obtain the initial tourist attraction route adjusted by the user, thereby deriving the tourist attraction route.

[0021] The display terminal is used to display the tourist attraction routes on the map of the corresponding tourist city.

[0022] The beneficial effects of the present application are: 1. A method and system for recommending tourist attraction routes provided by the present application solves the limitations of the current feasibility analysis process of recommending tourist attraction routes by calculating the recommendation coefficient and triggering an interactive pop-up window to prompt the user whether to change cities. The system generates initial routes based on road network modeling and an improved force-directed algorithm, optimizes the sequence of attractions by combining a dynamic programming algorithm, matches hotel locations according to the distribution of morning or night markets, hotel density, and user budget, and dynamically allocates attraction routes to different time periods. After the user interactively adjusts the route, the recommendation model is updated through Q-learning reinforcement learning, and the optimization results are finally displayed on the map terminal, ensuring the comprehensiveness of the final tourist attraction route.

[0023] 2. This application uses a multi-dimensional weight matrix, including the star rating of scenic spots, negative public opinion and historical satisfaction, to quantify the city recommendation degree, and combines real-time weather and crowd flow forecast data to improve the accuracy of the evaluation. The improved force-directed algorithm introduces the business hours overlap coefficient and user feedback data, breaking through the limitation of traditional models that only consider spatial distance; the real-time congestion index is embedded in the road network modeling, dynamically amplifying the travel time weight and giving priority to unobstructed paths; the pop-up interaction is triggered by the recommendation threshold, balancing automated recommendations with user autonomy, and at the same time, there are no intersecting routes in the initial scenic spot routes, reducing traffic intersections during travel and saving time and energy.

[0024] 3. This application plans hotels based on the location and price matching of morning or night markets to enhance itinerary convenience and budget adaptability; a dynamic allocation strategy allocates remaining attraction routes based on weather and distance to improve itinerary flexibility and rationality; a learning feedback loop is used, and users adjust their behavior through Q-learning to update the model and continuously optimize recommendation results and satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0026] Figure 1 The figure is a flowchart of the steps for implementing the application method.

[0027] Figure 2 This is a schematic diagram of the system structure connection for this application. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0029] See also Figure 1 As shown, the present application provides a method for recommending tourist attraction routes in the first aspect, including: Step 1, obtaining basic tourism data: obtaining the user's basic tourism data, wherein the basic tourism data includes tourist cities, attraction lists and itinerary time periods.

[0030] It should be noted that the user's travel cities, attractions list and travel time period are filled in by the user on the user side and then uploaded to the server.

[0031] Step 2: Acquisition and analysis of tourist city data: Based on the user's basic tourism data, obtain the tourist data of the tourist city, the weather data of the tourist city during the travel period, and the passenger flow data, and then analyze and derive the tourist score of the tourist city, and then analyze and derive the recommendation evaluation coefficient of the tourist city, triggering an interactive pop-up window to prompt the user to choose whether to change the city.

[0032] In a specific example, the tourism data of the tourist city includes the star rating of each attraction, the frequency ratio of negative keywords in the tourist city, and the user's historical travel satisfaction score.

[0033] It should be noted that the star ratings of various attractions are obtained from the public database of the Tourism Bureau, and the frequency ratio of negative keywords in tourist cities and the user's historical travel satisfaction scores are extracted from online media public opinion.

[0034] In a specific example, the analysis to obtain the tourism score of a tourist city includes: calculating the tourism score according to the formula The analysis yields the tourism score C of the tourist city, where i = 1, 2, ..., n, i represents the number of each scenic spot, n represents the total number of scenic spots, and n is a natural integer greater than or equal to 1; W represents the multi-dimensional weight matrix Each row corresponds to a scenic spot. S represents the star rating of the scenic spot, N represents the frequency ratio of negative keywords in the tourist city, U represents the user's historical itinerary satisfaction score of the tourist city, and w represents the star rating of the tourist city. S 、w N and w U They are respectively expressed as the weight factor corresponding to the star rating of the scenic spot, the weight factor corresponding to the frequency ratio of negative keywords in the tourist city, and the weight factor corresponding to the user's historical itinerary satisfaction score of the tourist city.

[0035] It should be noted that w S =0.4, w N 3, w U =0.3.

[0036] It should be noted that the server obtains real-time weather data by calling the Meteorological Bureau.

[0037] It should be noted that the specific acquisition process of the passenger flow data of the tourist city during the travel time period is as follows: at the same time, the server is connected to the municipal passenger flow monitoring system, the gate passage data of each scenic spot in the previous N days is extracted, and the ARIMA (P, D, Q) model is used for time series measurement to generate a passenger flow prediction curve, where the prediction model input parameters include the tourist city, travel time period and the gate passage data of each scenic spot in the previous N days, and then the passenger flow data of the tourist city during the travel time period is output.

[0038] In a specific example, the analysis to obtain the recommendation evaluation coefficient of a tourist city includes: according to the calculation formula The analysis yields the recommendation evaluation coefficient β of a tourist city, where C and C 1 They are respectively represented as the tourism score of the tourist city and the standard score of the tourist city, R and R 1 They represent the passenger flow data and the standard value of passenger flow data of the tourist city during the travel period, τ 1 They are weight factors corresponding to the user's historical itinerary satisfaction scores of tourist cities, τ 2 It is represented as the weight factor corresponding to the passenger flow data of the tourist city during the travel period, and γ is represented as the weather data of the tourist city during the travel period.

[0039] It should be noted that 0<τ 1 <1, 0<τ 2 <1,τ 1 +τ 2 =2.

[0040] In a specific example, the interactive pop-up window is triggered, and the user chooses whether to change the city: the recommendation evaluation coefficient of the tourist city is compared with the recommendation evaluation coefficient threshold of the tourist city. When the recommendation evaluation coefficient of the tourist city is less than or equal to the recommendation evaluation coefficient threshold of the tourist city, the interactive pop-up window is triggered to prompt the user that the recommendation evaluation coefficient of the tourist city is too low and whether to change the tourist city; when the user chooses yes, step one is performed; when the user chooses no, the tourist city selected by the user is recorded as the target city, and step three is performed.

[0041] Step 3: Initial route acquisition: Obtain the target city's travel data, perform road network modeling, and then use the improved force-directed algorithm to generate the initial routes to each attraction.

[0042] It should be noted that the road network modeling process is as follows: First, construct a weighted graph G = (V, E) where V is the vertex representing the scenic spot, E is the edge representing the path connecting the vertices, and W ij To represent the weight of the path, i and j are both represented as scenic spots. According to the calculation formula W ij =TR ij *(1+CO) gets the path weight W ij , where TR ij It is represented as the basic travel time from attraction i to j, calculated based on distance and average speed, such as the length of time spent walking or driving. CO is represented as the real-time congestion index, and CO takes values of 0 or 1, where 0 indicates smooth traffic and 1 indicates severe congestion. (1+CO) is represented as the congestion impact factor, which dynamically amplifies the basic travel time. For example, when CO = 0.3, the actual weight is equal to the basic travel time multiplied by 1.3.

[0043] It should be noted that the force-directed algorithm is an existing technology, which refers to calculating the combined force of attraction and repulsion through calculation of each node, and then moving the position of the node based on this combined force.

[0044] In a specific example, the target city's travel data includes the target city's traffic data, user feedback data, various tourist attractions, coordinates of each tourist attraction, business hours, recommended travel duration, and distances between each tourist attraction, wherein the traffic data includes road type, travel time, and real-time congestion index.

[0045] It should be noted that user feedback data includes the user's stay duration and route preference.

[0046] In a specific example, the improved force-directed algorithm includes: A1, repulsive force calculation formula The repulsive force between all non-directly connected scenic spots is obtained, where K is the repulsive force constant and 0.01 <K<0.1,d ij It is represented as the normalized distance between scenic spots i and j, and ΔFK is represented as user feedback data.

[0047] It should be noted that according to the calculation formula The normalized distance d between scenic spots i and j is obtained by analysis ij , and the value range after normalization is [0,1].

[0048] A2. Gravity calculation formula Obtain the gravitational force F between each scenic spot ij , where b ij It is represented as the business hours overlap coefficient, ρ is represented as the weight factor corresponding to the normalized distance, and σ is represented as the weight factor corresponding to the business hours overlap coefficient.

[0049] When necessary, ρ = 0.6, σ = 0.4.

[0050] It should be noted that according to The analysis yields the business hours overlap coefficient b ij , and when b ij When the value decreases, the temporal connection between attractions becomes worse. For example, when a certain attraction is temporarily closed, b ij =0, the scenic spot will automatically disconnect from the connection.

[0051] A3. Complete displacement adjustment.

[0052] It should be noted that the displacement adjustment process is as follows: Displacement adjustment formula Among them, M i =1 makes the quality of each scenic spot the same, Δt represents the time step; 100≤Δt≤500, the convergence speed is controlled by the number of iterations, total energy = node kinetic energy + potential energy, and the iteration stops when the total energy is less than the total energy threshold.

[0053] A4. Complete intersection detection and elimination, and finally output a non-intersection connection diagram on the map.

[0054] It should be noted that the process of completing intersection detection and elimination is to complete the scan line algorithm. First, the edges are sorted and all edges are arranged in ascending order according to the y-coordinate of the endpoints; then the scan line traversal is performed, and the vertical scan line is moved from left to right to maintain the active edge table; then the intersection detection is performed, and the geometric intersection calculation is performed on the newly added edges and the active edges, and the crossing edge pairs are marked; to complete the intersection elimination, the edges with small gravitational weights are first adjusted, and the intersections are eliminated by local node displacement or edge rewiring.

[0055] It should be noted that the active edge table represents the edges that currently intersect the scan line; the edges with small gravitational weights such as b ij <0.3 edge.

[0056] In a specific example, generating each initial scenic spot route includes: formulating the sequence of each scenic spot using a dynamic programming algorithm based on traffic network topology data, the business hours and recommended visiting duration of each scenic spot, thereby generating each initial scenic spot route.

[0057] It should be noted that the traffic network topology data is obtained based on the map service API.

[0058] It should be noted that the process of using the dynamic programming algorithm to formulate the order of each scenic spot is as follows: the maximum number of scenic spots i visited within the time period t is recorded as dp[i][t], and according to the calculation formula Perform state transfer, where WT represents the walking time between adjacent attractions and PT represents the recommended playing time; the constraints are WT ≤ 15 minutes and AT ∈ OT i ,CT i -PT i ], where AT represents the arrival time, OT and CT represent the opening and closing times of the scenic spot respectively.

[0059] It should be noted that t≤5h, which limits the time period t to 5 hours in half a day, thereby limiting each initial scenic spot route to within 5 hours.

[0060] This application uses a multi-dimensional weight matrix, including attraction star ratings, negative public opinion, and historical satisfaction, to quantify the city's recommendation level, and combines real-time weather and crowd flow forecast data to improve evaluation accuracy. The improved force-directed algorithm introduces business hours overlap coefficients and user feedback data, breaking through the limitations of traditional models that only consider spatial distance; a real-time congestion index is embedded in road network modeling, dynamically amplifying the weight of travel time and giving priority to unobstructed paths; pop-up window interactions are triggered by recommendation thresholds, balancing automated recommendations with user autonomy, while ensuring that there are no intersecting routes in the initial attraction routes, reducing traffic intersections during travel and saving time and energy.

[0061] Step 4: Initial tourist attraction route planning: Extract the recommended accommodation routes from each initial tourist attraction route, and then plan the location of each hotel in each recommended accommodation route, so as to analyze the attraction route arrangement for each route time period within the travel time period and obtain the initial tourist attraction route.

[0062] It should be noted that the specific process of extracting the recommended accommodation route from each initial route is as follows: the initial route containing a night market or a morning market in each initial route is recorded as the recommended accommodation route.

[0063] It should be noted that, according to the calculation formula Plan the location of each hotel, including JD n Expressed as the hotel distribution density next to each morning market or night market, calculate the number of hotels within a 500-meter radius around the location of the morning market or night market, JG n Indicates price matching, matching the user budget with the hotel price, and according to the calculation formula Analysis shows price matching degree JG n .

[0064] In a specific example, the analysis obtains the arrangement of scenic spot routes for each route time period within the travel time period, including: planning each initial scenic spot route containing a morning market in the morning route time period within the travel time period, planning each initial scenic spot route containing a night market in the evening route time period within the travel time period, and allocating the remaining initial scenic spot routes according to the distance from the initial scenic spot routes of each route time period within the arranged travel time period and the real-time weather, thereby obtaining the initial tourist attraction routes.

[0065] It should be noted that the specific implementation process of the scenic spot route arrangement for each route time period within the travel time period is as follows: First, the time period is locked Then dynamically allocate the remaining initial scenic spots according to the calculation formula The spatiotemporal weight SC is obtained by analysis o , where dd o It is represented as the initial scenic spot route o with the minimum distance to the assigned initial scenic spot route, TQ is represented as the weather impact factor, δ and ε are represented as the weight factor corresponding to the initial scenic spot route o with the minimum distance and the weight factor corresponding to the weather impact factor, respectively; where TQ = 1, δ = 0.6 and ε = 0.4 on sunny days, and TQ = 0.3, δ = 0.4 and ε = 0.6 on cloudy days; and the travel time between each initial scenic spot route is recorded as TT ol , and according to the calculation formula Complete the allocation of the remaining initial scenic spot routes.

[0066] It should be noted that, for example, when the weather is not sunny, the initial scenic spot routes in which indoor activities account for more than half will be arranged to be on a sunny day.

[0067] It should be noted that when the hotel is arranged to be near the night market and there is an initial scenic spot route with the shortest distance around it, the initial scenic spot route with the shortest distance will be arranged to the next morning.

[0068] It should be noted that when some of the initial scenic spots are far away, users are prompted to select hotels in sections, such as choosing hotels near different initial scenic spots for accommodation in the first half and the second half of the trip respectively.

[0069] Step 5: Dynamic adjustment: Send the initial tourist attraction route to the user end, and obtain the initial tourist attraction route adjusted by the user, thereby deriving the tourist attraction route.

[0070] It should be noted that the initial tourist attraction route adjusted by the user is obtained, and the tourist attraction route is derived from it. The specific process is as follows: the user behavior record is stored in the four-tuple original route, operation type, new route, and satisfaction is updated using Q-learning. The calculation formula is QQ(s,a)←QQ(s,a)+η[RR+γmaxQQ(s′,a′)-QQ(s,a)], where the state s represents the current route feature vector (such as the number of attractions, average distance or weather score), the action a represents operations such as adding or deleting attractions, changing time periods, etc., and the reward RR represents the user adoption result (1 / 0) + satisfaction score (0-5).

[0071] Step 6: Display terminal: Display the tourist attraction route on the map of the corresponding tourist city.

[0072] This application plans hotels based on the location and price matching of the morning or night market, enhancing itinerary convenience and budget adaptability; a dynamic allocation strategy allocates remaining attraction routes based on weather and distance to improve itinerary flexibility and rationality; a learning feedback loop is used, and users adjust their behavior through Q-learning to update the model and continuously optimize recommendation results and satisfaction.

[0073] See also Figure 2 As shown, the present application provides a tourist attraction route recommendation system in the second aspect, including: a tourist basic data acquisition module for acquiring the user's tourist basic data, wherein the tourist basic data includes tourist cities, attraction lists and travel time periods.

[0074] The tourist city data acquisition and analysis module is used to obtain the tourist data of the tourist city, the weather data and the passenger flow data of the tourist city during the travel period based on the user's basic tourism data, and then analyze the tourist score of the tourist city, and thus analyze the recommendation evaluation coefficient of the tourist city, triggering an interactive pop-up window to prompt the user to choose whether to change the city.

[0075] The initial route acquisition module is used to obtain the travel data of the target city, perform road network modeling, and then use the improved force-directed algorithm to generate the initial scenic spot routes.

[0076] The initial tourist attraction route planning module is used to extract the recommended accommodation attraction routes from each initial attraction route, and then plan the location of each hotel in each recommended accommodation attraction route, so as to analyze the attraction route arrangement for each route time period within the travel time period and obtain the initial tourist attraction route.

[0077] The dynamic adjustment module is used to send the initial tourist attraction route to the user terminal and obtain the initial tourist attraction route adjusted by the user, thereby deriving the tourist attraction route.

[0078] The display terminal is used to display the tourist attraction routes on the map of the corresponding tourist city.

[0079] The present application provides a method and system for recommending tourist attraction routes. By calculating the recommendation coefficient and triggering an interactive pop-up window to prompt the user whether to change cities, the system solves the limitations of the current feasibility analysis process for recommending tourist attraction routes. The system generates an initial route based on road network modeling and an improved force-directed algorithm, optimizes the sequence of attractions with a dynamic programming algorithm, matches hotel locations according to the distribution of morning or night markets, hotel density, and user budget, and dynamically allocates attraction routes to different time periods. After the user interactively adjusts the route, the system updates the recommendation model through Q-learning reinforcement learning, and finally displays the optimization results on the map terminal, ensuring the comprehensiveness of the final tourist attraction route.

[0080] The above content is merely an example and explanation of the concept of the present application. Technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the scope of protection of the present application.

Claims

1. A method for recommending tourist attraction routes, characterized in that: include: Step 1: Acquire basic travel data: Acquire the user's basic travel data, which includes travel cities, scenic spots, and itinerary time periods. Step 2: Acquisition and analysis of tourist city data: Based on the user's basic travel data, obtain the tourist data of the tourist city, the weather data of the tourist city during the travel period, and the passenger flow data of the tourist city. Then, analyze and derive the tourist score of the tourist city, and then analyze and derive the recommendation evaluation coefficient of the tourist city. Then, trigger an interactive pop-up window to prompt the user to choose whether to change the city. Step 3: Initial route acquisition: Obtain the target city's travel data, perform road network modeling, and then use the improved force-directed algorithm to generate the initial routes to each attraction. Step 4: Initial tourist attraction route planning: Extract recommended accommodation routes from each initial tourist attraction route, and then plan the location of each hotel in each recommended accommodation route, so as to analyze and obtain the route arrangement of tourist attractions for each route time period within the travel time period, and obtain the initial tourist attraction route; Step 5: Dynamic adjustment: Send the initial tourist attraction route to the user terminal, and obtain the initial tourist attraction route adjusted by the user, thereby deriving the tourist attraction route; Step 6: Display terminal: Display the tourist attraction route on the map of the corresponding tourist city.

2. A method for recommending tourist attractions routes according to claim 1, characterized in that: The tourism data of the tourist city includes the star rating of each attraction, the frequency ratio of negative keywords in the tourist city and the user's historical travel satisfaction score.

3. A method for recommending tourist attractions routes according to claim 2, characterized in that: The analysis yields a tourism score for a tourist city, including: According to the calculation formula The analysis yields the tourism score C of the tourist city, where i = 1, 2, ..., n, i represents the number of each scenic spot, n represents the total number of scenic spots, and n is a natural integer greater than or equal to 1; W represents the multi-dimensional weight matrix Each row corresponds to a scenic spot. S represents the star rating of the scenic spot, N represents the frequency ratio of negative keywords in the tourist city, U represents the user's historical itinerary satisfaction score of the tourist city, and w represents the star rating of the tourist city. S 、w N and w U They are respectively expressed as the weight factor corresponding to the star rating of the scenic spot, the weight factor corresponding to the frequency ratio of negative keywords in the tourist city, and the weight factor corresponding to the user's historical itinerary satisfaction score of the tourist city.

4. A method for recommending tourist attractions routes according to claim 3, characterized in that: The analysis yields a recommendation evaluation coefficient for a tourist city, including: According to the calculation formula The analysis yields the recommendation evaluation coefficient β of a tourist city, where C and C 1 They are respectively represented as the tourism score of the tourist city and the standard score of the tourist city, R and R 1 They represent the passenger flow data and the standard value of passenger flow data of the tourist city during the travel period, τ 1 They are weight factors corresponding to the user's historical itinerary satisfaction scores of tourist cities, τ 2 It is represented as the weight factor corresponding to the passenger flow data of the tourist city during the travel period, and γ is represented as the weather data of the tourist city during the travel period.

5. A method for recommending tourist attractions routes according to claim 4, characterized in that: The interactive pop-up window is triggered, and the user chooses whether to change the city: Compare the recommendation evaluation coefficient of a tourist city with the recommendation evaluation coefficient threshold of a tourist city. When the recommendation evaluation coefficient of a tourist city is less than or equal to the recommendation evaluation coefficient threshold of a tourist city, trigger an interactive pop-up window to prompt the user that the recommendation evaluation coefficient of the tourist city is too low and whether to change the tourist city. When the user selects yes, proceed to step 1; When the user selects "no", the tourist city selected by the user is recorded as the target city, and step three is performed.

6. A method for recommending tourist attraction routes according to claim 5, characterized in that: The target city's travel data includes the target city's traffic data, user feedback data, various tourist attractions, coordinates of each attraction, business hours, recommended travel duration, and distances between each attraction, wherein the traffic data includes road type, travel time, and real-time congestion index.

7. A method for recommending tourist attraction routes according to claim 6, characterized in that: The improved force-directed algorithm includes: A1. Repulsive force calculation formula The repulsive force between all non-directly connected scenic spots is obtained, where K is the repulsive force constant and 0.01 <K<0.1,d ij It is represented as the normalized distance between scenic spots i and j, and ΔFK is represented as user feedback data; A2. Gravity calculation formula Obtain the gravitational force F between each scenic spot ij , where b ij It is represented as the business hours overlap coefficient, ρ is represented as the weight factor corresponding to the normalized distance, and σ is represented as the weight factor corresponding to the business hours overlap coefficient; A3. Complete displacement adjustment; A4. Complete intersection detection and elimination, and finally output a non-intersection connection diagram on the map.

8. A method for recommending tourist attraction routes according to claim 7, characterized in that: The generating of each initial scenic spot route includes: According to the traffic network topology data, based on the business hours and recommended visiting duration of each tourist attraction, a dynamic programming algorithm is used to formulate the order of each attraction, thereby generating the initial attraction routes.

9. A method for recommending tourist attraction routes according to claim 8, characterized in that: The analysis results in the arrangement of scenic spots for each route time period within the travel time period, including: The initial scenic spot routes including the morning market are planned in the morning route time period within the itinerary time period, the initial scenic spot routes including the night market are planned in the evening route time period within the itinerary time period, and the remaining initial scenic spot routes are allocated according to the distance from the initial scenic spot routes of each route time period within the arranged itinerary time period and the real-time weather, thereby obtaining the initial tourist attraction routes.

10. A tourist attraction route recommendation system according to any one of claims 1 to 9, characterized in that: include: The basic tourism data acquisition module is used to obtain the user's basic tourism data, where the basic tourism data includes tourist cities, scenic spots list and itinerary time period; The tourist city data acquisition and analysis module is used to obtain tourist data of tourist cities, weather data and passenger flow data of tourist cities during the travel period based on the user's basic tourist data, and then analyze and derive the tourist rating of the tourist cities, thereby analyzing and deriving the recommendation evaluation coefficient of the tourist cities, triggering an interactive pop-up window to prompt the user to choose whether to change cities; The initial route acquisition module is used to obtain the target city's travel data, perform road network modeling, and then use the improved force-directed algorithm to generate the initial routes to various scenic spots; The initial tourist attraction route planning module is used to extract the recommended accommodation route from each initial attraction route, and then plan the location of each hotel in each recommended accommodation route, so as to analyze the route arrangement of the attractions in each time period within the travel time period and obtain the initial tourist attraction route; A dynamic adjustment module is used to send the initial tourist attraction route to the user terminal and obtain the initial tourist attraction route adjusted by the user, thereby deriving the tourist attraction route; The display terminal is used to display the tourist attraction routes on the map of the corresponding tourist city.

Images