Smart tourism online comprehensive management system

Abstract

The invention provides a smart tourism online comprehensive management system, belongs to the technical field of tourism data management, and can make tourism resource assessment more accurate. Static attributes and dynamic indexes are integrated by means of a regional industry knowledge graph, a unified data support network is constructed, a plurality of Pareto optimal routes are generated in combination with a multi-objective optimization algorithm, and personalized requirements of tourists on time, cost, interest matching degree and experience credibility are met; route adjustment can be carried out according to tourist real-time states, resource dynamic states and macroscopic passenger flow early warning automatic or response intention changes in the journey, and journey dynamic optimization is achieved; and meanwhile, passenger flow can be predicted through a graph neural network model, a risk early warning and multi-subject collaborative scheduling scheme can be generated, regional tourism collaborative management is assisted, the problems that a traditional system is large in recommendation deviation, cannot dynamically adapt and is insufficient in collaborative scheduling are effectively solved, and tourism experience of tourists and regional tourism service quality are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tourism data management, and particularly relates to a smart tourism online comprehensive management system. BACKGROUND

[0002] With the popularization of Internet technology, the smart tourism online management system has become an important part of the tourism industry. The tourism online comprehensive management system mainly provides basic query and reservation services for users by integrating network ticketing, hotel booking, and scenic spot introduction information. However, with the increasing demand for personalized and intelligent tourism experience of tourists, the existing tourism online comprehensive management system's tourism recommendation service is mostly based on static scenic spot information or single-dimensional user evaluation for route recommendation, and lacks dynamic mining and analysis of real behavior data of tourists, resulting in a large deviation between the recommended results and the actual travel experience.

[0003] At the same time, the real-time state of tourism resources, passenger flow distribution, and changes in tourist preferences cannot be effectively integrated, and dynamic adjustment of the itinerary and regional collaborative scheduling cannot be achieved, so as to effectively improve and optimize the quality of tourism and tourism services. Therefore, a smart tourism management system that can integrate data from different sources, real-time perceive the state of tourists, dynamically optimize the tourism route, and has regional collaborative scheduling capability is needed. Therefore, a smart tourism online comprehensive management system is proposed. SUMMARY

[0004] The present application provides a smart tourism online comprehensive management method, which can make the evaluation of tourism resources more accurate, effectively solve the problems of large recommendation deviation, inability to dynamically adapt, and insufficient collaborative scheduling of traditional systems, and significantly improve the tourist experience and regional tourism service quality.

[0005] The technical solution adopted by the present application is as follows: a smart tourism online comprehensive management method, comprising the following steps: Step S1: collecting group behavior sample data through a plurality of user-authorized mobile terminals, the group behavior sample data including anonymized location trajectory sequences, stay duration at tourism resources, and desensitized consumption association records; Step S2: based on the group behavior sample data, calculating the dynamic average real play duration of each tourism resource, and generating a comprehensive credibility index of each tourism resource by fusing network public evaluation data processed by credibility; Step S3: based on the public tourism resource attribute data and the dynamic average real play duration and the comprehensive credibility index generated in step S2, constructing a regional industry knowledge graph; Step S4: According to the initial preferences of the current tourists, the regional industry knowledge graph, and the real-time resource state, a plurality of Pareto optimal tourism routes are generated in parallel for the current tourists to choose from; Step S5: During the execution of the itinerary, based on the personal state data authorized by the current tourists through their mobile terminals in real time, the dynamically updated resource state data, and the intention change data actively input by the users, the selected tourism route is dynamically adjusted and the resources are re-recommended; Step S6: Based on the regional industry knowledge graph and macro-situation data, tourist flow prediction and risk early warning are performed, and a multi-agent collaborative scheduling suggestion scheme is generated.

[0006] Further, the step S1 includes: Step S11: In response to the authorization instructions of a plurality of users through their mobile terminal application programs, continuously receive the location data packets reported from the mobile terminals after removing the personal direct identifiers, forming anonymous group trajectory data; Step S12: From the group trajectory data, identify and extract group stay events for specific tourism resource identifiers, and calculate a group stay duration sample set; Step S13: By associating the consumption voucher information uploaded after user local desensitization processing with the group stay events, generate consumption association records, which include tourism resource identifiers, consumption categories, and statistical amount intervals.

[0007] Further, the step S2 includes: Step S21: Statistically analyze the group stay duration sample set, remove outliers, and calculate the dynamic average real play duration of each tourism resource; Step S22: Perform pattern recognition on the network comment text collected from the Internet public platform, and perform confidence weight reduction processing on the text identified as suspicious evaluation; Step S23: Take the statistical distribution characteristics of the group stay duration sample set and the intensity of the consumption association records as high-weight factors, and take the network evaluation sentiment score after weight reduction processing as a low-weight factor. The comprehensive confidence index is calculated and generated by a weighted fusion model.

[0008] Further, the step S3 includes: Step S31: Collect public attribute data of tourism resources through an application programming interface, the public attribute data including at least geographic location, opening time, ticket price, and capacity information; Step S32: Taking tourism resources as entity nodes, the public attribute data as node attributes, and the spatial distance between tourism resources and the theme similarity as relationship edges, a knowledge graph static skeleton is constructed; Step S33: associate the dynamic average real play time and the comprehensive credibility index as dynamic attributes to the corresponding tourism resource entity node, and complete the construction of the regional industry knowledge graph.

[0009] Further, the step S4 includes: Step S41: receive the initial preference parameters set by the current tourist and the time budget constraint; Step S42: take the initial preference parameters and the time budget constraint as the target, take the entities, attributes and relationships in the regional industry knowledge graph as the constraint network, and introduce the comprehensive credibility index as one of the optimization targets to construct a multi-objective optimization model; Step S43: use a multi-objective evolutionary algorithm to solve the multi-objective optimization model, and output multiple Pareto optimal tourism routes that are not dominated in time, cost, interest matching degree and experience credibility at one time; Step S44: for each of the Pareto optimal tourism routes, accumulate the dynamic average real play time of each scenic spot and the traffic time calculated based on real-time traffic data to obtain the estimated total time of the route.

[0010] Further, the step S5 includes: Step S51: real-time acquire the current position and travel progress data authorized by the current tourist to share through his mobile terminal; Step S52: monitor the state data of the tourism resources and receive the macroscopic flow early warning information from the step S6; Step S53: when the actual stay time of the current tourist at a scenic spot deviates from the dynamic average real play time of the scenic spot by more than a predetermined threshold, or when early warning information affecting the subsequent travel is received, automatically trigger the re-planning of the remaining route; Step S54: analyze the intention change request input by the current tourist through natural language, combine the current position and the remaining time, search for matching resources in the regional industry knowledge graph, and generate an adjustment scheme.

[0011] Further, the step S6 includes: Step S61: based on the spatial relationship of the resources in the regional industry knowledge graph and the transition probability mined from the historical group behavior sample data, use a graph neural network model to predict the future short-term flow distribution of the region; Step S62: generate congestion and risk early warning information according to the predicted flow distribution and the capacity threshold of each resource; Step S63: based on the early warning information, simulate the effect of different scheduling strategies, and generate the multi-agent collaborative scheduling suggestion scheme.

[0012] Further, the method further comprises a step S7: the behavior data generated by the current tourist in the itinerary execution is fed back to the group behavior sample data set in step S1 after anonymization processing, as new sample data, for updating the dynamic average real play time and the comprehensive credibility index.

[0013] An intelligent tourism online comprehensive management system, comprising: A group behavior collection module, configured to execute the method in step S1 of claim 1 to obtain group behavior sample data; A credibility and time calculation module, configured to execute the method in step S2 of claim 1 to calculate a dynamic average real play time and a comprehensive credibility index; A knowledge graph construction module, configured to execute the method in step S3 of claim 1 to construct a regional industry knowledge graph; An intelligent planning module, configured to execute the method in step S4 of claim 1 to generate a Pareto optimal tourism route; A real-time interactive adjustment module, configured to execute the method in step S5 of claim 1 to dynamically adjust the itinerary; A collaborative scheduling engine, configured to execute the method in step S6 of claim 1 to generate a passenger flow prediction and a scheduling scheme; The output of the group behavior collection module is connected to the input of the credibility and time calculation module; the credibility and time calculation module and the knowledge graph construction module are bidirectionally connected; the output of the knowledge graph construction module is connected to the input of the intelligent planning module and the collaborative scheduling engine; the output of the intelligent planning module is connected to the input of the real-time interactive adjustment module.

[0014] Further, the system further comprises a user mobile terminal, configured to: In response to user authorization, report anonymous behavior data to the group behavior collection module; Provide an interface for the current tourist to interact with the intelligent planning module and the real-time interactive adjustment module, and display the planning route, the adjustment scheme and the warning information.

[0015] The present application has the following advantages: The application collects group behavior sample data through anonymization and desensitization, guarantees user privacy and can mine real play rules, the dynamic average real play time calculated based on this and the comprehensive credibility index fused with objective behavior data and network evaluation after credibility processing make the tourism resource evaluation more accurate; with the help of regional industry knowledge graph, static attributes and dynamic indicators are integrated to build a unified data support network, and multiple Pareto optimal routes are generated by combining multi-objective optimization algorithm to meet the personalized needs of tourists in time, cost, interest matching degree and experience credibility; the route can be adjusted according to the real-time state of tourists, resource dynamics and macro passenger flow warning during the trip, and the route can be changed automatically or in response to the intention to realize dynamic optimization of the trip; at the same time, the passenger flow can be predicted by the graph neural network model, the risk warning and multi-agent collaborative scheduling scheme are generated to assist regional tourism collaborative management, effectively solve the problems of large deviation of traditional system recommendation, inability of dynamic adaptation and lack of collaborative scheduling, and significantly improve the tourist experience and the quality of regional tourism service. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flowchart of the intelligent tourism online comprehensive management method of the embodiment of the application; Figure 2 A group behavior sample data acquisition flowchart of one embodiment of the application; Figure 3 A Pareto optimal tourism route generation flowchart of one embodiment of the application; Figure 4 A dynamic time and comprehensive credibility index calculation flowchart of one embodiment of the application; Figure 5 A block schematic diagram of the intelligent tourism online comprehensive management system of the embodiment of the application. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0018] Embodiment one As shown in the figure, the intelligent tourism online comprehensive management method of the embodiment of the application includes the following steps: Figures 1-4 Step S1: Collect group behavior sample data through a plurality of user-authorized mobile terminals, the group behavior sample data including anonymized location trajectory sequence, stay time at tourism resources and desensitized consumption association records; ​Step S1: Collecting the crowd behavior sample data, the core mathematical principle involves the mathematical representation of trajectory sequence and event detection. Let the trajectory sequence of an anonymous user be represented as a set of longitude and latitude coordinates under a timestamp sequence . For a specific tourism resource , its geofence can be defined as a region . The detection of stay events is to identify the continuous subsequence that satisfies , and calculate the stay duration . The stay duration of all users in constitutes the original sample set .

[0019] At the same time, the desensitized consumption records are mapped to the corresponding stay events through the association function , forming the preliminary statistical basis of consumption intensity . The mathematical model output of this step and provides raw materials for the subsequent authenticity and credibility calculation.

[0020] Step S2: Based on the crowd behavior sample data, calculate the dynamic average real play duration of each tourism resource, and fuse the network public evaluation data after credibility processing to generate the comprehensive credibility index of each tourism resource; In the calculation of dynamic average real play duration and comprehensive credibility index, the system purifies the original behavior data and fuses its mathematical model to extract stable features from noise data and quantify experience credibility.

[0021] Firstly, the statistical distribution-based outlier rejection method is used to process . Calculate its first quartile , third quartile and interquartile range , and the cleaned sample set is .

[0022] This cleaning process removes extreme values that are too short (such as passing by) or too long (such as staying). The mean value calculated subsequently better reflects the central tendency of real play. The dynamic average real play duration is calculated as . This value is a dynamically updated statistic, and its dynamic nature is reflected in periodic recalculation as new samples are updated.

[0023] Step S3: Based on the publicly available tourism resource attribute data and the dynamic average real play duration and comprehensive credibility index generated in step S2, construct a regional industry knowledge graph; The mathematical essence of constructing a regional industry knowledge graph is to construct a heterogeneous graph with spatiotemporal and semantic attributes . Wherein, the node set corresponds to the tourism resource . The node attribute is divided into static part (such as geographical location , ticket price , capacity ) and dynamic part (namely S2 calculated and TrustScore ).

[0024] The construction of the edge set depends on the relationship calculation. The weight of the spatial relationship edge can be defined as the Euclidean distance distance . The weight of the theme similarity relationship edge can be obtained by calculating the cosine similarity of the theme label vector in the node static attribute, .

[0025] The construction of the knowledge graph integrates scattered and multi-modal data (static attributes, dynamic behavior statistics, network evaluation, spatial and theme relationships) into a unified graph structure model. This graph model becomes the knowledge base and data constraint network for all subsequent intelligent decision-making (route planning, passenger flow prediction).

[0026] Step S4: According to the initial preference of the current tourist, the regional industry knowledge graph and the real-time resource state, a plurality of Pareto optimal tourism routes are generated in parallel for the current tourist to choose; Based on the constructed knowledge graph , the Pareto optimal tourism route generation is formalized as a multi-objective optimization problem. The core of the mathematical model is to define the decision variable, the objective function and the constraint condition. The decision variable is an ordered node (scenic spot) sequence . The objective function usually includes: minimizing the total time travel_time , wherein the traffic time depends on the real-time traffic; minimizing the total cost ; maximizing the interest matching degree PreferenceVec, Theme ; and maximizing the route credibility TrustScore .

[0027] The constraint condition comes from the attributes and relationships in the knowledge graph , for example: start_time open_time travel_time TimeBudget. This is a typical NP-hard problem. The system employs a multi-objective evolutionary algorithm (such as NSGA-II) to solve it. This algorithm searches the solution space by simulating the selection, crossover, and mutation operations of biological evolution, ultimately outputting a Pareto optimal solution set. .

[0028] Each route in the set It is non-dominant to all objectives, meaning there is no other route that is no worse than the previous one for all objectives. And strictly superior to at least one objective This provides users with multiple options to achieve different balances between time, cost, interest, and credibility. The S4 model directly relies on all the attributes provided by the S3 knowledge graph. and relationships (Used to calculate traffic time), this is the first deep application of knowledge graphs.

[0029] Step S5: During the trip, based on the real-time personal status data authorized by the tourist through their mobile terminal, the dynamically updated resource status data, and the user's actively input intention change data, the selected travel route is dynamically adjusted and resources are re-recommended. Once the trip begins, the mathematical model for dynamic adjustment and resource reallocation focuses on real-time decision-making and local replanning. Its triggering logic is based on monitoring, assuming tourists are at attractions... The actual length of stay is , when the deviation ( When the preset threshold is reached, or when the system receives a warning message from S6 about subsequent attractions, replanning is triggered.

[0030] Replanning is modeled as a constraint satisfaction and optimization problem in the current state. The current state is defined as: the tourist's current position. Time consumed ,time left TimeBudget Updated preferences. The problem lies in the knowledge graph. In the middle, looking for one from Departure, meeting time constraints And optimize the current preferences PreferenceVe New node sequence for credibility target Since the calculations are performed in real time during the journey, heuristic algorithms or constraint programming are typically used for fast solutions. The S5 model is an online and incremental version of S4, sharing the optimization objective of S4, but starting from the current state and relying on the real-time macroscopic information provided by S6.

[0031] Step S6: Based on the regional industry knowledge graph and macroeconomic situation data, conduct passenger flow forecasting and risk warning, and generate a multi-entity collaborative scheduling suggestion scheme.

[0032] The mathematical model for passenger flow prediction and coordinated scheduling aims to utilize knowledge graphs and historical group behavior for macro-level situational awareness; its core is a spatiotemporal prediction model. This involves using knowledge graphs... Viewed as a spatial network, each node In different periods of history Passenger flow composition characteristics The probability matrix of passenger flow transfer between nodes can be extracted from historical group behavior sample data. , of which elements Indicates from tourist attractions Flowing to scenic spots The probability of.

[0033] Graph neural networks (GNNs), such as the Spatiotemporal Graph Convolutional Network (STGCN), can simultaneously capture spatial dependencies (through graph convolution) and temporal dependencies (through time series models such as GRU). The prediction model can be abstractly represented as: GNN ,in To predict the step size, It serves as a window for historical observation.

[0034] Based on predicted passenger flow and node capacity It can generate early warnings, if ( If a warning threshold (e.g., 0.8) is set, then it is marked as a congestion risk point. The generation of a collaborative scheduling scheme is a simulation optimization problem. On a multi-agent model including road networks and transportation vehicles, different scheduling strategies are simulated (e.g., guiding tourist flow through APP push notifications, adjusting shuttle bus frequency) to minimize the overall congestion level or the average tourist waiting time, selecting the optimal strategy. The S6 model is a knowledge graph. In another in-depth application at the macro level, the prediction results (early warning information) are in turn fed back as key inputs to the real-time adjustment module of S5, forming a closed loop of macro prediction guiding micro adjustment.

[0035] Furthermore, step S1 includes: Step S11: In response to authorization instructions from multiple users through their mobile terminal applications, continuously receive location data packets with removed personal direct identifiers reported from the mobile terminals to form anonymous group trajectory data; Step S12: From the group trajectory data, identify and extract the group stay events for the specific tourism resource identifier, and calculate the group stay duration sample set; Step S13: By associating the consumption voucher information uploaded after user local desensitization processing with the group stay events, generate consumption association records, which contain tourism resource identifier, consumption category and statistical amount interval.

[0036] Specifically, step S2 includes: Step S21: Statistically analyze the group stay duration sample set, and after removing outliers, calculate the dynamic average real play duration of each tourism resource; Step S22: Perform pattern recognition on the network comment text collected from the Internet public platform, and perform confidence weight reduction processing on the text identified as suspicious evaluation; Step S23: Take the statistical distribution characteristics of the group stay duration sample set and the strength of the consumption association record as high-weight factors, and take the network evaluation sentiment score after weight reduction processing as a low-weight factor, and calculate the comprehensive confidence index through a weighted fusion model.

[0037] At this time, in order to obtain a comprehensive experience evaluation beyond single behavior data, the system introduces network public evaluation data and designs a confidence fusion model. Suppose the comment text obtained from the network about has an original sentiment score obtained through sentiment analysis. Through pattern recognition (such as detecting brush evaluation mode and false text features), a confidence weight is assigned to it. The network evaluation score after weight reduction processing is . The comprehensive confidence index TrustScore is calculated through a weighted fusion model, which gives higher weight to objective behavior data: TrustScore Norm Norm . Among them, is the variance of the stay duration sample , and the smaller the variance, the more concentrated the play time and the more stable the experience. Its reciprocal after normalization Norm is used as the input of the high weight ; is the normalized consumption strength (such as per capita consumption, consumption proportion), which is used as the input of another high weight ; is the network evaluation score after weight reduction processing, which is used as the input of the low weight , and satisfies The principle of this model is that the data of the voting behavior of the group using experience time and consumption ratio is more objective than subjective text evaluation, but the latter can be supplemented. At this point, S2 is used to vote for each tourism resource Two key dynamic attributes are generated: and TrustScore Specifically, step S3 includes: Step S31: Collecting the public attribute data of the tourism resource through the application programming interface, and the public attribute data at least includes the geographic location, opening time, ticket price and capacity information; Step S32: Taking the tourism resource as an entity node, taking the public attribute data as a node attribute, and taking the spatial distance and theme similarity between the tourism resources as a relationship edge, a knowledge graph static skeleton is constructed; Step S33: The dynamic average real play time and the comprehensive trust index are associated to the corresponding tourism resource entity node as dynamic attributes, and the construction of the regional industry knowledge graph is completed.

[0038] Specifically, step S4 includes: Step S41: Receiving the initial preference parameters and time budget constraints set by the current tourist; Step S42: Taking the initial preference parameters and time budget constraints as the target, taking the entities, attributes and relationships in the regional industry knowledge graph as the constraint network, and introducing the comprehensive trust index as one of the optimization targets, a multi-objective optimization model is constructed; Step S43: A multi-objective evolutionary algorithm is used to solve the multi-objective optimization model, and a plurality of Pareto optimal tourism routes which are not dominated in time, cost, interest matching degree and experience trust are output at one time; Step S44: For each Pareto optimal tourism route, the dynamic average real play time of each scenic spot in the route and the traffic time calculated based on real-time traffic data are accumulated to obtain the estimated total time of the route.

[0039] Specifically, step S5 includes: Step S51: Real-time acquisition of the current position and travel progress data authorized by the current tourist through the mobile terminal; Step S52: Monitoring the state data of the tourism resource, and receiving the macroscopic flow warning information from step S6; Step S53: When the actual stay time of the current tourist at a scenic spot deviates from the dynamic average real play time of the scenic spot by more than a predetermined threshold, or when the warning information affecting the subsequent travel is received, the remaining route is automatically re-planned; Step S54: Analyze the intention change request of the current tourist through natural language input, combine the current location and the remaining time, search for matching resources in the regional industry knowledge graph, and generate an adjustment scheme.

[0040] Specifically, step S6 includes: Step S61: Based on the spatial relationship of resources in the regional industry knowledge graph and the transfer probability mined in the historical group behavior sample data, use a graph neural network model to predict the future short-term passenger flow distribution of the region; Step S62: According to the predicted passenger flow distribution and the capacity threshold of each resource, generate congestion and risk warning information; Step S63: Based on the warning information, simulate the effect of different scheduling strategies (such as guiding and shunting, adjusting traffic capacity), and generate a multi-agent collaborative scheduling suggestion scheme.

[0041] Specifically, the method further includes step S7: The behavior data generated by the current tourist in the execution of the itinerary is anonymized and fed back to the group behavior sample data set in step S1 as new sample data, which is used to update the dynamic average real play duration and the comprehensive trust score.

[0042] When the tourist completes the itinerary, his anonymized trajectory and stay data are fed back to the original data set in S1. This enables the and the consumption records used to calculate are updated, thereby triggering the and the recalculation of TrustScore . The updated dynamic attributes are synchronized to the node attributes of the knowledge graph . This means that the knowledge base used for S4 planning, S5 adjustment, and S6 prediction is continuously evolving, and the system can adapt to changes in group behavior patterns.

[0043] As shown in Figure 5 , the present application also proposes a smart tourism online comprehensive management system corresponding to the method of the above embodiment, which includes: A group behavior collection module for executing the method of step S1 in claim 1 to obtain group behavior sample data; A trust score and duration calculation module for executing the method of step S2 in claim 1 to calculate the dynamic average real play duration and the comprehensive trust score; A knowledge graph construction module for executing the method of step S3 in claim 1 to construct a regional industry knowledge graph; An intelligent planning module for executing the method of step S4 in claim 1 to generate a Pareto optimal tourism route; a real-time interactive adjustment module for performing the method of step S5 in claim 1 to make dynamic adjustment of the itinerary; a collaborative scheduling engine for performing the method of step S6 in claim 1 to make passenger flow prediction and scheduling scheme generation; wherein the output of the crowd behavior collection module is connected to the input of the credibility and duration calculation module; the credibility and duration calculation module is bidirectionally connected to the knowledge graph construction module; the output of the knowledge graph construction module is connected to the input of the intelligent planning module and the collaborative scheduling engine; the output of the intelligent planning module is connected to the input of the real-time interactive adjustment module.

[0044] Specifically, the system further comprises a user mobile terminal, which is configured to: report anonymous behavior data to the crowd behavior collection module in response to user authorization; provide an interface for the current tourist to interact with the intelligent planning module and the real-time interactive adjustment module, and display the planned route, adjustment scheme and warning information.

[0045] The system of the present application forms an intelligent closed loop capable of continuous learning and dynamic optimization. First, the system collects a large amount of real behavior data of tourists, such as location trajectory, stay duration and consumption record, under the authorization of the user, as the factual basis for all judgments.

[0046] Based on these crowd data, the system calculates the dynamic average real play duration of each scenic spot and a credibility index that integrates objective behavior and network evaluation. Then, these two dynamic indicators are fused with the static information of the scenic spot (such as location, theme, capacity) to construct a regional industry knowledge graph. This graph is equivalent to the intelligent center of the system, connecting scattered data into a knowledge network.

[0047] When the tourist uses it, the system uses this knowledge graph as a sand table to generate multiple recommended routes that achieve different balances in time, cost, interest and credibility according to the tourist's preferences and time budget through multi-objective optimization algorithm. During the trip, the system can real-time perceive the tourist's progress and surrounding state through the mobile phone, and once there is a deviation (such as overstay or congestion ahead), it will automatically trigger route re-planning and can understand the natural language request of the tourist for flexible adjustment.

[0048] At the macro level, the system predicts the regional passenger flow distribution based on the knowledge graph and historical passenger flow data, issues early warnings, and simulates collaborative scheduling schemes for scenic area, transportation and other management parties. These macro warnings are fed back to the individual route adjustment module in real time. Finally, the anonymous behavior data of each tourist will be fed back to the system for updating the play duration and credibility index.

[0049] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific feature, structure, material or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. Descriptive expressions of the above terms in the specification do not necessarily refer to the same embodiment or example. Also, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples, without contradiction.

[0050] Any process or method descriptions or descriptions of the flow diagrams in the flow charts described herein and elsewhere can be understood as representing the steps of the processes and methods as sets of operations that can be implemented in hardware, software, or a combination of both, and that can be implemented in an order different than the order described. One of ordinary skill in the art can also understand that the processes and methods described in the specification can be implemented using one or more articles of manufacture that store program code. Any resulting process or method can be made up of any number of processes or methods. The methods or processes described can also be undertaken as a computer program running on one or more computer systems and / or computer processors that each have computer program instructions implemented to perform the process or method.

[0051] The logic and / or steps represented in the flow charts and / or described herein, for example, can be embodied in computer executable code, that can be implemented by or in connection with an instruction execution system, apparatus, or device such as a computer based system, processor containing system, or other system that can fetch the instructions from a non-transitory computer readable storage medium and execute the instructions. For purposes of this specification, a "computer readable medium" can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. The computer readable medium can be a non-transitory computer readable medium. The non-transitory computer readable medium can include, but is not limited to, electronic, optical, magnetic, electromagnetic, infrared, semiconductor systems (or apparatuses) including a non-transitory computer readable medium. The non-transitory computer readable medium can also be paper or other suitable medium upon which the program is printed, as the program can be electronically captured, via optical scanning of the paper or other suitable medium, then compiled, interpreted, or otherwise processed in a suitable manner, if necessary, and stored in a computer memory to execute the computer program.

[0052] It should be understood that parts of the present application can be implemented in hardware, software, firmware or a combination thereof. In the above embodiments, a plurality of steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and as in another embodiment, it can be implemented using any or a combination of the following technologies, which are well known in the art: discrete logic circuitry having logic gates for implementing logic functions upon data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

[0053] Those skilled in the art can understand that all or part of the steps carried out by the above-mentioned embodiment methods can be completed by programs instructing relevant hardware, and the programs can be stored in a computer readable storage medium, and when the programs are executed, one or a combination of the steps of the method embodiments is included.

[0054] In addition, each functional unit in each embodiment of the present application can be integrated into one processing module, or each unit can exist physically, or two or more units can be integrated into one module. The above integrated module can be realized in the form of hardware or in the form of a software functional module. The integrated module, if realized in the form of a software functional module and sold or used as an independent product, can also be stored in a computer readable storage medium.

[0055] Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and cannot be understood as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A smart tourism online integrated management method, characterized in that, Includes the following steps: Step S1: Collect group behavior sample data through multiple user-authorized mobile terminals. The group behavior sample data includes anonymized location trajectory sequences, length of stay at tourist resources, and desensitized consumption association records. Step S2: Based on the group behavior sample data, calculate the dynamic average actual play time of each tourism resource, and integrate the publicly available online evaluation data after credibility processing to generate a comprehensive credibility index for each tourism resource. Step S3: Based on publicly available tourism resource attribute data, the dynamic average actual playtime generated in Step S2, and the comprehensive credibility index, construct a regional industry knowledge graph; Step S4: Based on the current tourist's initial preferences, the regional industry knowledge graph, and the real-time resource status, generate multiple Pareto optimal tourist routes in parallel for the current tourist to choose from; Step S5: During the trip, based on the current tourist's real-time authorized personal status data via their mobile terminal, dynamically updated resource status data, and user-initiated intention change data, the selected travel route is dynamically adjusted and resources are re-recommended. Step S6: Based on the regional industry knowledge graph and macroeconomic situation data, conduct passenger flow prediction and risk warning, and generate a multi-entity collaborative scheduling suggestion scheme.

2. The smart tourism online integrated management method according to claim 1, characterized in that, Step S1 includes: Step S11: In response to authorization instructions from multiple users through their mobile terminal applications, continuously receive location data packets with removed personal direct identifiers reported from the mobile terminals to form anonymous group trajectory data; Step S12: Identify and extract group stay events targeting specific tourism resource identifiers from the group trajectory data, and calculate the group stay duration sample set; Step S13: Generate a consumption association record by associating the consumption voucher information uploaded after local anonymization by the user with the group stay event. The consumption association record includes tourism resource identifier, consumption category and statistical amount range.

3. The smart tourism online integrated management method according to claim 1, characterized in that, Step S2 includes: Step S21: Perform statistical analysis on the sample set of group stay duration, remove outliers, and calculate the dynamic average real play duration of each tourism resource; Step S22: Perform pattern recognition on the online comment texts collected from publicly available internet platforms, and reduce the confidence weight of texts identified as suspicious comments; Step S23: The statistical distribution characteristics of the group stay duration sample set and the strength of the consumption association records are used as high-weight factors, and the reduced-weight network evaluation sentiment scores are used as low-weight factors. The comprehensive credibility index is calculated and generated through a weighted fusion model.

4. The smart tourism online integrated management method according to claim 1, characterized in that, Step S3 includes: Step S31: Collect publicly available attribute data of tourism resources through the application programming interface. The publicly available attribute data includes at least geographical location, opening hours, ticket prices, and capacity information. Step S32: Construct a static skeleton of a knowledge graph using tourism resources as entity nodes, the publicly available attribute data as node attributes, and the spatial distance and theme similarity between tourism resources as relation edges; Step S33: The dynamic average actual playtime and the comprehensive credibility index are used as dynamic attributes and associated with the corresponding tourism resource entity nodes to complete the construction of the regional industry knowledge graph.

5. The smart tourism online integrated management method according to claim 1, characterized in that, Step S4 includes: Step S41: Receive the initial preference parameters and time budget constraints set by the current tourist; Step S42: Using the initial preference parameters and the time budget constraint as objectives, the entities, attributes and relationships in the regional industry knowledge graph as constraint networks, and the comprehensive credibility index as one of the optimization objectives, a multi-objective optimization model is constructed. Step S43: Use a multi-objective evolutionary algorithm to solve the multi-objective optimization model and output multiple Pareto optimal travel routes that are independent of each other in terms of time, cost, interest matching degree and experience credibility at one time; Step S44: For each Pareto optimal travel route, sum the dynamic average actual playtime of each attraction and the travel time calculated based on real-time traffic data to obtain the estimated total time of the route.

6. The smart tourism online integrated management method according to claim 1, characterized in that, Step S5 includes: Step S51: Obtain in real time the current location and itinerary progress data authorized and shared by the current tourist through their mobile terminal; Step S52: Monitor the status data of tourism resources and receive macro-level passenger flow early warning information from step S6; Step S53: When the actual stay time of the current tourist at a certain attraction deviates from the dynamic average real play time of that attraction by more than a predetermined threshold, or when a warning message affecting the subsequent itinerary is received, the replanning of the remaining route is automatically triggered. Step S54: Analyze the intent change request input by the current tourist through natural language, combine it with the tourist's current location and remaining time, retrieve matching resources in the regional industry knowledge graph, and generate an adjustment plan.

7. The smart tourism online integrated management method according to claim 1, characterized in that, Step S6 includes: Step S61: Based on the spatial relationships of resources in the regional industry knowledge graph and the transfer probabilities mined from the historical group behavior sample data, use a graph neural network model to predict the future short-term passenger flow distribution in the region; Step S62: Generate congestion and risk warning information based on the predicted passenger flow distribution and the capacity thresholds of each resource; Step S63: Based on the early warning information, simulate the effects of different scheduling strategies and generate the multi-entity collaborative scheduling suggestion scheme.

8. The smart tourism online integrated management method according to claim 1, characterized in that, The method further includes step S7: anonymizing the behavioral data generated by the current tourist during the trip and feeding it back as new sample data to the group behavior sample dataset in step S1, so as to update the dynamic average real playtime and the comprehensive credibility index.

9. A smart tourism online integrated management system, characterized in that, include: A group behavior acquisition module is used to execute the method described in step S1 of claim 1 to obtain group behavior sample data; The credibility and duration calculation module is used to perform the method described in step S2 of claim 1 to calculate the dynamic average real playtime and the comprehensive credibility index. A knowledge graph construction module is used to perform the method described in step S3 of claim 1 to construct a regional industry knowledge graph; The intelligent planning module is used to perform the method described in step S4 of claim 1 to generate a Pareto optimal travel route; The real-time interactive adjustment module is used to execute the method described in step S5 of claim 1 to dynamically adjust the trip. A collaborative scheduling engine is used to execute the method described in step S6 of claim 1 to perform passenger flow prediction and scheduling scheme generation; Specifically, the output of the group behavior acquisition module is connected to the input of the credibility and duration calculation module; the credibility and duration calculation module is bidirectionally connected to the knowledge graph construction module; the output of the knowledge graph construction module is connected to the input of the intelligent planning module and the collaborative scheduling engine; and the output of the intelligent planning module is connected to the input of the real-time interactive adjustment module.

10. A smart tourism online integrated management system according to claim 9, characterized in that, The system also includes a user mobile terminal, which is used for: In response to user authorization, anonymous behavior data is reported to the group behavior collection module; The system provides the current tourist with an interface to interact with the intelligent planning module and the real-time interactive adjustment module, and displays the planned route, adjustment scheme, and early warning information.

Images