Intelligent tour guide scheduling system based on internet of things and tourist behavior portrait

Abstract

This invention belongs to the field of park tour guidance and scheduling, and involves data analysis technology to solve the problem that existing intelligent tour route recommendation technologies ignore the "endogenous conflicts" between recommended routes, which may exacerbate systemic congestion. Specifically, it is an intelligent tour guidance and scheduling system based on the Internet of Things and visitor behavior profiles, including a route recommendation module, a conflict analysis module, a profile analysis module, and a scheduling management module that are connected in sequence. The route recommendation module, conflict analysis module, and profile analysis module are all connected to a database. This application can comprehensively and without omission traverse and extract all continuous interest node sequences in the recommended route through a sliding window mechanism. This method ensures that even potential conflict points in the middle of the route, such as sequential intersections or overlaps, can be identified and evaluated by the system, thereby avoiding the problem of missed conflict evaluation that may exist in traditional methods.

Description

Technical Field

[0001] This invention belongs to the field of park tour scheduling and involves data analysis technology, specifically an intelligent tour scheduling system based on the Internet of Things and visitor behavior profiles. Background Technology

[0002] With the rapid development of IoT, location navigation and big data technologies, intelligent tour guide systems have become a core tool for smart scenic spots, museums, theme parks and other scenarios to improve visitor experience and operational efficiency. Existing intelligent tour guide route recommendation technologies are mainly based on personalized preference recommendations and static rule-based planning.

[0003] Existing intelligent tour guide route recommendation technologies neglect the "endogenous conflicts" between recommended routes, which may exacerbate systemic congestion. This is a core blind spot of existing technologies. When the system independently generates "optimal" recommended routes for tourist groups entering from different entrances, these routes are very likely to intersect or overlap in the sequence of popular attractions. At this time, the two main tourist flows continue to move towards each other between the two intersecting or overlapping tourist nodes, forming a "scissor flow" type of continuous congestion in bottleneck areas such as connecting passages and entrances / exits. This structural and predictable congestion caused by the design flaws of the recommendation scheme itself cannot be solved by existing "post-event diversion" systems. In addition, existing technologies cannot combine route recommendation analysis with user profiles to provide users with customized recommended routes.

[0004] To address the aforementioned technical problems, this application proposes a solution. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent tour guide scheduling system based on the Internet of Things and tourist behavior profiles, in order to solve the problem that existing intelligent tour guide route recommendation technologies ignore the "endogenous conflict" between recommended routes, which may exacerbate systemic congestion;

[0006] The technical problem that this invention aims to solve is: how to provide an intelligent tour guide scheduling system based on the Internet of Things and tourist behavior profiles that can be locally optimized to avoid "endogenous conflicts" between recommended routes.

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] The intelligent tour guide and dispatch system based on the Internet of Things and tourist behavior profiles includes a route recommendation module, a conflict analysis module, a profile analysis module, and a dispatch management module that are connected in sequence. The route recommendation module, conflict analysis module, and profile analysis module are all connected in communication with the database.

[0009] The route recommendation module is used to recommend and analyze the tour routes in the park: it generates a statistical period, marks the park's exits, entrances, and exhibition halls as nodes of interest, locates tourists through the Internet of Things positioning system (Bluetooth beacon, Wi-Fi probe) and gate access records within the statistical period, marks the first non-entrance node of interest visited by tourists after entering the park as the starting node, marks the number of times the exhibition hall is marked as the starting node as the starting tendency value, and generates L1 recommended routes starting from the L1 nodes of interest with the largest starting tendency values.

[0010] The conflict analysis module is used to analyze the conflict level of all recommended routes in the park: the recommended routes are sorted in order of generation time to obtain the generation sequence, a traversal window is generated and the traversal value of the traversal window is set, and the interest nodes of all recommended routes are traversed and the conflict assessment is performed through the traversal window.

[0011] The profile analysis module is used to analyze tourist behavior profiles: the behavior profiles include ticket booking profiles and tour profiles. The ticket booking profile features include solo tours, duo tours, and group tours; the tour profile features include light tours and in-depth tours.

[0012] The scheduling and management module is used to recommend routes to tourists and to manage and analyze resource scheduling within the park.

[0013] Furthermore, the process of generating recommended routes includes: marking the interest node with the largest initial tendency value as a priority node; marking the first visitor whose non-entry interest node is a priority node as a priority visitor; marking the number of times a priority visitor enters other interest nodes after leaving a priority node as the follow value of the corresponding interest node; and marking the follow nodes of the recommended route based on the follow value. Then, marking the interest node with the second largest initial tendency value as a priority node and generating the corresponding recommended route, and so on, until the number of recommended routes reaches L1.

[0014] Furthermore, the specific process of marking the following nodes of the recommended route includes: marking the interest node with the largest follow value as the following node; retrieving the follow threshold from the database; comparing the follow value of the following node with the follow threshold; if the follow value is less than the follow threshold, the recommended route is terminated; if the follow value is greater than or equal to the follow value, it is determined whether the following node is the exit: if yes, the recommended route is terminated; if no, the following node is added to the recommended route, and the marking of following nodes continues until the recommended route is terminated.

[0015] Furthermore, the specific process of traversing and extracting interest nodes from all recommended routes includes: marking the traversal value as k, extracting interest nodes with indices of 1-k from all recommended routes and forming a conflict analysis set, and performing conflict evaluation analysis on the conflict analysis set; then extracting interest nodes with indices of 2-k+1 from all recommended routes and forming a new conflict analysis set, and performing conflict evaluation analysis on the new conflict analysis set; and so on, until all interest nodes in all recommended routes have been added to the conflict analysis set and subjected to conflict evaluation analysis.

[0016] Furthermore, the specific process of conducting conflict assessment analysis on the conflict analysis set includes: marking the number of elements in the conflict analysis set corresponding to the interest node as the conflict value of the interest node; retrieving the conflict threshold and concentration threshold from the database; marking interest nodes with conflict values ​​not less than the conflict threshold as conflict nodes; marking recommended routes where the sequence number of the conflict node in the recommended route conforms to the truncation rule as the conflict route of the conflict node; calculating the variance of the node sequence number of the same conflict node in all conflict routes to obtain the concentration coefficient of the conflict node; comparing the concentration coefficient with the concentration threshold: if the concentration coefficient is greater than the concentration threshold, it is determined that the conflict node does not have a moderating characteristic; if the concentration coefficient is less than or equal to the concentration threshold, it is determined that the conflict node has a moderating characteristic; and inserting a buffer node before the last k / 2 conflict nodes in the conflict route.

[0017] Furthermore, the ticket booking profile features are obtained directly by retrieving tourists' ticket booking information; the labeling process for the tour profile features includes: after tourists enter the park, the average tour duration of tourists at all interest nodes other than entrances and exits is counted and labeled as an immersion value, the immersion threshold is retrieved from the database, and the immersion value is compared with the immersion threshold: if the immersion value is less than the immersion threshold, the corresponding tourist's tour profile features are labeled as light tour; if the immersion value is greater than or equal to the immersion threshold, the corresponding tourist's tour profile features are labeled as deep tour.

[0018] Furthermore, the specific process of the scheduling management module in analyzing routes for tourists includes: at the beginning of the statistical period, retrieving the selection frequency of all recommended routes corresponding to different booking profile features of tourists in the previous statistical period and marking it as the selection probability value of the recommended route corresponding to the booking profile feature; when recommending routes within the statistical period, obtaining the booking profile features of tourists, retrieving the selection probability values ​​of all recommended routes corresponding to the booking profile features, generating a recommendation sequence in descending order of selection probability value, and sending the recommendation sequence to the optimized mobile terminal.

[0019] Furthermore, the specific process of the scheduling management module in analyzing resource scheduling management within the park includes: periodically acquiring the number of tourists with in-depth tour profiles at each non-exit interest node within the statistical period and marking them as in-depth density values; comparing the in-depth density values ​​with preset in-depth density thresholds: if the in-depth density value is less than the in-depth density threshold, no processing is performed; if the in-depth density value is greater than or equal to the in-depth density threshold, the corresponding interest node is marked as an in-depth node, and scheduling information is released to the guides near the in-depth node.

[0020] The present invention has the following beneficial effects:

[0021] 1. Based on data such as IoT positioning systems and gate access records, the initial tendency value and follow value of each interest node are calculated by analyzing tourist behavior, and a recommended route is generated based on this. On this basis, this application further introduces a dynamic route termination mechanism. This mechanism first identifies the interest node with the largest follow value in the current path as the follow node, ensuring that the extension direction of the route is always based on the most attractive path in the tourist behavior data.

[0022] 2. This application can comprehensively and without omission traverse and extract all continuous interest node sequences in the recommended route through a sliding window mechanism. This method ensures that even potential conflict points in the middle of the route, such as sequential intersections or overlaps, can be identified and evaluated by the system, thereby avoiding the problem of missed conflict evaluation that may exist in traditional methods. This enables the intelligent tour guide scheduling system to more accurately identify and predict congestion risks in the park, providing a more reliable basis for subsequent scheduling management, effectively alleviating structural congestion caused by design defects in the recommended scheme itself, and significantly improving the system's early warning and management capabilities.

[0023] 3. Based on tourists' actual tour behavior, dynamically and accurately label their tour profile characteristics. Specifically, by directly obtaining ticket booking profile characteristics, basic information is provided for tourist profiling. More importantly, by statistically analyzing the average tour duration of tourists at interest nodes outside the entrances and exits as an immersion value and comparing it with an immersion threshold, an objective and quantitative distinction is achieved between light and deep tours. This solves the problem of traditional methods lacking specific data-driven approaches to accurately distinguish the depth of tourist tours, enabling the profile analysis module to generate more refined behavioral profiles that better reflect tourists' true intentions.

[0024] 4. During the statistical period, the number of tourists with in-depth tour characteristics within interest nodes outside the entrances and exits is periodically acquired and marked as in-depth density values. This enables dynamic and real-time monitoring of the density of specific high-value tourist groups (in-depth tour tourists) within the park. This allows the system to accurately identify potential congestion hotspots, rather than relying solely on the total number of tourists, thereby improving the accuracy and targeting of early warnings. The in-depth density values ​​are compared with preset in-depth density thresholds, and the system decides whether to take action based on the comparison results, avoiding unnecessary resource intervention and optimizing system operating efficiency. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a system block diagram of Embodiment 1 of the present invention;

[0027] Figure 2 This is a flowchart of the method in Embodiment 2 of the present invention. Detailed Implementation

[0028] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] In traditional intelligent tour guide systems, inherent conflicts between recommended routes are not identified and addressed, leading to systemic congestion. Specifically, when recommending routes independently for groups of tourists entering from different entrances, these routes intersect or overlap in the sequence of popular attractions, causing tourists to continuously move towards each other in bottleneck areas such as connecting passages and entrances / exits, resulting in scissor-flow congestion. Furthermore, route recommendation analysis is not integrated with user profiling, making it impossible to provide customized recommended routes, thus impacting key performance indicators such as tourist flow efficiency and system resource utilization.

[0030] For example, during peak holiday periods at a large museum, recommended routes are generated by the system based on IoT location data. Because visitor groups from multiple entrances tend to visit both the main exhibition hall and special exhibition areas, the system-generated recommended routes intersect at connecting passageways. Specifically, visitors entering from the east entrance are recommended to visit the main exhibition hall first, while those entering from the west entrance are recommended to visit the special exhibition area first. This causes the two visitor flows to move in opposite directions within the passageway, resulting in continuous congestion. Consequently, visitor delays occur, the system cannot promptly implement resource allocation, and the smooth flow of the visit is disrupted.

[0031] If these problems are not addressed, systemic congestion will inevitably worsen, hindering tourist movement and creating safety hazards. Furthermore, the lack of a recommendation mechanism that integrates user profiles makes it impossible to differentiate between booking profiles (solo, couple, and group travel) and tour profiles (light and in-depth travel), resulting in mismatches between recommended routes and actual tourist needs, further reducing service quality and operational efficiency. In the long run, this will damage the system's reliability and sustainable operation.

[0032] For ease of understanding, the following explains some key terms in this embodiment:

[0033] IoT positioning system (Bluetooth beacons, Wi-Fi probes): This system is configured to acquire real-time location information of visitors within the park. By deploying devices such as Bluetooth beacons and Wi-Fi probes, signals emitted by visitors' mobile devices can be collected, enabling precise tracking of visitors.

[0034] Gate access records: These records are used to obtain time information of tourists entering and exiting the park or specific areas, serving as a data source for auxiliary positioning and behavior analysis.

[0035] Interest nodes: These are locations within the park that are attractive to visitors, such as the park's exits, entrances, and various exhibition halls. These nodes are key points in the visitor's tour route.

[0036] Starting node: After a visitor enters the park, the first non-entrance interest node they visit is identified and marked as the starting node, which represents the actual starting point of the visitor's tour within the park.

[0037] Starting Tendency Value: This value measures the number of times a particular pavilion is marked as a starting point for visitors. The higher the value, the more popular the pavilion is as a starting point for tourists.

[0038] Traversal Window: When analyzing recommended routes, this logical window is used to slide and evaluate the route sequence. Its size is set by the traversal value.

[0039] Behavioral profile: This profile is an abstraction and summary of the characteristics of tourists' travel behavior, including ticket booking profile and tour profile, which is used for personalized recommendations and scheduling.

[0040] Ticketing profile features: These features are formed based on tourists' ticketing information, such as whether they are traveling alone, in pairs, or in groups, reflecting the size of the tourist group.

[0041] Visitor profile characteristics: These characteristics are formed based on the actual visitor behavior within the park, such as light or in-depth visits, reflecting the visitor's preferences and level of engagement.

[0042] Example 1: As Figure 1 As shown, the intelligent tour guide and dispatch system based on the Internet of Things and tourist behavior profiles includes a route recommendation module, a conflict analysis module, a profile analysis module, and a dispatch management module that are connected in sequence. The route recommendation module, conflict analysis module, and profile analysis module are all connected to the database.

[0043] The route recommendation module is used to recommend and analyze tour routes within the park. It generates a statistical period, marks the park's entrances, exits, and exhibition halls as nodes of interest, and locates visitors within this period using an IoT positioning system (Bluetooth beacons, Wi-Fi probes) and gate access records. The first non-entrance node visited by a visitor after entering the park is marked as the starting node. The number of times an exhibition hall is marked as a starting node is recorded as the initial tendency value. L1 recommended routes are generated starting from the L1 nodes with the highest initial tendency values. The node with the highest initial tendency value is marked as the priority node, and then the visitor whose first non-entrance node is the priority node is marked as the preferred node. First, for each visitor, the number of visitors who leave the priority node and enter other interest nodes is marked as the follow value of the corresponding interest node. The interest node with the largest follow value is marked as a follow node. The follow threshold is retrieved from the database, and the follow value of the follow node is compared with the follow threshold. If the follow value is less than the follow threshold, the recommended route is terminated. If the follow value is greater than or equal to the follow value, it is determined whether the follow node is the exit. If yes, the recommended route is terminated. If not, the follow node is added to the recommended route, and the marking of follow nodes continues until the recommended route is terminated. Then, the interest node with the second largest initial tendency value is marked as a priority node and a corresponding recommended route is generated. This process is repeated until the number of recommended routes reaches L1.

[0044] Specifically, marking the interest node with the highest initial tendency value as a priority node means that the system identifies the most popular starting point for visitors based on data from the first non-entrance interest node visited by visitors after entering the park within a statistical period. For example, the system can periodically (e.g., hourly or daily) calculate the initial tendency value of all interest nodes and automatically identify the interest node with the highest current value as a priority node. Alternatively, the system can combine historical big data analysis and real-time data to dynamically adjust and determine the most popular starting interest node as the priority node.

[0045] Then, the visitor whose first non-entry point of interest is a priority node is marked as a priority visitor. This means the system identifies the group of visitors who actually begin their tour from the currently marked priority node of interest. For example, IoT positioning systems (such as Bluetooth beacons, Wi-Fi probes) or gate access records continuously monitor the real-time location and access sequence of visitors. When a visitor enters the park, if their first non-entry point of interest matches the currently marked priority node, that visitor is identified and marked as a priority visitor by the system. Alternatively, the system can maintain a visitor status database, recording the real-time location and sequence of interest nodes visited by each visitor. Once a visitor's first non-entry point of interest is confirmed as a priority node, the system updates the visitor's status and classifies them as a priority visitor.

[0046] Next, the number of times a priority visitor enters other interest nodes after leaving the priority node is marked as the follow value of the corresponding interest node. This means the system quantifies the frequency or preference of priority visitors for subsequent specific interest nodes after leaving the priority node. For example, the system continuously tracks the movement trajectory of priority visitors. When a priority visitor leaves the priority node and enters the next interest node, the system increases the follow value of that "next interest node". If 30 out of 100 priority visitors go to Hall A after leaving the priority node, then the follow value of Hall A increases by 30. Alternatively, a directed graph model can be constructed, where nodes are interest nodes and edges represent the movement of visitors from one node to another. For priority visitors, when they move from the priority node to other interest nodes, the weight of the corresponding edge (i.e., the follow value) increases.

[0047] The following example will provide a more detailed explanation of the above technical solution:

[0048] Imagine a large theme park with multiple entrances, exits, and exhibits, such as "Fantasy Castle," "Future World," and "Adventure Island," all marked as nodes of interest. The system first generates a statistical period, for example, from 9:00 AM to 1:00 PM. Within this period, an IoT positioning system consisting of Bluetooth beacons and Wi-Fi probes deployed throughout the park, along with access records from each gate, continuously acquires real-time location data of visitors.

[0049] When visitor A enters the park through the entrance, their first non-entrance interest node is "Fantasy Castle," which is then marked as visitor A's starting node. The system counts the number of times all visitors use "Fantasy Castle" as their starting node within the current statistical period and marks this as the starting tendency value for "Fantasy Castle." Similarly, it also counts the starting tendency values ​​for exhibits such as "Future World" and "Adventure Island." Assuming that the statistics show "Fantasy Castle" has the highest starting tendency value, followed by "Future World," and then "Adventure Island," the route recommendation module will generate three recommended routes starting from the L1 interest nodes with the highest starting tendency values ​​(e.g., L1=3). For example, the first route might start from "Fantasy Castle," pass through "Dream Theater," and finally reach the exit; the second route might start from "Future World," pass through "Science Museum," and finally reach the exit; and the third route might start from "Adventure Island," pass through "Jungle Adventure," and finally reach the exit. These routes can be generated based on the simple principle of shortest distance.

[0050] The conflict analysis module analyzes the conflict level of all recommended routes within the park. It sorts the recommended routes in chronological order of generation time to obtain a generation sequence, generates a traversal window, and sets the traversal value. The module then iterates through and extracts interest nodes from all recommended routes. The traversal value is marked as k, and interest nodes with indices 1-k from all recommended routes are extracted and grouped into a conflict analysis set. Conflict evaluation analysis is performed on this set. Then, interest nodes with indices 2-k+1 from all recommended routes are extracted and grouped into a new conflict analysis set. This process is repeated until all interest nodes from all recommended routes have been added to the conflict analysis set and undergone conflict evaluation analysis. Finally, conflict evaluation analysis is performed on the entire conflict analysis set. The specific process includes: marking the number of elements in the conflict analysis set corresponding to the interest node as the conflict value of the interest node; retrieving the conflict threshold and concentration threshold from the database; marking interest nodes with conflict values ​​not less than the conflict threshold as conflict nodes; marking recommended routes where the sequence number of the conflict node in the recommended route conforms to the truncation rule as the conflict route of the conflict node; calculating the variance of the node sequence number of the same conflict node in all conflict routes to obtain the concentration coefficient of the conflict node; comparing the concentration coefficient with the concentration threshold: if the concentration coefficient is greater than the concentration threshold, it is determined that the conflict node does not have a moderating characteristic; if the concentration coefficient is less than or equal to the concentration threshold, it is determined that the conflict node has a moderating characteristic; inserting buffer nodes before the last k / 2 conflict nodes in the conflict route, the buffer nodes include small viewpoints, rest halls, and restaurants.

[0051] The traversal value is denoted as k, which defines the fixed length of the sequence of interest nodes used for analysis, i.e., the size of the sliding window. This traversal value k can be a preset fixed integer, such as 3, 5, or 7, or it can be dynamically adjusted based on the park's size, historical visitor traffic data, or real-time congestion conditions. Its purpose is to ensure that conflict analysis focuses on a specific length of interest node sequence, avoiding evaluation bias caused by inconsistent analysis window sizes.

[0052] The process involves extracting interest nodes with indices ranging from 1 to k from the beginning of each recommended route and creating a conflict analysis set. This step aims to extract the first sequence of interest nodes of length k from the start of each recommended route. This can be achieved by iterating through each recommended route and extracting its first k interest nodes. These extracted k interest node sequences are then grouped into a set and used as input for the initial conflict assessment.

[0053] A conflict assessment analysis is performed on the conflict analysis set. This step involves detecting conflicts among all k interest node sequences in the current conflict analysis set. Specifically, it compares the k interest node sequences of different recommended routes in the set to identify potential conflict points by checking for common interest nodes or similar node orders. For example, the number of common interest nodes in each sequence can be calculated, or a predefined conflict algorithm can be used to assess congestion risk.

[0054] Then, interest nodes with indices matching 2-k+1 are extracted from all recommended routes and grouped into a new conflict analysis set. This step represents sliding the analysis window forward by one interest node. For example, if the previous window extracted interest nodes 1 to k, the current window will extract interest nodes 2 to k+1. This can be achieved by updating the starting index of each recommended route and re-extracting k interest nodes.

[0055] The conflict value of an interest node is determined by marking the number of elements within its conflict analysis set. This quantifies the degree to which a specific interest node is accessed by multiple recommended routes within the current analysis window, thereby identifying potential congestion points. The conflict value can be understood as the "access frequency" of the interest node within the conflict analysis set. For example, it can be obtained by iterating through each recommended route in the conflict analysis set; if the route contains the target interest node, the counter for that interest node is incremented, and the final count is the conflict value. Alternatively, an association matrix between interest nodes and routes can be constructed, and the conflict value can be obtained by summing the columns corresponding to the interest nodes in the matrix.

[0056] The profile analysis module is used to analyze visitor behavior profiles: behavior profiles include ticket booking profiles and tour profiles. Ticket booking profile features include solo tours, duo tours, and group tours; tour profile features include light tours and in-depth tours; ticket booking profile features are obtained directly by retrieving visitor ticket booking information; the labeling process for tour profile features includes: after visitors enter the park, the average tour duration of visitors at all interest nodes (excluding entrances and exits) is calculated and labeled as an immersion value; an immersion threshold is retrieved from the database, and the immersion value is compared with the immersion threshold: if the immersion value is less than the immersion threshold, the corresponding visitor's tour profile feature is labeled as a light tour; if the immersion value is greater than or equal to the immersion threshold, the corresponding visitor's tour profile feature is labeled as an in-depth tour.

[0057] The booking profile features refer to the behavioral patterns or attributes exhibited by tourists when booking tickets or services, such as solo travel, duo travel, and group travel. Obtaining these features provides the foundational data for tourist behavior profiling, reflecting tourists' initial intentions and group attributes. In practice, the system can interface with ticketing systems (such as online ticketing platforms and official scenic area apps) via API to obtain tourists' booking records in real time or periodically. Furthermore, when tourists use the smart guide system for the first time, relevant information can be extracted from their purchase vouchers (such as e-ticket QR codes and order numbers) through authorization.

[0058] The average visit duration of visitors at all non-entrance / exit interest nodes is statistically analyzed and denoted as the immersion value. The immersion value is an indicator that quantifies the time visitors spend at interest nodes within the park, reflecting their level of interest and depth of exploration in a specific area. This step aims to objectively measure visitors' actual visit behavior, providing a quantitative basis for distinguishing between light and in-depth visits. Specifically, IoT positioning systems (such as Bluetooth beacons and Wi-Fi probes) can be used to continuously track visitors' dwell time at each interest node and calculate its average. Alternatively, by combining gate access records and entrance / exit sensor data for each interest node, the entry and exit times of visitors at each interest node can be inferred, and then the dwell time can be calculated and averaged.

[0059] The scheduling management module is used for route recommendations for tourists and resource scheduling management and analysis within the park: At the beginning of the statistical period, it retrieves the selection frequency of all recommended routes corresponding to different booking profile characteristics of tourists in the previous statistical period and marks them as the selection probability value of the recommended routes corresponding to the booking profile characteristics. When recommending routes within the statistical period, it obtains the booking profile characteristics of tourists, retrieves the selection probability values ​​of all recommended routes corresponding to the booking profile characteristics, and generates a recommendation sequence in descending order of selection probability values. The recommendation sequence is then sent to the optimized mobile terminal. Within the statistical period, it periodically obtains the number of tourists with in-depth tour profile characteristics in each non-exit interest node and marks them as in-depth density values. The in-depth density value is compared with a preset in-depth density threshold: if the in-depth density value is less than the in-depth density threshold, no processing is performed; if the in-depth density value is greater than or equal to the in-depth density threshold, the corresponding interest node is marked as an in-depth node, and scheduling information is issued to the guides near the in-depth node.

[0060] The statistical period refers to the preset time frame used for data collection and analysis. Its length can be set according to actual operational needs, such as a calendar day, a week, or a month, or a dynamically adjusted time window based on peak and off-peak periods. Retrieving the selection frequency of all recommended routes for tourists with different booking profiles within the previous statistical period refers to the number or proportion of times a specific recommended route was selected by tourists with specific booking profiles, retrieved from historical data. This can be achieved by querying historical tourist behavior logs stored in the database; for example, recording the actual routes selected by tourists after each recommendation and associating them with their booking profile characteristics.

[0061] Selection frequency can be a simple count of selections or a weighted statistical value. Labeling selection frequency as the probability value of a recommended route corresponding to a booking profile feature means converting the statistically obtained selection frequency into a probability value to quantify the likelihood of a tourist with a specific booking profile feature choosing a particular recommended route. This can be calculated by dividing the selection frequency of a route by the total number of selections by tourists with that booking profile feature, or by using a more complex statistical model (such as smoothed frequency). Obtaining the tourist's booking profile features means that when route recommendations are needed, the system identifies the characteristics reflected in the current tourist's booking information, such as whether they are traveling alone, in pairs, or in groups. This is usually obtained directly by reading the tourist's electronic ticketing information or their registration information in the system. Retrieving the selection probability values ​​of all recommended routes corresponding to the booking profile feature means that the system, based on the currently identified booking profile feature of the tourist, filters out the probability values ​​of all recommended routes associated with that feature from pre-calculated and stored probability values. This usually involves a database query operation to obtain the probability data of all routes under a specific profile feature. Generating a recommended sequence by ordering the selection probability values ​​from highest to lowest involves sorting the retrieved recommended routes and their corresponding selection probability values, placing the route with the highest probability value at the top, and so on, forming an ordered recommended list. This can be achieved using standard sorting algorithms. Sending the recommended sequence to optimized mobile terminals means transmitting the generated list of recommended routes over the network to the mobile devices used by tourists. These mobile devices typically run dedicated navigation applications that can receive, parse, and display the recommended information in a user-friendly way, such as highlighting recommended routes on a map and providing navigation guidance.

[0062] Example 2: Figure 2 As shown, the intelligent tour guide scheduling method based on the Internet of Things and tourist behavior profiles includes the following steps:

[0063] Step 1: Analyze recommended tour routes in the park: Generate statistical periods, mark the park's exits, entrances, and exhibition halls as interest nodes, mark the number of times an exhibition hall is marked as an interest node as a starting node as the initial tendency value, and generate L1 recommended routes starting from the L1 interest nodes with the largest initial tendency values.

[0064] Step 2: Analyze the conflict level of all recommended routes in the park: Sort the recommended routes in order of generation time to obtain the generation sequence, generate a traversal window and set the traversal value of the traversal window, and traverse and extract the interest nodes of all recommended routes through the traversal window.

[0065] Step 3: Analyze tourist behavior profiles: Behavioral profiles include ticket booking profiles and tour profiles. Ticket booking profiles include characteristics of solo tours, duo tours, and group tours; tour profiles include characteristics of light tours and in-depth tours.

[0066] Step 4: Analyze tourist routes and resource allocation management within the park by analyzing recommended routes, ticket booking profiles, and visitor profiles.

[0067] The intelligent tour guide and scheduling system based on the Internet of Things and visitor behavior profiles generates statistical cycles during operation. It marks park exits, entrances, and exhibition halls as nodes of interest, and the number of times an exhibition hall is marked as a starting node is recorded as an initial tendency value. Starting from the L1 nodes with the highest initial tendency values, it generates L1 recommended routes. The recommended routes are sorted in chronological order of generation time to obtain a generation sequence. A traversal window is generated, and its traversal value is set. The system then traverses and extracts nodes of interest from all recommended routes. Behavioral profiles include booking profiles and tour profiles. Booking profile features include solo, duo, and group tours; tour profile features include light and in-depth tours. Based on recommended routes, booking profile features, and tour profile features, the system provides route recommendations and analyzes resource scheduling and management within the park.

[0068] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

[0069] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0070] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. An intelligent tour guide and dispatch system based on the Internet of Things and tourist behavior profiling, characterized in that, It includes a route recommendation module, a conflict analysis module, a profile analysis module, and a scheduling management module that are connected in sequence. The route recommendation module, conflict analysis module, and profile analysis module are all connected in communication with the database. The route recommendation module is used to recommend and analyze the tour routes in the park: generate statistical periods, mark the park's exits, entrances, and exhibition halls as interest nodes, mark the first non-entrance interest node visited by tourists after entering the park as the starting node, mark the number of times the exhibition hall is marked as the starting node as an interest node as the starting value, and generate L1 recommended routes starting from the L1 interest nodes with the largest starting value. The conflict analysis module is used to analyze the conflict level of all recommended routes in the park: the recommended routes are sorted in order of generation time to obtain the generation sequence, a traversal window is generated and the traversal value of the traversal window is set, and the interest nodes of all recommended routes are traversed and the conflict assessment is performed through the traversal window. The profile analysis module is used to analyze tourist behavior profiles: the behavior profiles include ticket booking profiles and tour profiles. The ticket booking profile features include solo tours, duo tours, and group tours; the tour profile features include light tours and in-depth tours. The scheduling and management module is used to recommend routes to tourists and to manage and analyze resource scheduling within the park.

2. The intelligent tour guide and scheduling system based on the Internet of Things and tourist behavior profiling as described in claim 1, characterized in that, The process of generating recommended routes includes: marking the interest node with the largest initial tendency value as the priority node; marking the first visitor whose non-entry interest node is the priority node as the priority visitor; marking the number of times the priority visitor enters other interest nodes after leaving the priority node as the follow value of the corresponding interest node; and marking the follow nodes of the recommended route based on the follow value; then marking the interest node with the second largest initial tendency value as the priority node and generating the corresponding recommended route, and so on, until the number of recommended routes reaches L1.

3. The intelligent tour guide and scheduling system based on the Internet of Things and tourist behavior profiling as described in claim 2, characterized in that, The specific process of marking follower nodes for the recommended route includes: marking the interest node with the largest follower value as a follower node; retrieving the follower threshold from the database; comparing the follower value of the follower node with the follower threshold; if the follower value is less than the follower threshold, the recommended route is terminated; if the follower value is greater than or equal to the follower value, it is determined whether the follower node is an exit: if yes, the recommended route is terminated; if no, the follower node is added to the recommended route, and the marking of follower nodes continues until the recommended route is terminated.

4. The intelligent tour guide and scheduling system based on the Internet of Things and tourist behavior profiling as described in claim 3, characterized in that, The specific process of traversing and extracting interest nodes for all recommended routes includes: marking the traversal value as k, extracting interest nodes with indices of 1-k from all recommended routes and forming a conflict analysis set, and performing conflict evaluation analysis on the conflict analysis set; then extracting interest nodes with indices of 2-k+1 from all recommended routes and forming a new conflict analysis set, and performing conflict evaluation analysis on the new conflict analysis set; and so on, until all interest nodes in all recommended routes have been added to the conflict analysis set and subjected to conflict evaluation analysis.

5. The intelligent tour guide and scheduling system based on the Internet of Things and tourist behavior profiling as described in claim 4, characterized in that, The specific process of conducting conflict assessment analysis on the conflict analysis set includes: marking the number of elements in the conflict analysis set corresponding to the interest node as the conflict value of the interest node; retrieving the conflict threshold and concentration threshold from the database; marking interest nodes with conflict values ​​not less than the conflict threshold as conflict nodes; marking recommended routes where the sequence number of the conflict node in the recommended route conforms to the truncation rule as the conflict route of the conflict node; calculating the variance of the node sequence number of the same conflict node in all conflict routes to obtain the concentration coefficient of the conflict node; comparing the concentration coefficient with the concentration threshold: if the concentration coefficient is greater than the concentration threshold, it is determined that the conflict node does not have a moderating characteristic; if the concentration coefficient is less than or equal to the concentration threshold, it is determined that the conflict node has a moderating characteristic; and inserting a buffer node before the last k / 2 conflict nodes in the conflict route.

6. The intelligent tour guide and scheduling system based on the Internet of Things and tourist behavior profiling as described in claim 5, characterized in that, Ticket booking profile features are obtained directly by retrieving tourists' ticket booking information; the labeling process for tour profile features includes: after tourists enter the park, the average tour duration of tourists at all interest nodes other than entrances and exits is counted and labeled as immersion value. Immersion threshold is retrieved from the database and compared with immersion value: if immersion value is less than immersion threshold, the corresponding tourist's tour profile feature is labeled as light tour; if immersion value is greater than or equal to immersion threshold, the corresponding tourist's tour profile feature is labeled as deep tour.

7. The intelligent tour guide and scheduling system based on the Internet of Things and tourist behavior profiling as described in claim 6, characterized in that, The specific process of the scheduling management module in analyzing routes for tourists includes: at the beginning of the statistical period, retrieving the selection frequency of all recommended routes corresponding to different booking profile features in the previous statistical period and marking it as the selection probability value of the recommended route corresponding to the booking profile feature; when recommending routes within the statistical period, obtaining the booking profile features of tourists, retrieving the selection probability values ​​of all recommended routes corresponding to the booking profile features, generating a recommendation sequence in descending order of selection probability value, and sending the recommendation sequence to the optimized mobile terminal.

8. The intelligent tour guide and scheduling system based on the Internet of Things and tourist behavior profiling as described in claim 7, characterized in that, The specific process of the scheduling management module in analyzing resource scheduling management within the park includes: periodically acquiring the number of tourists with in-depth tour profiles at each non-exit interest node within the statistical period and marking them as in-depth density values; comparing the in-depth density values ​​with preset in-depth density thresholds: if the in-depth density value is less than the in-depth density threshold, no processing is performed; if the in-depth density value is greater than or equal to the in-depth density threshold, the corresponding interest node is marked as an in-depth node, and scheduling information is released to the guides near the in-depth node.

Images