Personalized customized big data smart tourism system

Abstract

The invention discloses a personalized customized big data smart tourism system, and the system comprises a user login module which comprises a data collection and input unit and a user tag library; the tourism data statistical analysis module comprises an information calling system and a tourism information base; the information calling system is used for collecting information of scenic spots, hotels and traffic and storing the information into a tourism information base; the query display module is used for a user to query and display information of corresponding scenic spots, hotels and traffic according to query content; the tourism scheme planning module is used for customizing a tourism scheme for the user and comprises a tourist data analysis unit and a scheme customizing and pushing unit; the tourism data recommendation module comprises an intelligent recommendation system and a real-time dynamic recommendation unit; the user interaction module comprises a tourist evaluation and feedback system and a social sharing function unit; by analyzing the crowd portrait and screening the tourism data, the personalized customized tourism scheme meeting the requirements of the user is accurately generated, and the tourism experience and the tourism efficiency of the user are improved.

Description

Technical Field

[0001] The present invention relates to a big data smart tourism system, in particular to a personalized big data smart tourism system, and belongs to the technical fields of tourism information technology, big data analysis and route planning. Background Art

[0002] With the rapid development of the economy and the improvement of people's living standards, tourism has gradually become an important part of people's daily lives. Tourism is not only a simple process of traveling from one place to another, but also a comprehensive activity including outings, sightseeing, and entertainment. With the end of the epidemic and the accelerated pace of modern people's lives, the tourism industry has ushered in a blowout development, which not only promotes the development of the urban economy, but also meets people's growing demand for a better life.

[0003] However, there are still many problems with the current travel experience. Most tourists tend to follow the trend when choosing a travel destination and tend to choose more famous tourist attractions or Internet celebrity attractions, resulting in a single choice of tourist attractions. Especially during holidays, tourists flock in, causing overcrowding at the attractions, affecting the overall travel experience. In addition, the current travel methods are mainly self-driving and group tours, but these two methods also have their own shortcomings: group tours have compact itineraries, strong constraints, and lack of in-depth experience; although self-driving tours are flexible and free, they are often difficult to make reasonable travel plans due to the limited scenic spot information provided by the Internet.

[0004] In the existing technology, the matching method of travel guides also has significant deficiencies. The existing system usually relies on users to input the expected travel destination, and then displays several guides for users to choose from. However, these guides often lack relevance to the personalized needs of users, and users need to filter them by themselves, which easily leads to boredom and reduces the efficiency of use. In addition, most of the existing travel guides are based on the subjective sharing of other users, lacking systematic and standardized analysis, resulting in users being unable to obtain comprehensive reference information in actual travel, further affecting their travel experience.

[0005] In order to solve the above problems, it is necessary to design a travel guide intelligent matching generation method based on user behavior analysis. By analyzing crowd portraits and screening travel data, personalized customized travel plans that meet user needs can be accurately generated to improve users' travel experience and travel efficiency. Summary of the invention

[0006] Purpose of the invention: The purpose of the invention is to provide a personalized big data smart tourism system that can improve the rationality of tourism customization plans.

[0007] Technical solution: The personalized big data smart tourism system described in the present invention includes:

[0008] The user login module is used to provide users with registration and login functions to protect user information security; it also includes a data collection and entry unit and a user tag library; the data collection and entry unit is used to collect the user's basic information, travel preference information and historical travel data, generate the user's user tag, and store it in the user tag library; the travel preference information includes travel type, budget range, travel time, length of stay and transportation preference; the user tag library is used to store user tags;

[0009] The tourism data statistical analysis module includes an information calling system and a tourism information database; the information calling system is used to collect information on scenic spots, hotels and transportation, and store it in the tourism information database;

[0010] The query display module is used for user query and displays the corresponding scenic spots, hotels and transportation information according to the query content;

[0011] A travel plan planning module is used to customize travel plans for users, including a tourist data analysis unit and a plan customization push unit;

[0012] Tourism data recommendation module, including intelligent recommendation system and real-time dynamic recommendation unit;

[0013] User interaction module, including visitor evaluation and feedback system and social sharing function unit.

[0014] Furthermore, the specific implementation method of the information calling system of the tourism data statistical analysis module is to obtain data from major attractions, hotels and transportation platforms in real time, perform data cleaning and preprocessing, integrate and classify the processed information based on big data and artificial intelligence algorithms, and continuously update and maintain the information stored in the tourism information database in real time through a scheduled API calling mechanism.

[0015] Furthermore, the tourist data analysis unit of the tourism program planning module is specifically implemented as follows:

[0016] (1) Calculate the user's preference weight for each tourism type. Specifically, the attribute data of users and attractions are vectorized and expressed as feature vectors. Suppose user U i The eigenvector of U is i =[u 1 ,u 2 ,...,u n ], Attractions S j The eigenvector of j =[s 1 ,s 2 ,...,s m]; Based on the user's historical travel data and system browsing data, use weighted average or machine learning model to calculate the user's preference weight for different travel types. i For travel type T k The preference weight w ik :

[0017]

[0018] Among them, λ is the balance coefficient, which indicates the influence of user tags and access behaviors on preference weights, α iL Indicates the user label L for user U i The importance weight, M kL For label L and travel type T k Each element M of the matching matrix kL , indicating the tourism type T corresponding to the label L k The degree of adaptation, the value is between 0 and 1, 1 means complete match, 0 means no match at all, f ij Is user U i For attractions j The number of visits or tags, t ij Is user U i Attractions S j The visit duration, S k Is of type T k If the user has not searched or visited a certain type of attraction, the behavior data of the attraction will not be considered. i j =0 or t ij =0;

[0019] (2) Matching the user's preference weight and the scenic spot feature vector is performed, and the matching recommendation coefficient between each user and the scenic spot is calculated, and the recommendation is completed through screening and optimization;

[0020] Specifically, let the user and the scenic spot match the recommendation coefficient M ij Indicates that through user U i The preference weight of attraction S j The weighted cosine similarity of the feature vector is expressed as:

[0021]

[0022] (3) Based on the matching recommendation coefficient M ij , combined with the geographical location and influencing factors of the scenic spot, calculate the comprehensive matching coefficient C of the scenic spot ij :

[0023]

[0024] dj It is a tourist attraction j The distance from the user's destination, D max is the maximum permissible distance, O j It is a tourist attraction j The impact factor, T arr is the user arrival time, α, β and γ are the weight coefficients of each factor;

[0025] According to the comprehensive matching coefficient C ij Sort all attractions and generate a recommended list of attractions that meets user preferences.

[0026] Furthermore, the specific implementation method of the plan customization push unit of the tourism plan planning module is to determine the optimal travel route through the traveling salesman problem TSP method, and define each scenic spot, hotel and transportation hub in the results of the tourist data analysis unit as a node in the TSP problem; according to the actual map data or online API, calculate the travel distance or time between each node, construct a distance matrix, solve it with minimizing the total travel distance or time as the optimization goal, and push the plan to the user.

[0027] Furthermore, the intelligent recommendation system of the tourism data recommendation module is used to recommend attractions, routes, dining and shopping information to users based on a tourism information database through big data and artificial intelligence technology.

[0028] Furthermore, the real-time dynamic recommendation unit of the travel data recommendation module is used to dynamically recommend surrounding attractions, restaurants and shopping information based on the user's location and travel preference information.

[0029] Furthermore, the tourist evaluation and feedback system of the user interaction module is used for users to evaluate tourist attractions.

[0030] Furthermore, it is also used for users to evaluate and provide feedback on the travel plans customized by the travel plan planning module. When the feedback result is unsatisfactory, it returns to the travel plan planning module, updates the user tag library and evaluation parameters, and re-customizes the travel plan.

[0031] Furthermore, the social sharing function unit of the user interaction module is used for users to share travel plans and travel experiences on social platforms for interaction, and to collect corresponding travel preference information and historical travel data, and update them to the user tag library.

[0032] Furthermore, the user tag library also includes updates based on the interaction results of the artificial intelligence clustering algorithm and the user interaction module.

[0033] Beneficial effects: Compared with the prior art, the present invention adopts multi-level feature analysis. First, preliminary crowd labels are generated according to user registration information, and then labels are dynamically updated according to user historical information, browsing information, etc. Finally, label information is updated again through user social sharing feedback. In this process, artificial intelligence algorithms such as machine learning are used to not only extract and analyze basic primary information features, but also generate more complex and multi-dimensional secondary information features, improve the accuracy and robustness of the model, and truly realize accurate recommendation and personalized customization of travel plans; this system integrates and updates the latest data on tourist attractions, transportation and accommodation in real time to ensure that users obtain the most timely and accurate information and enhance the travel experience. The system innovatively uses the solution of the Traveling Salesman Problem (TSP) to plan travel routes. First, it combines the K-means clustering algorithm to integrate the travel information required by users, and then quickly obtains the optimal travel route through the simulated annealing algorithm. Through this method, the system can provide efficient and reasonable travel route planning based on user preferences and characteristics of tourist attractions. Relying on the efficient parallel computing and fast training algorithms of this system, users can provide real-time feedback on tourist attractions and routes, and the system can quickly adjust travel strategies based on these data to achieve dynamic adjustment of travel plans. Through the above means, this system can truly provide personalized services for tourists, significantly improving users' travel experience and the overall performance of the system. These beneficial effects not only improve the practicality and reliability of the system, but also provide new ideas and directions for the development of the smart tourism industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a system framework diagram of an embodiment of the present invention;

[0035] Figure 2 A schematic diagram of a method for establishing a user tag library according to an embodiment of the present invention;

[0036] Figure 3 A schematic diagram of a method for establishing a tourism information database according to an embodiment of the present invention;

[0037] Figure 4 Schematic diagram of the system operation mode of an embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0039] As attached Figure 1 As shown, the personalized big data smart tourism system of this embodiment includes:

[0040] The user login module is used to provide users with registration and login functions to protect user information security; it also includes a data collection and entry unit and a user tag library; the data collection and entry unit is used to collect the user's basic information, travel preference information and historical travel data, generate the user's user tag, and store it in the user tag library; the travel preference information includes travel type, budget range, travel time, length of stay and transportation preference; the user tag library is used to store user tags;

[0041] The tourism data statistical analysis module includes an information calling system and a tourism information database; the information calling system is used to collect information on scenic spots, hotels and transportation, and store it in the tourism information database;

[0042] The query display module is used for user query and displays the corresponding scenic spots, hotels and transportation information according to the query content;

[0043] A travel plan planning module is used to customize travel plans for users, including a tourist data analysis unit and a plan customization push unit;

[0044] Tourism data recommendation module, including intelligent recommendation system and real-time dynamic recommendation unit;

[0045] User interaction module, including visitor evaluation and feedback system and social sharing function unit.

[0046] Specifically, the user login module collects basic information entered by the user during registration, such as name, age, gender, contact information, etc. It also collects information such as travel preferences filled in by the user, including travel type (such as natural scenery, history and culture, city sightseeing, etc.), budget range, travel time, length of stay, transportation preference, etc. The system initially creates user tags based on the information filled in by the user, and stores the tags in the user tag library. The tag library establishment diagram is shown in the figure below. Figure 2 shown.

[0047] The specific implementation method of the information calling system of the tourism data statistical analysis module is to obtain data from major attractions, hotels and transportation platforms in real time, perform data cleaning and preprocessing, integrate and classify the processed information based on big data and artificial intelligence algorithms, and continuously update and maintain the information stored in the tourism information database in real time through a timed API calling mechanism. The construction method of the tourism information database is as follows Figure 3 The information of the scenic spot includes geographical location, attraction rating, popularity, opening hours, facilities and services, etc.

[0048] The tourist data analysis unit of the travel program planning module of this embodiment vectorizes the attribute data of users and attractions and expresses them as feature vectors. i The eigenvector of U is i =[u 1,u 2 ,...,u n ], Attractions S j The eigenvector of j =[s 1 ,s 2 ,...,s m ]. Based on the user's historical travel data, a weighted average or machine learning model is used to calculate the user's preference weights for different travel types.

[0049] According to the matching degree between the user's tag and the travel type, the user U i For travel type T k Preference weights:

[0050]

[0051] Among them, λ is the balance coefficient, which indicates the influence of user tags and access behaviors on preference weights. Here, λ is determined by the reinforcement learning method in machine learning, and is adjusted dynamically in real time based on user tag data and browsing behaviors. is the label part, α iL Indicates the user label L for user U i The importance weight, M kL For label L and travel type T k Each element M of the matching matrix kL , indicating the tourism type T corresponding to the label L k The degree of adaptation, the value is between 0 and 1, 1 means complete match, 0 means no match at all, here M kL The acquisition method also relies on generating an initial matching matrix based on the labels, and using a gradient descent algorithm to update user preference weights and matching degrees based on user behavior feedback. is the user behavior part, f ij Is user U i For attractions j The number of visits or marks reflects the user's interest in the attraction. ij Is user U i Attractions S j The length of stay reflects the depth of user participation. k Is of type T k The formula combines the user’s tag information and actual visit behavior to calculate the preference weight. If the user has not searched or visited a certain type of attraction, the behavior data of the attraction will not be considered, that is, f i j =0 or t ij= 0. This design can not only make accurate recommendations when users search for attractions, but also make accurate recommendations based on their preferences when users have no idea of ​​a travel destination, making it more personalized.

[0052] Then, the user's intention information is matched with the scenic spot information: the system matches the user's preference information with the characteristics of the scenic spot, calculates the matching recommendation coefficient of each scenic spot with the user's preference, and further screens and optimizes the recommended content. According to the matching degree, the priority of the recommendation list is adjusted, and personalized travel suggestions are provided to users.

[0053] In this embodiment, the user and scenic spot matching recommendation coefficient M ij It can be represented as user U i The preference vector and the attraction S j The weighted cosine similarity of the feature vectors

[0054]

[0055] In this embodiment, the system further combines the actual factors such as the geographical location and opening hours of the scenic spot to calculate the comprehensive matching coefficient C of the scenic spot. ij Combined with the matching recommendation coefficient M ij and the actual factors of the attractions,

[0056]

[0057] d j It is a tourist attraction j The distance from the user's destination, D max is the maximum permissible distance O j It is a tourist attraction j factors such as opening time or rating level, T arr is the user's arrival time, α, β, and γ are the weight coefficients of each factor, which are adjusted based on historical data or user preferences in the system. ij Sort all attractions and generate a recommended list of tourist attractions that meet user preferences.

[0058] Furthermore, the customized travel plan push unit of the travel plan planning module obtains the customized travel plan and pushes the customized travel plan to the customer: the system generates a complete customized travel plan through the TSP method, including detailed itinerary arrangements, recommended attractions, accommodation arrangements, transportation plans, etc. The system pushes the customized plan to the user through in-application notifications, emails or text messages. The system can generate an efficient and reasonable travel route for the user, minimizing travel time and costs while ensuring the richness of the itinerary and the satisfaction of the user experience. This plan is pushed to the user through in-application notifications, emails or text messages, ensuring that the user can obtain an accurate and customized travel plan before departure.

[0059] In this embodiment, in order to optimize the itinerary and transportation plan, the traveling salesman problem (TSP) method can be applied to determine the optimal travel path. Each tourist attraction, hotel or transportation hub is defined as a node in the TSP problem. According to the actual map data or online API (such as Gaode Map or Baidu Map), the travel distance or time between each node is calculated to construct a complete distance matrix. The optimization goal is to minimize the total travel distance or time, ensuring that users cover more attractions within a limited time and reduce unnecessary time waste.

[0060] Furthermore, the intelligent recommendation system of the tourism data recommendation module is used to recommend attractions, routes, dining and shopping information to users based on a tourism information database through big data and artificial intelligence technology.

[0061] Furthermore, the real-time dynamic recommendation unit of the travel data recommendation module is used to dynamically recommend surrounding attractions, restaurants and shopping information based on the user's location and travel preference information.

[0062] Furthermore, the tourist evaluation and feedback system of the user interaction module is used for users to evaluate tourist attractions.

[0063] It is also used for users to evaluate and give feedback on the travel plans customized by the travel plan planning module. Users can view the received travel customized plans and make evaluations. Users can directly accept the recommended plan, or make modification suggestions or feedback according to their own needs. When the feedback result is unsatisfactory, it returns to the travel plan planning module, updates the user label and evaluation parameters, and customizes the travel plan again. The overall operation mode of this embodiment is as follows: Figure 4 shown.

[0064] Furthermore, the social sharing function unit of the user interaction module is used for users to share travel plans and travel experiences on social platforms for interaction, and to collect corresponding travel preference information and historical travel data, and update them to the user tag library.

[0065] Furthermore, the user tag library also includes updates based on the interaction results of the artificial intelligence clustering algorithm and the user interaction module.

[0066] The above disclosure is only a few specific embodiments of the present invention, and those skilled in the art can make various modifications and equivalent substitutions to the present invention. Any modification, equivalent substitution, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A personalized big data smart tourism system, characterized by: include: The user login module is used to provide users with registration and login functions to protect user information security; it also includes a data collection and entry unit and a user tag library; the data collection and entry unit is used to collect the user's basic information, travel preference information and historical travel data, generate the user's user tag, and store it in the user tag library; the travel preference information includes travel type, budget range, travel time, length of stay and transportation preference; the user tag library is used to store user tags; The tourism data statistical analysis module includes an information calling system and a tourism information database; the information calling system is used to collect information on scenic spots, hotels and transportation, and store it in the tourism information database; The query display module is used for user query and displays the corresponding scenic spots, hotels and transportation information according to the query content; A travel plan planning module is used to customize travel plans for users, including a tourist data analysis unit and a plan customization push unit; Tourism data recommendation module, including intelligent recommendation system and real-time dynamic recommendation unit; User interaction module, including visitor evaluation and feedback system and social sharing function unit.

2. The personalized big data smart tourism system according to claim 1 is characterized in that: The specific implementation method of the information calling system of the tourism data statistical analysis module is to obtain data from major attractions, hotels and transportation platforms in real time, perform data cleaning and preprocessing, integrate and classify the processed information based on big data and artificial intelligence algorithms, and continuously update and maintain the information stored in the tourism information database in real time through a scheduled API calling mechanism.

3. The personalized big data smart tourism system according to claim 1 is characterized in that: The specific implementation method of the tourist data analysis unit of the tourism program planning module is: (1) Calculate the user's preference weight for each tourism type. Specifically, the attribute data of users and attractions are vectorized and expressed as feature vectors. Suppose user U i The eigenvector of U is i =[u1,u2,...,u n ], Attractions S j The eigenvector of j =[s1,s2,...,s m ]; Based on the user's historical travel data and system browsing data, use weighted average or machine learning model to calculate the user's preference weight for different travel types. i For travel type T k The preference weight w ik : Among them, λ is the balance coefficient, which indicates the influence of user tags and access behaviors on preference weights, α iL Indicates the user label L for user U i The importance weight, M kL For label L and travel type T k Each element of the matching matrix represents the tourism type T corresponding to the label L k The degree of adaptation, the value is between 0 and 1, 1 means complete match, 0 means no match at all, f ij Is user U i For attractions j The number of visits or tags, t ij Is user U i Attractions S j The visit duration, S k Is of type T k If the user has not searched or visited a certain type of attraction, the behavior data of the attraction will not be considered. ij =0 or t ij =0; (2) Matching is performed based on the user's preference weight and the scenic spot feature vector, calculating the matching recommendation coefficient between each user and the scenic spot, and completing the recommendation through screening and optimization; Specifically, let the user and the scenic spot match the recommendation coefficient M ij Indicates that through user U i The preference weight of attraction S j The weighted cosine similarity of the feature vector is expressed as: (3) Based on the matching recommendation coefficient M ij , combined with the geographical location and influencing factors of the scenic spot, calculate the comprehensive matching coefficient C of the scenic spot ij : d j It is a tourist attraction j The distance from the user's destination, D max is the maximum permissible distance, O j It is a tourist attraction j The impact factor, T arr is the user arrival time, α, β and γ are the weight coefficients of each factor; According to the comprehensive matching coefficient C ij Sort all attractions and generate a recommended list of attractions that meets user preferences.

4. The personalized big data smart tourism system according to claim 1 is characterized in that: The specific implementation method of the scheme customization push unit of the tourism scheme planning module is to determine the optimal tourism route through the traveling salesman problem TSP method, and define each scenic spot, hotel and transportation hub in the result of the tourist data analysis unit as a node in the TSP problem; According to the actual map data or online API, the travel distance or time between each node is calculated, and the distance matrix is ​​constructed. The solution is solved with the optimization goal of minimizing the total travel distance or time, and the solution is pushed to the user.

5. The personalized big data smart tourism system according to claim 1 is characterized in that: The intelligent recommendation system of the tourism data recommendation module is used to recommend attractions, routes, catering and shopping information to users based on the tourism information database through big data and artificial intelligence technology.

6. The personalized big data smart tourism system according to claim 1 is characterized in that: The real-time dynamic recommendation unit of the travel data recommendation module is used to dynamically recommend surrounding attractions, restaurants and shopping information based on the user's location and travel preference information.

7. The personalized big data smart tourism system according to claim 1 is characterized in that: The visitor evaluation and feedback system of the user interaction module is used for users to evaluate tourist attractions.

8. The personalized big data smart tourism system according to claim 1 is characterized in that: The tourist evaluation and feedback system of the user interaction module is also used for users to evaluate and provide feedback on the travel plans customized by the travel plan planning module. When the feedback result is unsatisfactory, it returns to the travel plan planning module, updates the user tag library and evaluation parameters, and re-customizes the travel plan.

9. The personalized big data smart tourism system according to claim 1 is characterized in that: The social sharing functional unit of the user interaction module is used for users to share travel plans and travel experiences on social platforms for interaction, and to collect corresponding travel preference information and historical travel data, and update them to the user tag library.

10. The personalized big data smart tourism system according to claim 1 is characterized in that: The user tag library also includes updating according to the interaction results of the artificial intelligence clustering algorithm and the user interaction module.

Images