Intelligent matching tourist destination recommendation method
Through intelligent matching methods, text segmentation, AI semantic modules and collaborative filtering algorithms are used to build a tourist city scenic spot library, solving the problem that traditional tourism websites are difficult to match user needs, realizing personalized and accurate tourism destination recommendations, and enhancing users Experience and satisfaction.
Patent Information
- Application Number
- CN202411006103.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-25
- Publication Date
- 2024-11-19
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The data queries provided by traditional tourism websites are huge and difficult to match and recommend according to user needs, which reduces the quality of tourism services and traveler experience.
Using intelligent matching tourism destination recommendation method, text segmentation and keyword extraction are performed on the information input by users, combined with national urban yearbook data, meteorological information and geographical location, a tourist city scenic spot library is built, and AI semantic modules and collaborative filtering algorithms are used. Perform personalized recommendations and generate radar maps for city screening and sorting.
It realizes accurate recommendations for users' personalized preferences, saves users' screening time, discovers suitable niche tourist destinations, and improves travel experience and satisfaction.
Smart Images

Figure CN118981574A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of tourism recommendation, and in particular to a method for recommending tourism destinations by intelligent matching. Background Art
[0002] The intelligent travel destination recommendation system can accurately find your ideal travel destination based on your personalized needs. You only need to fill in basic information, travel preferences, travel time, budget and travel method, and the system will match it with the huge travel destination database through advanced algorithms. Whether it is a seaside vacation, cultural exploration, or a natural scenery tour, it can quickly recommend suitable destinations for you, and provide itinerary planning, budget estimation and user evaluation references to help you easily start a perfect journey, discover the beautiful corners of the world, and make every trip an unforgettable experience.
[0003] Existing technologies include an invention with a publication number of CN111931070B, which discloses a method, device, readable storage medium and computer equipment for recommending tourist destinations. The method for recommending tourist destinations includes: obtaining travel time and target tourist destinations, and obtaining the first candidate scenic spot in the target tourist destination; searching for vegetation information corresponding to each of the first candidate scenic spots in a pre-stored database, wherein the vegetation information at least includes vegetation types and corresponding vegetation flowering periods; determining whether there is a first destination scenic spot among all the first candidate scenic spots, and the vegetation flowering period in the vegetation information corresponding to the first destination scenic spot matches the travel time; if the first destination scenic spot exists, the first destination scenic spot is pushed. The present invention enables users to see vegetation in flowering when they arrive at the first destination scenic spot at the travel time, thereby improving the user's travel experience.
[0004] With the development of information technology and the Internet, people have gradually entered an era of information overload from an era of information scarcity. Recommendation systems are one of the most effective ways to solve information overload. However, traditional travel websites provide a huge amount of data queries, and tourist destinations often list some popular attractions. It is difficult to match recommendations according to user needs, which reduces the quality of tourism services and the travel experience of travelers. To address this problem, it is necessary to design a solution to provide users with personalized attraction recommendation services, save user time and economic costs, and enhance users' travel experience. This is of great significance for fully building and developing smart tourism, meeting the personalized needs of tourists, and improving the quality of tourism services. Summary of the invention
[0005] The purpose of the present invention is to solve the problems in the prior art that traditional travel websites provide a huge amount of data queries, and travel destinations often list some popular attractions, which makes it difficult to match and recommend according to user needs, reducing the quality of travel services and the travel experience of travelers, and proposes an intelligent matching travel destination recommendation method.
[0006] In order to achieve the above object, the present invention adopts the following technical solution: a method for recommending a tourist destination by intelligent matching, comprising the following steps:
[0007] S1. Performing text segmentation on the voice description or text description input by the user to form a word sequence, extracting destination dimension keywords from the word sequence, and extracting overall information of the associated city from the city scenic spot tourism information database according to the dimension keywords;
[0008] S2. Use software to crawl the classification and ranking data of 5A / 4A-level scenic spots across the country from Maidian Cultural Tourism Big Data Center, including a small number of 3A-level scenic spots. The obtained data is divided into 5 categories according to the classification standards of the data center, with a total of 28 different types of scenic spots. Perform geographic analysis (POI analysis) based on Baidu API to obtain the latitude and longitude of the corresponding scenic spots and the cities they belong to. Classify and sort the cities where the various types of scenic spots are located based on the MBI value ranking of the Maidian website to form a tourist city scenic spot type library;
[0009] S3. Combined with the data from the National City Yearbook, the urban green coverage rate and the average concentration of fine air particles are used as references for the environmental classification of prefecture-level cities across the country. The per capita medical resources in the entire city are used as references for the convenience classification of prefecture-level cities across the country. The per capita available public transportation and rental cars are used as references for the transportation classification of prefecture-level cities across the country. The national urban meteorological data from the National Meteorological Information Center is used as references for the climate classification of prefecture-level cities across the country. Together with the type library, it constitutes a tourist city scenic spot library for urban feature analysis.
[0010] S4. The data in the tourist city scenic area database are classified into 11 levels, with the lowest being 1 and the highest being 11. Based on the keywords, radar maps of all cities can be obtained. Different radar maps of different cities can be displayed according to different needs.
[0011] Preferably, the scenic spot type library in step 2 includes: scenic spot information, destination climate, consumption level, and travel safety factor.
[0012] Preferably, scenic spot information: including scenic spot type (such as beaches, mountains, historical sites, museums, etc.), scenic spot features, opening hours, ticket prices, etc.; destination climate: average temperature in different seasons, precipitation conditions, probability of extreme weather; consumption level: local accommodation, catering, transportation and other cost levels; travel safety factor: local security conditions, frequency of natural disasters, etc.
[0013] Preferably, the information input by the user includes collecting basic information such as the user's age, gender, occupation, income level, etc. through user registration or filling out a questionnaire, and allowing the user to choose the type of travel they like, such as seaside vacation, historical and cultural tour, natural scenery tour, city sightseeing tour, adventure tour, parent-child tour, etc., understanding the time period when the user plans to travel, including season, vacation length, etc., asking the user for the approximate budget for the trip, and understanding the user's preferred travel method, such as airplane, train, self-driving, cruise ship, etc.
[0014] Preferably, for travel preferences, if the user chooses a seaside vacation, the system will give priority to recommending coastal cities or coastal areas; if a historical and cultural tour is chosen, historical and cultural cities and areas with concentrated historical sites will be recommended; in terms of travel time, if the user plans to travel in winter and likes skiing, the system will recommend areas in the north with high-quality ski resorts; if traveling in summer and likes seaside vacations, seaside destinations with suitable temperatures will be recommended; within the budget, destinations that meet the user's consumption capacity will be screened out, excluding those places with too high or too low consumption levels; as for travel mode preference, if the user prefers to drive, areas with convenient transportation and developed road networks will be recommended.
[0015] Preferably, in step 4, a commercial AI semantic module is used to perform keyword analysis on the input fuzzy tourist destination, or to sort the keywords, and the weight information is initially 1; a destination demand radar chart is formed with the input keyword as a dimension, such as if you want to go to a small city with rich history and culture, it would be even better if you can go rafting, etc.; compare the coverage area of all cities with that of the radar chart, and select the N cities with the largest repeated areas under these five dimensions (which can be 10 or more, depending on the scope of the next step) to complete the first step of screening of tourist destinations; then, based on the latitude and longitude coordinates of the user's travel location, combined with factors such as travel time and distance, a spatial range circle that conforms to the time distance is drawn with the user's location as the center of the circle. If the cities in the circle are repeated with the cities screened out in the first step, the repeated cities are intelligently matched cities.
[0016] Preferably, the system architecture of the solution includes: a user interface layer, a data processing layer, a matching algorithm layer, and a data storage layer.
[0017] Preferably, the user interface layer includes websites and mobile applications, where users can fill in personal information and travel preferences after registering and logging in; the data processing layer is responsible for collecting, organizing and analyzing information input by users, as well as stored travel destination data; the matching algorithm layer uses specific algorithms to match user needs and destination data to generate recommendation results; the data storage layer stores user information, travel destination information, evaluation data, matching algorithm models, etc.
[0018] Preferably, a corresponding weight is assigned to each demand factor based on the information input by the user. For example, if the user clearly indicates that the travel theme is the primary consideration, then the weight of the travel theme is relatively high; the weights of budget and travel time are also set according to the user's emphasis; a collaborative filtering algorithm is used to recommend travel destinations to target users based on the choices and evaluations of other users with similar preferences; a content-based recommendation algorithm is used to make recommendations based on the degree of match between the characteristics of the destination (such as attraction type, consumption level, climate conditions, etc.) and user needs.
[0019] Preferably, after using the recommendation service, users can evaluate and provide feedback on the recommendation results. The system continuously optimizes the intelligent matching algorithm based on user feedback to improve the accuracy and satisfaction of recommendations.
[0020] Compared with the prior art, the advantages and positive effects of the present invention are:
[0021] 1. The present invention provides a method for recommending tourist destination scenic spots based on tourism intelligent matching, which can recommend scenic spots that better meet the personalized preferences of users, making it easier for users to plan the best travel routes. At the same time, the recommendation results are more accurate and explainable.
[0022] 2. In the present invention, tourists do not need to spend a lot of time to screen through a large amount of tourism information, quickly determine the appropriate destination, and plan a travel plan. For example, for busy office workers, the recommendation system can quickly help them find travel destinations that meet their vacation duration and budget, saving time on searching and comparing on the Internet.
[0023] 3. The present invention helps tourists break through conventional cognition and discover some niche or emerging tourist destinations that they have not paid attention to in the past but are very suitable for them. For example, some villages or towns that are not yet well-known to the public but have beautiful scenery and unique culture enter the tourists' field of vision through the recommendation system.
[0024] 4. The present invention can accurately recommend suitable destinations based on tourists' interests, budget, time and other personalized factors. For example, tourists who like history and culture can be recommended to cities with rich historical sites and cultural heritage; students with limited budgets can get recommendations for cost-effective tourist destinations.
[0025] 5. In the present invention, reasonable destination recommendations can help tourists better match their expectations, thereby achieving higher satisfaction and better experience during their travels. For example, the recommended destinations have suitable weather, rich activities, and accommodation and dining conditions that meet tourists' preferences, making tourists more comfortable and happy during their travels. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] Figure 1 A system schematic diagram of an intelligent matching tourist destination recommendation method proposed by the present invention;
[0027] Figure 2 A radar map of Beijing in a method for recommending a tourist destination by intelligent matching is provided in the present invention;
[0028] Figure 3 A radar map of Harbin City in a method for recommending a tourist destination by intelligent matching is proposed in the present invention;
[0029] Figure 4 A radar map of Nanjing City in a method for recommending a tourist destination by intelligent matching is provided in the present invention;
[0030] Figure 5 A radar map of Yangzhou City in a method for recommending a tourist destination by intelligent matching proposed by the present invention;
[0031] Figure 6 A radar map of Guangzhou City in a method for recommending a tourist destination by intelligent matching is provided in the present invention;
[0032] Figure 7 A radar map of Xi'an in a method for recommending a tourist destination by intelligent matching is provided in the present invention;
[0033] Figure 8 A radar map of Linzhi City in the intelligent matching tourist destination recommendation method proposed by the present invention;
[0034] Fig. 9 A radar chart of target demand in a method for recommending tourist destinations by intelligent matching is proposed for the present invention;
[0035] Fig.10 A schematic diagram of matching results in an intelligent matching tourist destination recommendation method proposed by the present invention. DETAILED DESCRIPTION
[0036] See also Figure 1-Figure 10 The present invention provides a technical solution: a method for recommending a tourist destination by intelligent matching, comprising the following steps:
[0037] S1. Performing text segmentation on the voice description or text description input by the user to form a word sequence, extracting destination dimension keywords from the word sequence, and extracting overall information of the associated city from the city scenic spot tourism information database according to the dimension keywords;
[0038] S2. Use software to crawl the classification and ranking data of 5A / 4A-level scenic spots across the country from Maidian Cultural Tourism Big Data Center, including a small number of 3A-level scenic spots. The obtained data is divided into 5 categories according to the classification standards of the data center, with a total of 28 different types of scenic spots. Perform geographic analysis (POI analysis) based on Baidu API to obtain the latitude and longitude of the corresponding scenic spots and the cities they belong to. Classify and sort the cities where the various types of scenic spots are located based on the MBI value ranking of the Maidian website to form a tourist city scenic spot type library;
[0039] S3. Combined with the data from the National City Yearbook, the urban green coverage rate and the average concentration of fine air particles are used as references for the environmental classification of prefecture-level cities across the country. The per capita medical resources in the entire city are used as references for the convenience classification of prefecture-level cities across the country. The per capita available public transportation and rental cars are used as references for the transportation classification of prefecture-level cities across the country. The national urban meteorological data from the National Meteorological Information Center is used as references for the climate classification of prefecture-level cities across the country. Together with the type library, it constitutes a tourist city scenic spot library for urban feature analysis.
[0040] S4. The data in the tourist city scenic area database are classified into 11 levels, with the lowest being 1 and the highest being 11. Based on the keywords, radar maps of all cities can be obtained. Different radar maps of different cities can be displayed according to different needs.
[0041] Specifically, the scenic spot type library in step 2 includes: scenic spot information, destination climate, consumption level, and travel safety factor.
[0042] Specifically, information about tourist attractions: including types of attractions (such as beaches, mountains, historical sites, museums, etc.), features of attractions, opening hours, ticket prices, etc.; climate of destination: average temperature in different seasons, precipitation conditions, probability of extreme weather; consumption level: local accommodation, catering, transportation and other cost levels; travel safety factor: local security conditions, frequency of natural disasters, etc.
[0043] In this embodiment: the information input by the user includes collecting the user's basic information such as age, gender, occupation, income level, etc. through user registration or filling out a questionnaire, and allowing the user to choose the type of travel they like, such as seaside vacation, historical and cultural tour, natural scenery tour, city sightseeing tour, adventure tour, parent-child tour, etc., to understand the time period when the user plans to travel, including season, vacation length, etc., to ask the user about the approximate budget for the trip, and to understand the user's preferred travel mode, such as airplane, train, self-driving, cruise ship, etc.
[0044] Specifically, for travel preferences, if the user chooses a seaside vacation, the system will give priority to recommending coastal cities or coastal areas; if a historical and cultural tour is chosen, historical and cultural cities and areas with concentrated historical sites will be recommended; in terms of travel time, if the user plans to travel in winter and likes skiing, the system will recommend areas in the north with high-quality ski resorts; if traveling in summer and likes seaside vacations, seaside destinations with suitable temperatures will be recommended; within the budget, destinations that meet the user's consumption capacity will be screened out, excluding those places with too high or too low consumption levels; in terms of travel mode preference, if the user prefers to drive, areas with convenient transportation and developed road networks will be recommended.
[0045] In this embodiment: Step 4 uses a commercial AI semantic module to perform keyword analysis on the input fuzzy tourist destination, or sort the keywords, and the weight information is initially 1; using the input keyword as a dimension, a destination demand radar chart is formed, such as if you want to go to a small city with a rich history and culture, it would be even better if you can go rafting, etc., this can be formed; comparing the coverage area of all cities with that of the radar chart, selecting the N cities with the largest repeated areas under these five dimensions (which can be 10 or more, depending on the scope of the next step), to complete the first step of screening of tourist destinations; then based on the latitude and longitude coordinates of the user's travel location, combined with factors such as travel time and distance, a spatial range circle that conforms to the time distance is drawn with the user's location as the center, and if the cities in the circle are repeated with the cities screened out in the first step, the repeated cities are the cities intelligently matched.
[0046] Specifically, the system architecture of the solution includes: user interface layer, data processing layer, matching algorithm layer, and data storage layer.
[0047] In this embodiment: User interface layer: including websites and mobile applications, users can fill in personal information and travel preferences after registration and login; Data processing layer: responsible for collecting, organizing and analyzing information input by users, as well as stored travel destination data; Matching algorithm layer: using specific algorithms to match user needs and destination data to generate recommendation results; Data storage layer: storing user information, travel destination information, evaluation data, matching algorithm models, etc.
[0048] Specifically, based on the information input by the user, a corresponding weight is assigned to each demand factor. For example, if the user clearly indicates that the travel theme is the primary consideration, then the weight of the travel theme is relatively high; the weights of budget and travel time are also set according to the user's emphasis; a collaborative filtering algorithm is used to recommend travel destinations to target users based on the choices and evaluations of other users with similar preferences; a content-based recommendation algorithm is used to make recommendations based on the degree of match between the characteristics of the destination (such as attraction type, consumption level, climate conditions, etc.) and user needs.
[0049] Specifically, after using the recommendation service, users can evaluate and provide feedback on the recommendation results. The system continuously optimizes the intelligent matching algorithm based on user feedback to improve the accuracy and satisfaction of recommendations.
[0050] Working principle: In the present invention, a method for recommending tourist destinations based on intelligent tourism matching is provided, which can recommend scenic spots that better meet personalized preferences for users, facilitate users to plan the best travel routes, and the recommendation results are more accurate and explainable; in the present invention, tourists do not need to spend a lot of time to screen in a large amount of tourism information, quickly determine the appropriate destination, and plan travel plans. For example, for busy office workers, the recommendation system can quickly help them find travel destinations that meet the holiday duration and budget, saving time for searching and comparing on the Internet; in the present invention, it helps tourists break through conventional cognition and discover some niche or emerging tourist destinations that have not been paid attention to in the past but are very suitable for them. For example, some villages or towns that are not yet well-known to the public but have beautiful scenery and unique culture enter the field of vision of tourists through the recommendation system; in the present invention, it can accurately recommend suitable destinations according to personalized factors such as tourists' interests, budgets, time, etc. For example, tourists who like history and culture can be recommended to cities with rich monuments and cultural heritage; students with limited budgets can get cost-effective tourist destination recommendations; in the present invention, reasonable destination recommendations can allow tourists to better match their expectations, thereby obtaining higher satisfaction and better experience in travel. For example, the recommended destinations have suitable weather, abundant activities, and accommodation and dining conditions that meet tourists' preferences, making tourists more comfortable and enjoyable during their trip.
Claims
1. A method for recommending a tourist destination by intelligent matching, comprising the following steps: S1. Performing text segmentation on the voice description or text description input by the user to form a word sequence, extracting destination dimension keywords from the word sequence, and extracting overall information of the associated city from the city scenic spot tourism information database according to the dimension keywords; S2. Use software to crawl the classification and ranking data of 5A / 4A-level scenic spots across the country from Maidian Cultural Tourism Big Data Center, including a small number of 3A-level scenic spots. The obtained data is divided into 5 categories according to the classification standards of the data center, with a total of 28 different types of scenic spots. Perform geographic analysis (POI analysis) based on Baidu API to obtain the latitude and longitude of the corresponding scenic spots and the cities they belong to. Classify and sort the cities where the various types of scenic spots are located based on the MBI value ranking of the Maidian website to form a tourist city scenic spot type library; S3. Combined with the data from the National City Yearbook, the urban green coverage rate and the average concentration of fine air particles are used as references for the environmental classification of prefecture-level cities across the country. The per capita medical resources in the entire city are used as references for the convenience classification of prefecture-level cities across the country. The per capita available public transportation and rental cars are used as references for the transportation classification of prefecture-level cities across the country. The national urban meteorological data from the National Meteorological Information Center is used as references for the climate classification of prefecture-level cities across the country. Together with the type library, it constitutes a tourist city scenic spot library for urban feature analysis. S4. The data in the tourist city scenic area database are classified into 11 levels, with the lowest being 1 and the highest being 11. Based on the keywords, radar maps of all cities can be obtained. Different radar maps of different cities can be displayed according to different needs.
2. The method for recommending a tourist destination by intelligent matching according to claim 1, characterized in that: The scenic spot type library in step 2 includes: scenic spot information, destination climate, consumption level, and travel safety factor.
3. The method for recommending a tourist destination by intelligent matching according to claim 2, characterized in that: Attraction information: including attraction type (such as beaches, mountains, historical sites, museums, etc.), attraction features, opening hours, ticket prices, etc.; destination climate: average temperature in different seasons, precipitation conditions, probability of extreme weather; consumption level: local accommodation, catering, transportation and other cost levels; travel safety factor: local security conditions, frequency of natural disasters, etc.
4. The method for recommending a tourist destination by intelligent matching according to claim 1, characterized in that: The information input by the user includes collecting basic information such as the user's age, gender, occupation, income level, etc. through user registration or filling out questionnaires, and allowing the user to choose the type of travel they like, such as seaside vacations, historical and cultural tours, natural scenery tours, city sightseeing tours, adventure tours, parent-child tours, etc., to understand the time period when the user plans to travel, including season, holiday length, etc., ask the user for the approximate budget for the trip, and understand the user's preferred travel method, such as airplane, train, self-driving, cruise ship, etc.
5. The method for recommending a tourist destination by intelligent matching according to claim 1, characterized in that: Regarding travel preferences, if the user chooses a seaside vacation, the system will give priority to recommending coastal cities or coastal areas; if a historical and cultural tour is chosen, historical and cultural cities and areas with concentrated historical sites will be recommended; in terms of travel time, if the user plans to travel in winter and likes skiing, the system will recommend areas in the north with high-quality ski resorts; if the user travels in summer and likes a seaside vacation, the system will recommend seaside destinations with suitable temperatures; within the budget, the system will screen out destinations that meet the user's spending power and exclude those places with too high or too low spending levels; as for travel mode preferences, if the user prefers to drive, the system will give priority to recommending areas with convenient transportation and developed road networks.
6. The method for recommending a tourist destination by intelligent matching according to claim 1, characterized in that: Step 4: Use a commercial AI semantic module to perform keyword analysis on the input fuzzy tourist destination, or sort the keywords, with the initial weight information being 1; use the input keywords as dimensions to form a destination demand radar chart, such as if you want to go to a small city with a rich history and culture, it would be even better if you can go rafting, etc.; compare the coverage area of all cities with that of the radar chart, and select the N cities with the largest repeated areas under these five dimensions (which can be 10 or more, depending on the scope of the next step) to complete the first step of screening the tourist destination; then, based on the latitude and longitude coordinates of the user's travel location, combined with factors such as travel time and distance, draw a spatial range circle that conforms to the time distance with the user's location as the center. If the cities in the circle are repeated with the cities screened out in the first step, the repeated cities are the cities that are intelligently matched.
7. The method for recommending a tourist destination by intelligent matching according to claim 1, characterized in that: The system architecture of the solution includes: user interface layer, data processing layer, matching algorithm layer, and data storage layer.
8. The method for recommending a tourist destination by intelligent matching according to claim 7, characterized in that: User interface layer: including websites and mobile applications, where users can fill in personal information and travel preferences after registering and logging in; Data processing layer: responsible for collecting, organizing and analyzing user input information and stored travel destination data; Matching algorithm layer: using specific algorithms to match user needs and destination data to generate recommendation results; Data storage layer: stores user information, travel destination information, evaluation data, matching algorithm models, etc.
9. The method for recommending a tourist destination by intelligent matching according to claim 1, characterized in that: Based on the information input by the user, each demand factor is assigned a corresponding weight. For example, if the user clearly indicates that the travel theme is the primary consideration, then the weight of the travel theme is relatively high; the weights of budget and travel time are also set according to the user's emphasis; using the collaborative filtering algorithm, based on the choices and evaluations of other users with similar preferences, recommend travel destinations to target users; using a content-based recommendation algorithm, recommendations are made based on the degree of match between the characteristics of the destination (such as attraction type, consumption level, climate conditions, etc.) and user needs.
10. The method for recommending a tourist destination by intelligent matching according to claim 1, characterized in that: After using the recommendation service, users can evaluate and provide feedback on the recommendation results. The system continuously optimizes the intelligent matching algorithm based on user feedback to improve the accuracy and satisfaction of recommendations.
Citation Information
Cited By
Cultural tourism route real-time optimization and adjustment system based on big data analysis
CN119250330A
Method for dividing and screening urban tourism data
CN121301633A
