Tourist route recommendation system and method adopting big data retrieval
The travel route recommendation system, which uses big data retrieval, solves the problem that traditional recommendation methods cannot meet personalized needs, and achieves accurate and scientific travel route recommendations, thereby improving user experience and efficiency.
Patent Information
- Application Number
- CN202511850368.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-09
- Publication Date
- 2026-03-06
AI Technical Summary
Traditional travel route recommendation methods cannot meet users' personalized needs, nor can they effectively integrate and analyze multi-source data, resulting in recommendations that lack relevance and timeliness. Existing recommendation algorithms lack in-depth mining of user interests and preferences and multi-dimensional feature analysis.
This travel route recommendation system, which utilizes big data retrieval, generates personalized travel routes through data collection, storage, cleaning, and processing. It employs a big data processing framework and recommendation algorithms, and includes a data collection unit, a data storage unit, a big data processing unit, and a travel route generation unit. By combining user behavior data and attraction feature analysis, it calculates the matching degree between users and attractions to generate reasonable travel routes.
It achieves precise recommendations, improves the accuracy and relevance of recommendations, generates scientific and reasonable travel routes, adapts to diverse needs, and enhances users' travel experience and efficiency.
Smart Images
Figure CN121614673A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of data processing technology, and in particular relates to a travel route recommendation system and method using big data retrieval. Background Technology
[0002] With the continuous improvement of people's living standards, tourism has become one of the important ways for people to relax and have fun. The booming development of the tourism market has led to an increasingly rich selection of travel routes, and users' demand for personalized and precise travel route recommendations is becoming stronger.
[0003] Traditional methods of recommending tour routes rely primarily on human experience or simple rule matching. Tourism professionals design fixed itineraries based on their experience and knowledge of tourist attractions for users to choose from. This approach has several limitations. Firstly, manually designed routes struggle to comprehensively cover the diverse interests and needs of users, failing to meet individual requirements regarding the type of tourist attraction, tour duration, budget, and other personalized needs. For instance, some users prefer natural scenery, while others favor historical and cultural sites; traditional methods struggle to accurately match these different preferences.
[0004] On the other hand, with the widespread adoption of the internet, tourism-related information has exploded, providing users with access to a vast amount of travel information. However, this information is scattered across various travel websites, social media platforms, and online travel booking systems, making it difficult for users to quickly and accurately filter out travel routes that meet their needs. Furthermore, traditional recommendation systems cannot effectively integrate and analyze this multi-source data, resulting in recommendations that lack specificity and timeliness.
[0005] In recommendation algorithms, existing algorithms often consider only a single factor, such as recommending attractions solely based on their popularity or simply repeating recommendations based on a user's past travel destinations. They lack in-depth analysis of user interests and preferences, as well as comprehensive analysis of the multi-dimensional characteristics of tourist attractions. This results in a significant discrepancy between recommended travel routes and users' actual needs, failing to provide a satisfactory travel experience. Summary of the Invention
[0006] To address the problems existing in the prior art, this invention provides a travel route recommendation system and method using big data retrieval, which has the advantage of personalized travel route planning through big data and recommendation algorithms, thus solving the existing problems.
[0007] This invention is implemented as follows: a travel route recommendation system using big data retrieval, comprising: The data acquisition unit is used to collect target data, which is multi-source data containing tourism target keywords. The multi-source data includes tourist attractions, hotels, transportation information, as well as users' historical travel data and preferences. The data storage unit uses a distributed storage system to store the target data and cleans and preprocesses the target data, including deduplication, filling in missing values, and data format conversion. The big data processing unit uses big data processing frameworks (such as Apache Spark) to perform cleaning, transformation, and aggregation operations on stored data. It extracts user preference types from user historical travel data, extracts attraction features from tourist attraction data, uses cluster analysis to classify tourist attractions, obtains the feature types of attractions, and calculates the matching degree based on user preference types and attraction feature types. The tourism route generation unit selects attractions with a matching degree exceeding a threshold as recommended attractions, and uses route planning technology to generate tourism routes from several recommended attractions.
[0008] As a preferred embodiment of the present invention, the data acquisition unit includes: Data source management subunit: Used to maintain and expand the list of data sources, including tourism websites, social media platforms, IoT devices, government tourism department data interfaces, and meteorological data interfaces. It regularly evaluates the reliability and data quality of data sources and dynamically adjusts the collection priority of data sources according to needs. The crawler scheduling subunit uses distributed crawler technology to manage the task allocation of multiple crawler nodes. It adopts an incremental crawler strategy, records the collected data, and only crawls the updated or newly added data to improve the collection efficiency. User behavior collection subunit: During the user's use of the system, user behavior data is collected in real time, including search keywords, page browsing, link clicking, route saving, and attraction rating. This behavior data is classified and preliminarily organized in real time to provide a foundation for subsequent analysis and recommendations.
[0009] As a preferred embodiment of the present invention, the travel route recommendation system employing big data retrieval further includes a user interface unit for users to input their departure point, destination, travel time, and budget, and to display recommended travel routes. The interface is interactive, allowing users to rate and save recommended routes, and the system can adjust the recommendation results based on user feedback. As a preferred embodiment of the present invention, the data storage unit stores structured core information using a relational database (such as MySQL) and semi-structured and unstructured data using a NoSQL database (such as MongoDB).
[0010] As a preferred embodiment of the present invention, the matching degree formula is used in the recommendation algorithm: in, This represents the matching degree between user u and scenic spot i. User preference features The weight, is the feature of attraction i, and n is the number of features, used to adjust recommendation priority and weight in collaborative filtering and content-based recommendation.
[0011] As a preferred embodiment of the present invention, the outlier detection formula is used in data quality assessment: ;in, This indicates the degree of anomaly of data point x. It is the mean of the data. It is the standard deviation, used to identify and process outlier data.
[0012] As a preferred embodiment of the present invention, the weight of user preference features is calculated using the following formula: Where B represents the quantified value of user behavior data, and S represents the seasonal influencing factor of tourist attractions. Indicates the degree of matching between user budget and actual budget. They are The corresponding adjustment coefficient is used to automatically adjust the weights to improve calculation accuracy.
[0013] As a preferred embodiment of the present invention, the method for calculating the quantified value B of user behavior data is as follows: Search frequency standardization formula: ;in, The number of searches after standardization. The number of times a user searches for a certain type of tourism element. This represents the maximum number of searches for this type of tourism element by all users in the system within a certain time period. Standardized formula for browsing time: ;in, For standardized browsing time, The time a user spends browsing a particular page. The longest time all users in the system have viewed this type of page; Standardized formula for the number of collections: ;in This refers to the standardized number of collections. The number of times a user has favorited their work. This represents the maximum number of times a item can be saved in the system. Standardized formula for the number of likes: ;in The number of likes after standardization. The number of likes a user receives. This represents the maximum number of likes in the system. Standardized formula for booking and purchasing behavior: ;in A is a standardized value representing the combined amount of booking and purchasing behavior, where A is the booking or purchasing amount. F represents the maximum booking or purchase amount in the system, and F represents the booking or purchase frequency. This represents the highest booking or purchase frequency in the system. General formula for calculating the quantified value of user behavior data: Where B is the quantified value of user behavior data. The weights for search behavior, browsing behavior, collection and liking behavior, and booking and purchasing behavior are respectively: .
[0014] As a preferred embodiment of the present invention, the method for calculating the quantified value B of user behavior data is as follows: The method for calculating the seasonality impact factor S of tourist attractions is as follows: Collect tourist flow or popularity index data from different seasons over many years; Calculate the average tourist flow or popularity index for each season and the whole year; calculate User budget matching degree The calculation method is as follows: obtain user budget And the cost of the tour route or attractions (C); calculate the difference. Using formula Calculate the user's budget match.
[0015] A method for recommending travel routes using big data retrieval includes the following steps: Collect target data, which is multi-source data containing tourism target keywords. The multi-source data includes tourist attractions, hotels, transportation information, as well as users' historical travel data and preferences. The target data is stored using a distributed storage system, and the target data is cleaned and preprocessed, including deduplication, filling in missing values, and data format conversion. Using a big data processing framework, the stored data is cleaned, transformed, and aggregated. User preference types are extracted from historical travel data, and attraction features are extracted from tourist attraction data. Cluster analysis is used to classify tourist attractions to obtain their feature types. The matching degree is calculated based on the user preference type and the attraction feature type. Tourist attractions with a matching degree exceeding a threshold are selected as recommended attractions, and route planning technology is used to generate tourist routes from several recommended attractions.
[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: Precise Recommendations: By collecting and deeply analyzing multi-source data, we can accurately grasp users' interests and travel needs, recommending highly matching tourist attractions and routes to users, improving the accuracy and targeting of recommendations, avoiding the blindness of traditional recommendation methods, and increasing the probability of users finding their ideal travel routes.
[0017] Rational tour route planning: Tour routes are generated based on route planning technology, taking into full account the actual conditions between attractions, making the routes more scientific and rational. Users can more efficiently visit multiple attractions of interest within a limited time, reducing travel time and wasted time, and improving the quality and efficiency of the tour.
[0018] Adapting to diverse needs: Due to its wide range of data sources, the system can cover various types of tourism resources and user needs. Whether you are a travel enthusiast who enjoys in-depth tours of niche attractions, a general tourist who seeks to visit popular spots, or a user with special preferences (such as family trips or honeymoons), you can find suitable travel routes through this system, making it highly adaptable and versatile. Attached Figure Description
[0019] Figure 1 This is a schematic diagram of a travel route recommendation system using big data retrieval provided in an embodiment of the present invention; Figure 2 This is a flowchart of a tourism route recommendation method using big data retrieval provided in an embodiment of the present invention. Detailed Implementation
[0020] To further understand the invention's content, features, and effects, the following embodiments are provided, and detailed descriptions are given in conjunction with the accompanying drawings.
[0021] The structure of the present invention will now be described in detail with reference to the accompanying drawings.
[0022] like Figures 1 to 2 As shown in the figure, an embodiment of the present invention provides a travel route recommendation system using big data retrieval, comprising: The data acquisition unit is used to collect target data, which is multi-source data containing tourism target keywords. The multi-source data includes tourist attractions, hotels, transportation information, as well as users' historical travel data and preferences. The data storage unit uses a distributed storage system (such as Hadoop's HDFS or Apache Cassandra) to store the target data and cleans and preprocesses the target data, including deduplication, filling missing values, and data format conversion. The big data processing unit uses big data processing frameworks (such as Apache Spark) to perform cleaning, transformation, and aggregation operations on stored data. It extracts user preference types from user historical travel data, extracts attraction features from tourist attraction data, uses cluster analysis to classify tourist attractions, obtains the feature types of attractions, and calculates the matching degree based on user preference types and attraction feature types. The tourism route generation unit selects attractions with a matching degree exceeding a threshold as recommended attractions, and uses route planning technology to generate tourism routes from several recommended attractions.
[0023] Specifically, the data acquisition unit includes: Data source management subunit: Used to maintain and expand the list of data sources, including tourism websites, social media platforms, IoT devices, government tourism department data interfaces, and meteorological data interfaces. It regularly evaluates the reliability and data quality of data sources and dynamically adjusts the collection priority of data sources according to needs. The crawler scheduling subunit uses distributed crawler technology to manage the task allocation of multiple crawler nodes. It adopts an incremental crawler strategy, records the collected data, and only crawls the updated or newly added data to improve the collection efficiency. User behavior collection subunit: During the user's use of the system, user behavior data is collected in real time, including search keywords, page browsing, link clicking, route saving, and attraction rating. This behavior data is classified and preliminarily organized in real time to provide a foundation for subsequent analysis and recommendations.
[0024] Furthermore, the travel route recommendation system using big data retrieval also includes a user interface unit, which allows users to input their departure point, destination, travel time, and budget, and displays recommended travel routes. The interface is interactive, allowing users to rate and save recommended routes, and the system can adjust the recommendation results based on user feedback.
[0025] Specifically, the data storage unit stores structured core information in a relational database (such as MySQL) and semi-structured and unstructured data in a NoSQL database (such as MongoDB).
[0026] Furthermore, a matching degree formula is used in the recommendation algorithm: ;in, This represents the matching degree between user u and scenic spot i. User preference features The weight, is the feature of attraction i, and n is the number of features, used to adjust recommendation priority and weight in collaborative filtering and content-based recommendation.
[0027] This formula measures the degree of match between user u and attraction i. The numerator is derived from user preference features. and scenic spot features A weighted sum is performed to reflect the correlation between the two features across various feature dimensions. The denominator calculates the vector magnitudes of the user preference features and the attraction features for normalization, thus improving the matching degree. The value ranges from [-1, 1] (when the feature is a vector) to [0, 1] (when the feature is non-negative), which facilitates the comparison of matching degree between different users and attractions.
[0028] Example as follows: Suppose we consider two features: natural landscape preference and historical and cultural preference, n=2. User A's preference for natural landscapes. Preference for history and culture The weights are respectively Scenic Spot X has natural landscape features Historical and cultural characteristics
[0029]
[0030] Length of the user preference feature vector in the denominator:
[0031] Length of the feature vector of the scenic spot in the denominator:
[0032] Then the matching degree For example, if the matching threshold is set to 0.6, a spot with a matching degree greater than 0.6 will be designated as a recommended attraction, and a spot with a matching degree less than 0.6 will not be designated as a recommended attraction.
[0033] Furthermore, an outlier detection formula is used in data quality assessment: ;in, This indicates the degree of anomaly of data point x. It is the mean of the data. It is the standard deviation, used to identify and process outlier data.
[0034] This formula, based on the Z-score principle, is used to determine whether a data point x is an outlier. It works by calculating the difference between the data point x and the dataset mean. The difference, divided by the standard deviation of the dataset. Obtain the degree of anomaly of data point x Usually, when A data point is considered an outlier if its value exceeds a threshold (e.g., 3). Standard deviation The mean reflects the dispersion of the data. These two statistics represent the central tendency of the data, and can be used to effectively identify data points that deviate from the normal range.
[0035] Suppose we have a set of daily visitor flow data for tourist attractions: First, calculate the mean. Then calculate the standard deviation. (Calculation process omitted), assuming For data point 500, calculate... Although 2.17 does not exceed the common threshold of 3, it is relatively large, indicating that the data point of 500 deviates significantly from the mean and may require further analysis. If there is another data point of 1000, then... If the value exceeds the threshold of 3, it can be identified as an outlier.
[0036] Furthermore, the weights of user preference features are calculated using the following formula: Where B represents the quantified value of user behavior data, and S represents the seasonal influencing factor of tourist attractions. Indicates the degree of matching between user budget and actual budget. They are The corresponding adjustment coefficient is used to automatically adjust the weights to improve calculation accuracy.
[0037] This formula comprehensively considers the quantified value of user behavior data (B), the seasonal impact factor of tourist attractions (S), and the matching degree of user budget. Three factors, adjusted by coefficients This allows for dynamic adjustment of the relative importance of each factor in the weight calculation. The larger the value, the greater the proportion of user behavior data in the weighting calculation; The larger the value, the more important seasonal factors are for tourist attractions; The larger the value, the greater the impact of the user budget matching degree on the weight. In this way, weights can be flexibly assigned to user preference features according to different scenarios and needs, improving the accuracy and adaptability of the recommendation system.
[0038] - Example: Suppose Users have recently been searching for and browsing seaside attractions frequently, resulting in a significant increase in the quantification of user behavior data. Currently, it is the direct season; what are the seasonal influencing factors for a certain seaside scenic spot? The user's budget is 5000 yuan, and the cost of this seaside tour is 4500 yuan.
[0039] - Calculate weights This indicates that, under the current circumstances, the seaside attraction has a high weight in the user's preference features and will be given priority in the recommendation system.
[0040] Furthermore, the formula for standardizing search counts: ;in, The number of searches after standardization. The number of times a user searches for a certain type of tourism element. This represents the maximum number of searches for this type of tourism element by all users in the system within a certain time period. Standardized formula for browsing time: ;in, For standardized browsing time, The time a user spends browsing a particular page. The longest time all users in the system have viewed this type of page; Standardized formula for the number of collections: ;in This refers to the standardized number of collections. The number of times a user has favorited their work. This represents the maximum number of times a item can be saved in the system. Standardized formula for the number of likes: ;in The number of likes after standardization. The number of likes a user receives. This represents the maximum number of likes in the system. Standardized formula for booking and purchasing behavior: ;in A is a standardized value representing the combined amount of booking and purchasing behavior, where A is the booking or purchasing amount. F represents the maximum booking or purchase amount in the system, and F represents the booking or purchase frequency. This represents the highest booking or purchase frequency in the system. General formula for calculating the quantified value of user behavior data: Where B is the quantified value of user behavior data. The weights for search behavior, browsing behavior, collection and liking behavior, and booking and purchasing behavior are respectively: .
[0041] The calculation example is as follows: Search behavior: The number of times a user searched for "seaside vacation" within a month. It is 10 times.
[0042] The maximum number of searches for "seaside vacation" by all users in the system during that month. It is 50 times.
[0043] Browsing behavior: The user spent a certain amount of time browsing pages related to "seaside vacation". The longest browsing time for all users in the system on pages related to "Beach Resort" was 45 minutes. It lasts 90 minutes.
[0044] Assume the user's browsing depth normalization value is 0.7 (e.g., based on a comprehensive evaluation of the user's browsing page hierarchy and the number of links clicked).
[0045] Collection and Like Behavior: The number of times this user collects "seaside vacation" related travel guides. The maximum number of times a user can save content related to "seaside vacation" in the system is 5. The number is 20. This indicates the number of times the user liked "seaside resort" related attractions. The maximum number of likes a user can give to content related to "seaside vacation" is 3. It is 10 times.
[0046] Reservation and purchase behavior: The user booked a travel package under the "Beach Resort" category, costing A = 2000 yuan. What is the maximum amount a user can book for a "Beach Resort" package in the system? The price is 5000 yuan. This user's booking frequency under the "Beach Resort" category is F=1 time. This is the maximum booking frequency for this user under the "Beach Resort" category in the system. It was 3 times.
[0047] Weight settings: Weights are determined through data analysis and expert experience: Search behavior weights Browsing behavior weight Weighting of collection and like behaviors Weighting of booking and purchasing behavior
[0048] The calculation process is as follows:
[0049] Furthermore, the calculation method for the seasonality impact factor S of tourist attractions is as follows: Collect tourist flow or popularity index data from different seasons over many years; Calculate the average tourist flow or popularity index for each season and the whole year;
[0050] The following is an example of calculating the seasonality impact factor for tourist attractions: The seasonality impact factor is calculated using tourist traffic data from the past 5 years. 1. Data collection and organization: Spring (March-May): The tourist numbers for the past 5 years during spring were 200,000, 220,000, 250,000, 230,000, and 240,000 respectively, with an average tourist number of...
[0051] Summer (June-August): The tourist numbers for the past 5 years were 300,000, 320,000, 350,000, 330,000, and 340,000 respectively, with an average tourist number of...
[0052] Autumn (September-November): The tourist numbers for the past five autumns were 280,000, 300,000, 310,000, 290,000, and 300,000 respectively, with an average tourist number of...
[0053] Winter (December-February): Tourist numbers for the past five winters were 150,000, 160,000, 180,000, 170,000, and 160,000 respectively, with an average tourist number of...
[0054] Average annual tourist volume: The total annual tourist volume over the past 5 years was 5.34 million, and the average annual tourist volume was...
[0055] 2. Calculate the seasonality influencing factor:
[0056] Here's an example of how user budget matching is calculated: Suppose a user plans to travel to Sanya, and their set travel budget... The price was 8,000 yuan, while the tour package he was interested in, C, cost 7,000 yuan.
[0057] 1. Calculate the difference between the budget and the cost:
[0058] 2. Calculate the user's budget fit: based on the formula achievable This indicates that the tour itinerary is a good fit for Xiao Zhang's budget, and according to the recommendation system, it's a suitable choice for him. If the other tour costs 9500 yuan, then... In comparison, the former route has a greater advantage in terms of budget matching.
[0059] This application also provides a method for recommending travel routes using big data retrieval, including the following steps: step Collect target data, which is multi-source data containing tourism target keywords. The multi-source data includes tourist attractions, hotels, transportation information, as well as users' historical travel data and preferences. step The target data is stored using a distributed storage system, and the target data is cleaned and preprocessed, including deduplication, filling in missing values, and data format conversion. step Using a big data processing framework, the stored data is cleaned, transformed, and aggregated. User preference types are extracted from historical travel data, and attraction features are extracted from tourist attraction data. Cluster analysis is used to classify tourist attractions to obtain their feature types. The matching degree is calculated based on the user preference type and the attraction feature type. step Tourist attractions with a matching degree exceeding a threshold are selected as recommended attractions, and route planning technology is used to generate tourist routes from several recommended attractions.
[0060] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0061] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A travel route recommendation system using big data search, characterized by, Comprise: Data acquisition unit for collecting target data, the target data is multi-source data containing tourism target keywords, the multi-source data contains tourist attractions, hotels, traffic information and user historical travel data and preferences; Data storage unit, using distributed storage system to store the target data, and cleaning and preprocessing the target data, including deduplication, filling missing values, data format conversion; Big data processing unit, with the help of big data processing framework, the stored data is executed cleaning, conversion, aggregation operation, the user's preference type is extracted from the user historical travel data, the scenic spot features are extracted from the tourist attraction data, the tourist attractions are classified by using clustering analysis, the feature type of the scenic spot is obtained, and the matching degree is calculated according to the user's preference type and the feature type of the scenic spot; Tourist route generation unit, selecting the scenic spots with matching degree exceeding threshold value as recommended scenic spots, and generating tourist routes through route planning technology.
2. The tourist route recommendation system using big data retrieval according to claim 1, wherein: The data acquisition unit comprises: Data source management subunit: for maintaining and expanding the data source list, including tourism websites, social media platforms, Internet of Things devices, government tourism department data interfaces, and meteorological data interfaces, regularly evaluating the reliability and data quality of the data sources, and dynamically adjusting the collection priority of the data sources according to the demand; Crawler scheduling subunit: using distributed crawler technology to manage the task allocation of multiple crawler nodes, adopting incremental crawler strategy, recording the collected data, and only grabbing updated or new data to improve collection efficiency; User behavior acquisition subunit: in the process of user using the system, real-time collection of user behavior data, including search keywords, page browsing, link clicking, route collection, and scenic spot evaluation, real-time classification and preliminary arrangement of these behavior data, providing basis for subsequent analysis and recommendation.
3. The tourist route recommendation system using big data retrieval according to claim 1, wherein: The tourist route recommendation system using big data retrieval further comprises a user interface unit for user to input departure place, destination, travel time, budget, and display recommended tourist routes, and the interface has interactive function, user can evaluate and collect recommended routes, and the system can adjust the recommendation results according to user feedback.
4. The tourist route recommendation system using big data retrieval according to claim 1, wherein: In the data storage unit, structured data core information is stored by relational database storage structure, and semi-structured and unstructured data is stored by NoSQL database. 5.The travel route recommendation system employing big data search of claim 1, wherein, In the recommendation algorithm, the matching degree formula is adopted: ; wherein, represents the matching degree of user u and scenic spot i, is the weight of the user preference feature , and is the feature of scenic spot i, and n is the number of features, which is used to adjust the recommendation priority and weight in collaborative filtering and content-based recommendation. 6.The travel route recommendation system employing big data search of claim 5, wherein, In the data quality evaluation, the data outlier detection formula is adopted: ; wherein, represents the degree of abnormality of data point x, is the mean of the data, is the standard deviation, used to identify and treat abnormal data.
7. The travel route recommendation system employing big data search of claim 5, wherein, The user preference feature weight adopts the following calculation formula: ; wherein B represents the user behavior data quantitative value, S represents the seasonal influence factor of the tourist attraction, represents the user budget matching degree, respectively The corresponding adjustment coefficient is used for automatically adjusting the weight to improve the calculation accuracy. 8.The travel route recommendation system employing big data search of claim 7, wherein, The calculation method of user behavior data quantitative value B is: Search frequency normalization formula: ; wherein, is the normalized search frequency, is the search frequency of a user for a certain type of tourism element, is the maximum search frequency of all users in the system for this type of tourism element within a certain period of time; The browsing time standardization formula is: ; wherein, is the normalized browsing time, is the time of a user browsing a certain page, is the longest time of all users in the system browsing the page. Collection frequency normalization formula: ; wherein is the normalized collection frequency, is the collection frequency of the user, is the maximum collection frequency in the system; The like number standardization formula is: ; wherein is the normalized like number, is the like number of the user, is the maximum like number in the system; The comprehensive standardization formula of reservation and purchase behavior is: ; wherein is the comprehensive standardization value of reservation and purchase behavior, A is the reservation or purchase amount, is the maximum reservation or purchase amount in the system, F is the reservation or purchase frequency, is the maximum reservation or purchase frequency in the system; The total formula for calculating user behavior data quantitative value is: ; wherein B is a user behavior data quantification value, are weights of search behavior, browsing behavior, collection and like behavior, reservation and purchase behavior, respectively, and . 9.The travel route recommendation system employing big data search of claim 8, wherein, The calculation method of user behavior data quantitative value B is: the calculation method of tourism scenic spot seasonal influence factor S is: collect tourist flow or heat index data in different seasons for many years; Calculate the average tourist flow or heat index of each season and the whole year; calculate ; user budget matching degree The calculation method is: obtain the user budget and the cost C of the travel route or scenic spot; calculate the difference ; calculate the user budget matching degree by the formula . 10.A method for recommending a travel route using big data search, the method comprising: receiving a travel route request from a user; and providing a travel route recommendation to the user based on the travel route request. Comprise the following steps: Collecting target data, the target data is multi-source data containing tourism target keywords, the multi-source data contains tourist attractions, hotels, traffic information and user historical travel data and preferences; The target data is stored by using a distributed storage system, and the target data is cleaned and preprocessed, including deduplication, missing value filling and data format conversion; The stored data is cleaned, converted and aggregated by means of a big data processing framework, the preference type of the user is extracted from the historical travel data of the user, the feature of the scenic spot is extracted from the scenic spot data, the scenic spots are classified by using clustering analysis, the feature type of the scenic spot is obtained, and the matching degree is calculated according to the preference type of the user and the feature type of the scenic spot; The scenic spots with the matching degree exceeding a threshold value are selected as recommended scenic spots, and a travel route is generated by using route planning technology.