Travel route mining method based on signaling big data
Through cleaning and clustering, mobile signaling data is processed, combined with European-style distance hierarchical clustering method, tourism route mining is optimized, and the data support for traditional tourism route planning and scenic spot operation problems are solved, achieving more accurate and valuable Recommended travel routes.
Patent Information
- Application Number
- CN202110597996.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-05-31
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2041-05-31
AI Technical Summary
Traditional tourism route planning lacks comprehensive, accurate and dynamic scenic spot data support, resulting in overload operation of scenic spots, traffic congestion and low tourist satisfaction. The existing technology based on signaling big data has failed to consider the location, size and adjacent scenic spots of the scenic spot. Factors have led to limited reference value of the excavated tourist routes.
By obtaining the original mobile phone signaling data, the ping-pong effect noise, non-scenic base station connection data and the base station connection data of passers-by, scenic spot staff and permanent residents are removed, the base station-scenic spot table is constructed, and the preliminary tourist tour route is obtained by aggregating calculation, and the first tourist tour route is adopted. The European-style distance hierarchical clustering method combines similar scenic spots to optimize the results of tourism route mining.
It has achieved more accurate and valuable tourism route mining, ensured the accuracy of results and user experience, provided more accurate recommendations of hot tourist routes, and solved the problems of overload operation and traffic congestion in scenic spots.
Smart Images

Figure CN113313307B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent travel technology, and in particular to a travel route mining method based on signaling big data. Background Art
[0002] With the development of society and the improvement of people's living standards, tourism has become an increasingly important way of life. However, there are many problems in traditional tourism route planning and scenic spot management, such as the inability to share data, low data utilization, and the inability to accurately grasp the flow of people in scenic spots, resulting in a series of problems such as overloaded operation of scenic spots, traffic congestion, and low tourist satisfaction. Therefore, accurately grasping hot tourist routes has become an urgent problem to be solved. With the proposal of "Smart Earth" by IBM, a wave of research and application of smart cities and smart tourism has been brought about. Tourism route mining is one of the main contents of smart tourism. The purpose is to analyze tourist big data, identify popular tourist routes, provide reference for tourists, scenic spots and tourism management departments, and provide precise marketing support for tourism-related industries.
[0003] With the widespread popularization of mobile Internet and smart terminals, operators have accumulated a large amount of mobile phone signaling data. Mining tourist routes based on signaling big data has made it possible to accurately grasp tourists' travel behavior and scenic spot flow data. The main problem with traditional tourist route mining is the lack of comprehensive, accurate, and dynamic tourist attraction data support, which cannot solve a series of problems such as overloaded operation of the above-mentioned scenic spots, traffic congestion, and low tourist satisfaction. Recently, some tourist route mining technologies based on signaling big data have also emerged, but they do not take into account factors such as the location, size, and proximity of scenic spots, and the reference value of the mined tourist routes is limited. Summary of the invention
[0004] The present invention aims to solve one of the technical problems in the related art at least to a certain extent.
[0005] To this end, the purpose of the present invention is to propose a tourist route mining method based on signaling big data, which optimizes the effect of tourist route mining.
[0006] To achieve the above-mentioned purpose, an embodiment of the present invention proposes a tourist route mining method based on signaling big data, comprising the following steps: step S1, obtaining original mobile phone signaling data of a preset area, calculating the duration of each connection of a mobile phone to a base station, removing short-term connection noise data generated by the ping-pong effect in the original mobile phone signaling data, and obtaining a global mobile phone signaling data table; step S2, crawling scenic spot POI data of a preset area to obtain a scenic spot list, screening base stations within the scope of each scenic spot according to the scenic spot list and the base station list in the global mobile phone signaling data table, constructing a base station-scenic spot table, and then filtering the base station-scenic spot table according to the base station- The scenic area table removes the non-scenic area base station connection data in the global mobile phone signaling data table to obtain the scenic area mobile phone signaling data table; step S3, removes the noise data generated by passers-by, scenic area staff and nearby permanent residents in the scenic area mobile phone signaling data table to obtain the scenic area tourist mobile phone signaling data table; step S4, constructs a scenic area sequence in chronological order for the scenic area tourist mobile phone signaling data table, and performs clustering calculation to obtain a preliminary tourist tour route table; step S5, adopts a hierarchical clustering method based on Euclidean distance to merge adjacent scenic spots in the preliminary browsing route to obtain the optimal travel route.
[0007] The tourist route mining method based on signaling big data in an embodiment of the present invention removes ping-pong effect data, non-scenic area base station connection data, and base station connection data of passers-by, scenic area staff, and permanent residents in turn to ensure the accuracy of the results, and then constructs a preliminary scenic area tourist route based on the cleaned data, and merges and optimizes similar scenic spots based on hierarchical clustering of Euclidean distance, so that the tourist route finally mined is more accurate and gives users a better experience.
[0008] In addition, the tourist route mining method based on signaling big data according to the above embodiment of the present invention may also have the following additional technical features:
[0009] Further, in one embodiment of the present invention, the formula for calculating the time duration of the mobile phone connecting to the base station is:
[0010] lasttime(x)=endtime(x)-starttime(x)
[0011] Among them, lasttime(x) is the time when the mobile phone is connected to the base station, endtime(x) is the time when the mobile phone is disconnected from the base station, and starttime(x) is the time when the mobile phone starts to connect to the base station.
[0012] Furthermore, in one embodiment of the present invention, the global mobile phone signaling data table includes: date, user ID, base station, time when the mobile phone is connected to the base station, time when the mobile phone is disconnected from the base station, and duration of time the mobile phone is connected to the base station.
[0013] Furthermore, in one embodiment of the present invention, the scenic spot list includes the name of each scenic spot, the longitude and latitude of the center of the scenic spot, and the radius of the scenic spot, and the base station list includes the base station ID and the longitude and latitude of the base station.
[0014] Furthermore, in one embodiment of the present invention, the scenic spot (tourist) mobile phone signaling data table includes: date, user ID, base station ID, time when the mobile phone is connected to the base station, time when the mobile phone is disconnected from the base station, duration of mobile phone connection to the base station and scenic spot name.
[0015] Furthermore, in one embodiment of the present invention, in step S3, the scenic spot mobile phone signaling data table is clustered and calculated using user ID, scenic spot name and date as keywords, and the passers-by are removed by using the daily base station connection time less than a preset time threshold; the scenic spot tourist mobile phone signaling data table is clustered and calculated using user ID, scenic spot name and week as keywords, and the scenic spot staff and permanent residents whose frequency of appearance in the scenic spot tourist mobile phone signaling data table reaches a preset period are removed from their signaling data to obtain the scenic spot tourist mobile phone signaling data table.
[0016] Further, in one embodiment of the present invention, the step S4 specifically includes: aggregating the scenic spot tourist mobile phone signaling data table with user ID as the keyword to generate a scenic spot list sorted by start time; using the method of connecting and eliminating fields by regular expression, the start connection time of the base station and the scenic spot name in the scenic spot tourist mobile phone signaling data table are connected and sorted, and then the time is eliminated to obtain a tourist-scenic spot list; the tourist-scenic spot list includes a scenic spot sequence sorted by user ID and time; aggregating and calculating the tourist-scenic spot list with the scenic spot sequence as the keyword, counting the number of people in the same scenic spot list, sorting with the number of people as the keyword, and obtaining the preliminary tourist tour route table, the preliminary tourist tour route table includes the number of people on the route and the scenic spot sequence sorted by the number of people on the route.
[0017] Furthermore, in one embodiment of the present invention, the step S5 specifically includes: obtaining the scenic spot sequence according to the preliminary tourist tour route table; clustering each scenic spot as a separate cluster and initializing the scenic spot distance matrix; traversing the scenic spot sequence to calculate the Euclidean distance between any two scenic spots; when the distance matrix is not empty, selecting the minimum Euclidean distance, and judging whether the minimum Euclidean distance is less than or equal to a preset threshold, if so, merging the two scenic spots with the minimum Euclidean distance, and deleting the minimum Euclidean distance from the distance matrix, iterating the process until it is greater than the preset threshold, completing the scenic spot clustering, and obtaining the optimal browsing route.
[0018] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:
[0020] Figure 1 It is a flow chart of a travel route mining method based on signaling big data according to an embodiment of the present invention;
[0021] Figure 2 It is an execution flow chart of a tourist route mining method based on signaling big data according to an embodiment of the present invention;
[0022] Figure 3 is a flow chart of scenic spot clustering according to an embodiment of the present invention;
[0023] Figure 4 This is a base station information display diagram according to an embodiment of the present invention;
[0024] Figure 5 This is a display diagram of permanent residents in a scenic area according to an embodiment of the present invention;
[0025] Figure 6 This is a preliminary display diagram of the number of people on a weekend tour route obtained in one embodiment of the present invention;
[0026] Figure 7 This is a display diagram of scenic spot clustering results according to an embodiment of the present invention;
[0027] Figure 8 It is a display diagram of the best tour route table of one embodiment of the present invention;
[0028] Fig. 9 This is a display diagram of TOP10 tour routes according to an embodiment of the present invention. DETAILED DESCRIPTION
[0029] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.
[0030] The following describes a travel route mining method based on signaling big data proposed according to an embodiment of the present invention with reference to the accompanying drawings.
[0031] Figure 1 It is a flowchart of a travel route mining method based on signaling big data according to an embodiment of the present invention.
[0032] Figure 2It is an execution flow chart of a travel route mining method based on signaling big data according to an embodiment of the present invention.
[0033] like Figure 1 and 2 As shown, the travel route mining method based on signaling big data includes the following steps:
[0034] In step S1, the original mobile phone signaling data of the preset area is obtained, the duration of each mobile phone connection to the base station is calculated, and the short-term connection noise data generated by the ping-pong effect in the original mobile phone signaling data is removed to obtain a global mobile phone signaling data table.
[0035] It is understandable that in a mobile communication system, if a mobile phone is in a position where the signals of two base stations overlap, it will switch back and forth between the two base stations, resulting in the so-called "ping-pong effect", causing the two base stations to generate a large amount of short-term connection signaling data, resulting in a large error. In order to reduce the impact of this short-term connection on subsequent calculations and remove this type of noise data, it is necessary to calculate the duration of each mobile phone connection, and obtain a tourist mobile phone signaling data table including user ID, base station ID, time when the mobile phone starts connecting to the base station, time when the mobile phone disconnects from the base station, and time the mobile phone connects to the base station. The formula for calculating the time the mobile phone connects to the base station is:
[0036] lasttime(x)=endtime(x)-starttime(x)
[0037] Among them, lasttime(x) is the time when the mobile phone is connected to the base station, endtime(x) is the time when the mobile phone is disconnected from the base station, and starttime(x) is the time when the mobile phone starts to connect to the base station.
[0038] It should be noted that data with a duration of less than 1 minute, that is, data with a timestamp less than 60000, can be considered as noise data generated by the ping-pong effect, and this part of the data is removed.
[0039] In step S2, the scenic spot POI data of the preset area is crawled to obtain a scenic spot list, and the base stations within the range of each scenic spot are screened according to the scenic spot list and the base station list in the global mobile phone signaling data table to construct a base station-scenic spot table, and then the non-scenic spot base station connection data in the global mobile phone signaling data table are removed according to the base station-scenic spot table to obtain the scenic spot mobile phone signaling data table;
[0040] Specifically, step S2 may include the following steps:
[0041] When constructing the scenic spot list, it is considered that scenic spots can be divided into small scenic spots in the city (such as churches, museums, etc.), medium scenic spots in the city (blocks, parks, etc.), and large scenic spots outside the city. Small scenic spots in the city are usually covered by a single base station, so the radius of the scenic spot can be set to 500m, and the radius of the medium scenic spot can be set to within 1km; for suburban natural scenic spots, due to the large area and no interference with the surrounding scenic spots, in order to ensure that all tourists entering the scenic spot can be counted, the scenic spot center is selected as the scenic spot range within 2km. Therefore, the scenic spot list includes the name of the scenic spot, the longitude and latitude of the scenic spot center, and the radius of the scenic spot.
[0042] Traverse the base station list in the global mobile phone signaling data table, determine whether each base station is covered by a scenic spot in the scenic spot list according to the base station longitude and latitude, filter out the base stations within the range of each scenic spot, realize the binding of the base station with the corresponding scenic spot, and build a base station-scenic spot list. The base station list includes the base station ID and the base station longitude and latitude, and the base station-scenic spot table includes the base station ID and the scenic spot name.
[0043] The specific algorithm can be as follows:
[0044] Input: scenic spot list, base station list
[0045] Output: Base station-scenic area table
[0046] Step 1:
[0047] Traverse the base station list
[0048] Browse the scenic spot list
[0049] If the base station latitude and longitude are within the scenic area
[0050] Bind the base station to the scenic spot and write it into the base station-scenic spot table.
[0051] Step 2:
[0052] Return to Base Station-Scenic Area Table
[0053] Then, the base station-scenic area table is connected to the global mobile phone signaling data table with the ping-pong effect removed, using the base station ID as the keyword, to obtain a signaling data table only within the scenic area, namely, the scenic area mobile phone signaling data table. The table includes date, user ID, base station ID, start connection time, end connection time, connection duration and scenic area name, as shown in Table 1 below.
[0054] Table 1 Scenic spot tourists mobile phone signaling data table structure
[0055]
[0056] At this time, the scenic area's mobile phone signaling data table still cannot be used directly for subsequent calculations. It is also necessary to screen passers-by, scenic area staff and permanent residents, remove data interference from non-tourist groups, and ensure the accuracy of the results.
[0057] In step S3, the noise data generated by passers-by, scenic area staff and nearby permanent residents in the scenic area mobile phone signaling data table are removed to obtain the scenic area tourist mobile phone signaling data table.
[0058] Furthermore, in one embodiment of the present invention, in step S3, the scenic spot signaling data table is clustered and calculated using user ID, scenic spot name and date as keywords, and passers-by are removed by using the daily base station connection time less than a preset time threshold; the scenic spot mobile phone signaling data table is clustered and calculated using user ID, scenic spot name and week as keywords, and scenic spot staff and permanent residents who appear in the scenic spot mobile phone signaling data table within a preset time range are removed.
[0059] Specifically, the embodiment of the present invention takes into account the fact that pedestrians passing through a scenic spot have a short time, which is reflected in the mobile phone signaling as a short total base station connection time.
[0060] Using user ID, scenic spot name and date as keywords, we perform aggregation calculations and use the following screening methods:
[0061] sum(lasttime)≥1800000
[0062] Even if the total connection time of the mobile phone to the base station in the scenic area exceeds 30 minutes, the passers-by can be removed with a high probability.
[0063] Secondly, it is necessary to identify the local residents and scenic spot staff at the scenic spot, and then delete the mobile phone signaling data of these people from the scenic spot signaling data table to ensure the accuracy of tourist route mining.
[0064] It is understandable that the identification of permanent residents and staff members is carried out by counting the number of days they appear around the scenic area in a week. For ordinary tourists, there is only a very small probability that they will stay in the same scenic area for multiple days in a week.
[0065] The user ID, scenic spot name and week (which can be obtained by date conversion) are used as keywords for aggregation calculation, and the following screening method is used:
[0066] count(day)≥3
[0067] If the user appears in the scenic area for more than three days a week in the observation time window, he or she can be considered to be a nearby resident or scenic area staff, and the selected time avoids holidays to avoid the problem of holiday staff being unable to be included in the statistics.
[0068] After the above processing, the final scenic spot tourist mobile phone signaling data table is obtained.
[0069] In step S4, a scenic spot sequence is constructed based on the scenic spot tourist mobile phone signaling data table in chronological order, and aggregation calculation is performed to obtain a preliminary tourist tour route table.
[0070] Furthermore, in one embodiment of the present invention, step S4 specifically includes:
[0071] The mobile phone signaling data table of scenic spot tourists is processed with user ID as the keyword to generate a tourist-scenic spot list sorted by start time, the tourist-scenic spot list includes the user ID and the scenic spot sequence sorted by time, and then the tourist-scenic spot list is clustered and calculated with the scenic spot sequence as the keyword, the number of people in the same scenic spot sequence is counted, and the list is sorted with the number of people as the keyword to obtain a preliminary tourist tour route table, which includes the route number and the scenic spot list sorted by time.
[0072] It should be noted that the scenic spot sequence needs to be sorted according to time. Therefore, after sorting using the collect_list function of Hive SQL, the embodiment of the present invention uses a method of adding and removing fields using regular expressions to ensure that the scenic spot sequence is arranged in chronological order. Specifically, the time is connected to the string, and then the obtained string is sorted in lexicographic order. Since the time is a number in front, the lexicographic order at this time is the time sorting. The Hive regular expression replacement operation is also used to identify numbers using regular expressions, and the time part is replaced with an empty string to obtain a scenic spot sequence sorted in chronological order. As shown in Table 2 below, the tourist-scenic spot list includes a user ID and a scenic spot sequence sorted by time.
[0073] Table 2 Tourist-attraction list
[0074]
[0075] Cluster calculation is performed using the scenic spot list as the keyword, the number of people in the same scenic spot list is counted, and sorted using the number of people as the keyword to obtain a preliminary tourist route list, including the keyword scenic spot list and the number of people.
[0076] Table 3 Tour route table structure
[0077]
[0078] In step S5, a hierarchical clustering method based on Euclidean distance is used to merge adjacent scenic spots in the preliminary browsing route to obtain the optimal travel route.
[0079] It is understandable that since the initial mining results contain a large number of different routes to similar scenic spots, considering that the order of visiting similar scenic spots is meaningless, it is chosen to intelligently cluster similar scenic spots, merge similar scenic spots, and optimize the mining results of tourist routes. In the embodiment of the present invention, a hierarchical clustering algorithm based on Euclidean distance is used to cluster scenic spots whose longitude and latitude differences are within 0.015.
[0080] Step S5 specifically includes:
[0081] Acquire the scenic spot sequence according to the preliminary tourist tour route table;
[0082] Treat each scenic spot as a separate cluster and initialize the scenic spot distance matrix;
[0083] Traversing the scenic spot sequence to calculate the Euclidean distance between any two scenic spots;
[0084] When the distance matrix is not empty, the minimum Euclidean distance is selected and it is determined whether the minimum Euclidean distance is less than or equal to the preset threshold. If so, the two scenic spots with the minimum Euclidean distance are merged and the minimum Euclidean distance is deleted from the distance matrix. The process is iterated until it is greater than the preset threshold, and the scenic spot clustering is completed. Figure 3 shown.
[0085] The specific algorithm processing flow is as follows:
[0086] Input: scenic spot list x
[0087] Output: Clustering of similar scenic spots
[0088] Step 1:
[0089] Get the list of scenic spots;
[0090] Treat each scenic spot as a separate cluster, and the parent node of each scenic spot is its own
[0091] Initialize the scenic spot distance matrix distance
[0092] Step 2:
[0093] Initialize the minimum heap q
[0094] Traverse all scenic spots in x for i and j
[0095] Calculate the Euclidean distance distance(i,j)
[0096] Put the triple (i, j, distance (i, j)) into the minimum heap q
[0097] Step 3:
[0098] While the minimum heap q is not empty
[0099] p = q top element of the heap
[0100] If the minimum distance exceeds the threshold, the loop ends.
[0101] If the cluster where px is located is not equal to the cluster where py is located
[0102] Merge the cluster where px is located with the cluster where py is located
[0103] q Pop the top element of the heap
[0104] Output scenic spot clustering results
[0105] It is possible to consider selecting scenic spots within a threshold of 1.5 kilometers as similar scenic spots, that is, scenic spots with a longitude and latitude distance within 0.015 as similar scenic spots.
[0106] The scenic spots with the same clustering results are considered as the same scenic spot, and the tourist route analysis results are recalculated to obtain the optimal tourist route list.
[0107] The travel route mining method based on signaling big data in the embodiment of the present invention is verified by a specific example below.
[0108] In order to verify the effectiveness of the method of the present invention, the signaling big data of Central Street in Harbin in a certain week was selected for verification.
[0109] Select a specific scenic spot and you can see the basic information of base stations around the scenic spot, including the total number of base station connections per day, the number of connections per base station, and the total number of base station connections per day after removing the ping-pong effect, and the number of connections per base station. After removing the ping-pong effect, the number of daily connections of each base station is greatly reduced, greatly reducing the negative impact of the ping-pong effect. Figure 4 As shown, the base station information of a certain scenic spot on January 11, 2020 is taken as an example.
[0110] For permanent residents screening, select a specific scenic spot from the drop-down list to see a list of permanent residents around the scenic spot. Figure 5 As shown, taking the list of residents near a certain scenic spot as an example, the resident’s mobile phone number is replaced by *.
[0111] In terms of tourist route analysis, the number of people in the same scenic spot list is counted for the tourist route mining results, and the number of people is sorted as the keyword to obtain a preliminary tourist route list. The results obtained on a certain weekend are as follows: Figure 6 shown.
[0112] In terms of scenic spot clustering analysis, scenic spots within 1.5 kilometers are selected as similar scenic spots, that is, scenic spots with a latitude and longitude distance difference of less than 0.015 are selected as similar scenic spots. The hierarchical clustering algorithm based on Euclidean distance is used to merge similar scenic spots. The clustering results are as follows: Figure 7 shown.
[0113] The scenic spot clustering results and the preliminary tourist route table are recalculated to obtain the final tourist route mining results, such as Figure 8 Shown are the tour routes and number of people for a particular week.
[0114] The top 10 tour routes and the number of visitors can also be displayed in a horizontal bar chart, which can more intuitively display the results of tour route mining, such as Fig. 9 shown.
[0115] Therefore, the tourist route mining method based on signaling big data proposed in the embodiment of the present invention establishes a tourist route system, which can rank the popularity of different routes at the same time and achieve the expected purpose.
[0116] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.
[0117] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.
[0118] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.
Claims
1. A tourist route mining method based on signaling big data, characterized in that: The following steps are involved: Step S1, obtaining the original mobile phone signaling data of the preset area, calculating the duration of each mobile phone connection to the base station, removing the short-term connection noise data generated by the ping-pong effect in the original mobile phone signaling data, and obtaining a global mobile phone signaling data table; wherein, the data of the duration of the mobile phone connection to the base station of less than 1 minute is used as the short-term connection noise data generated by the ping-pong effect; the formula for calculating the duration of the mobile phone connection to the base station is: lasttime(x)=endtime(x)-starttime(x) Among them, lasttime(x) is the time the mobile phone is connected to the base station, endtime(x) is the time when the mobile phone is disconnected from the base station, and starttime(x) is the time when the mobile phone starts to connect to the base station; Step S2, crawling scenic spot POI data of a preset area to obtain a scenic spot list, screening base stations within the scope of each scenic spot according to the scenic spot list and the base station list in the global mobile phone signaling data table, constructing a base station-scenic spot table, and then removing non-scenic spot base station connection data in the global mobile phone signaling data table according to the base station-scenic spot table to obtain a scenic spot mobile phone signaling data table; the scenic spot list includes the name of each scenic spot, the longitude and latitude of the scenic spot center, and the radius of the scenic spot, and the base station list includes the base station ID and the longitude and latitude of the base station; Step S3, removing the noise data generated by passers-by, scenic area staff and nearby permanent residents in the scenic area mobile phone signaling data table, and obtaining the scenic area tourist mobile phone signaling data table; specifically: using user ID, scenic area name and date as keywords to perform cluster calculation on the scenic area mobile phone signaling data table, using the daily base station connection time less than the preset time value to remove the passers-by; using user ID, scenic area name and week as keywords to perform cluster calculation on the scenic area tourist mobile phone signaling data table, removing the signaling data of the scenic area staff and permanent residents in the scenic area tourist mobile phone signaling data table that reaches the frequency of occurrence in the preset period, and obtaining the scenic area tourist mobile phone signaling data table; the frequency of occurrence in the preset period is more than three days a week; Step S4, constructing a scenic spot sequence from the scenic spot tourist mobile phone signaling data table in chronological order, and performing clustering calculation to obtain a preliminary tourist tour route table; specifically: clustering the scenic spot tourist mobile phone signaling data table with the user ID as the keyword to generate a scenic spot sequence sorted by the start time; using the method of connecting and eliminating fields in the scenic spot tourist mobile phone signaling data table by regular expression, connecting and sorting the start connection base station time and the scenic spot name, and then eliminating the time to obtain a tourist-scenic spot list; the tourist-scenic spot list includes the scenic spot sequence sorted by user ID and time; then clustering the tourist-scenic spot list with the scenic spot sequence as the keyword, counting the number of people in the same scenic spot sequence, sorting with the number of people as the keyword, and obtaining a preliminary tourist tour route table; the preliminary tourist tour route table includes the scenic spot sequence sorted by the number of people on the route and time; Step S5, using a hierarchical clustering method based on Euclidean distance to merge adjacent scenic spots in the preliminary tourist tour route table to obtain an optimal tour route; The step S5 specifically includes: Acquire the scenic spot sequence according to the preliminary tourist tour route table; Treat each scenic spot as a separate cluster and initialize the scenic spot distance matrix; Traversing the scenic spot sequence to calculate the Euclidean distance between any two scenic spots; When the distance matrix is not empty, the minimum Euclidean distance is selected, and it is determined whether the minimum Euclidean distance is less than or equal to a preset threshold. If so, the two scenic spots with the minimum Euclidean distance are merged, and the minimum Euclidean distance is deleted from the distance matrix. The process is iterated until it is greater than the preset threshold, the scenic spot clustering is completed, and the optimal travel route is obtained.
2. The tourist route mining method based on signaling big data according to claim 1 is characterized in that: The global mobile phone signaling data table includes: date, user ID, base station ID, time when the mobile phone is connected to the base station, time when the mobile phone is disconnected from the base station, and duration of time when the mobile phone is connected to the base station.
3. The tourist route mining method based on signaling big data according to claim 1 is characterized in that: The scenic spot tourist mobile phone signaling data table includes: date, user ID, base station ID, time when the mobile phone is connected to the base station, time when the mobile phone is disconnected from the base station, duration of mobile phone connection to the base station and scenic spot name.
