Method and device for determining permanent residence

Abstract

The invention provides a permanent residence determination method and device, and belongs to the technical field of data processing, and the method comprises the steps: analyzing the permanent residence of a passenger from multiple dimensions according to the flight travel record of the passenger, obtaining a multi-dimensional judgment result of the permanent residence, and determining the permanent residence; wherein the permanent residence multi-dimensional judgment result comprises multiple items of a flight frequency judgment result, a holiday and festival judgment result, a travel judgment result, a stay duration judgment result or an identity card judgment result. According to the method, various factors and business characteristics are comprehensively considered, and at least two of five dimension analysis including flight frequency judgment, holiday and festival judgment, travel judgment, stay duration judgment or identity card judgment is carried out on the flight travel record of the passenger, so that the method can adapt to different application scenes; and finally determining the permanent residence of the passenger by using the city which may be used as the permanent residence of the passenger and the corresponding confidence result, so that the permanent residence of the passenger can be accurately determined without obtaining the total position information of the passenger.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and device for determining a permanent residence. Background Art

[0002] The determination of the passenger’s permanent residence is very important for civil aviation enterprises to provide targeted services and personalized product design to passengers.

[0003] At present, the common method to determine the passenger's permanent residence is to use the buried point analysis method to obtain the passenger's longitude and latitude information when the passenger performs the event operation of the flight travel, and use the preset map library to match the longitude and latitude information to the address, and then use the obtained address information as the passenger's permanent residence.

[0004] However, when determining the permanent residence of a passenger, the above method first needs to obtain the full location information of the passenger, such as longitude and latitude information. If the full location information of the passenger cannot be obtained, the permanent residence of the passenger cannot be determined. In addition, determining the permanent residence of a passenger based on longitude and latitude information is a single-threaded logic that does not comprehensively consider various factors, lacks business support, and cannot adapt to different application scenarios. As a result, the address determined based on longitude and latitude information may not be the permanent residence of the passenger, but the passenger's travel destination, which in turn leads to inaccurate results of the determined permanent residence of the passenger. Summary of the invention

[0005] The present invention provides a method and device for determining a permanent residence, which are used to solve the defects in the prior art that the permanent residence of a passenger cannot be determined when the longitude and latitude information of the passenger cannot be obtained, and that the permanent residence of the passenger is determined only based on the longitude and latitude information, resulting in inaccurate results, so that the permanent residence of the passenger can be accurately determined without obtaining the full location information of the passenger.

[0006] In a first aspect, the present invention provides a method for determining a permanent residence, comprising: According to the flight travel record of the passenger, the passenger's permanent residence is analyzed from multiple dimensions to obtain a multi-dimensional determination result of the passenger's permanent residence; Determine the passenger's permanent residence based on the multi-dimensional determination result of the permanent residence; The multi-dimensional determination result of the permanent residence includes multiple items of the flight number determination result, holiday determination result, business trip determination result, stay duration determination result or ID card determination result.

[0007] According to a method for determining a permanent residence provided by the present invention, when the multi-dimensional determination result of the permanent residence includes the flight number determination result, the flight number determination result is determined based on the following method: Determine each departure city and its corresponding departure times, and each arrival city and its corresponding arrival times according to the flight travel records; Taking all the departure cities and all the arrival cities as candidate permanent residences, a first candidate permanent residence set is constructed; determining a confidence level of each of the candidate permanent residences in the first candidate permanent residence set being judged as a permanent residence as the flight number determination result; The confidence level of each candidate permanent residence being judged as a permanent residence is determined based on the number of flights and the total number of flights of the candidate permanent residence. The number of flights is determined based on the number of departures and arrivals of the candidate permanent residence. The total number of flights is determined based on the total number of departures and the total number of arrivals.

[0008] According to a method for determining a permanent residence provided by the present invention, when the multi-dimensional determination result of the permanent residence includes the holiday determination result, the holiday determination result is determined based on the following method: Obtaining the first flight travel record of the passenger in each holiday time period, wherein each holiday time period is determined according to a holiday time period data set pre-retrieved from a third-party database; According to the first flight travel record, determine the departure city of each first flight and its corresponding first flight departure number, and determine the arrival city of each last flight and its corresponding last flight arrival number; Taking all the first flight departure cities and all the last flight arrival cities as candidate permanent residences, a second candidate permanent residence set is constructed; determining a confidence level of each of the candidate permanent residences in the second candidate permanent residence set being judged as a permanent residence as the holiday determination result; The confidence level of each candidate permanent residence being judged as the permanent residence is determined based on the number of holiday flights and the total number of holiday flights of the candidate permanent residence, the number of holiday flights is determined based on the number of first flight departures and the number of last flight arrivals of the candidate permanent residence, the total number of holiday flights is determined based on the total number of first flight departures and the total number of last flight arrivals; The first flight departure city is the departure city in the passenger's first flight record in the first half of the holiday time period, and the last flight arrival city is the arrival city in the passenger's last flight record in the second half of the holiday time period.

[0009] According to a method for determining a permanent residence provided by the present invention, whether a passenger is a student is determined based on the passenger's age when traveling on a flight; If the passenger is a student, the holiday time period is determined as the student holiday time period.

[0010] According to a method for determining a permanent residence provided by the present invention, when the multi-dimensional determination result of the permanent residence includes a travel determination result, the travel determination result is determined based on the following method: Obtain the passenger's second flight travel record during non-holiday periods; Determine, from the second flight travel record, a continuous business trip itinerary whose first business trip flight and / or last business trip flight meets the discount rate condition; the continuous business trip itinerary is a continuous trip in which the passenger stays at the travel destination for less than a preset time; According to the continuous business trip itinerary, determine the departure city of each first business trip and its corresponding first business trip departure number, and determine the arrival city of each last business trip and its corresponding last business trip arrival number; Taking all the departure cities of the first business trip and all the arrival cities of the last business trip as candidate permanent residences, a third candidate permanent residence set is constructed; Determining the confidence level of each candidate permanent residence in the third candidate permanent residence set being judged as a permanent residence as the business trip determination result; The confidence level of each candidate permanent residence being judged as the permanent residence is determined based on the number of business travel flights and the total number of business travel flights of the candidate permanent residence. The number of business travel flights is determined based on the number of first business travel departures and the number of last business travel arrivals of the candidate permanent residence. The total number of business travel flights is determined based on the total number of first business travel departures and the total number of last business travel arrivals.

[0011] According to a method for determining a permanent residence provided by the present invention, when the multi-dimensional determination result of the permanent residence includes a determination result of the length of stay, the determination result of the length of stay is determined based on the following method: Determine the effective length of stay of the passenger in each arrival city based on the flight travel record; Taking all the arrival cities as candidate permanent residences, a fourth candidate permanent residence set is constructed; Determining the confidence level of each candidate permanent residence in the fourth candidate permanent residence set being judged as a permanent residence as the stay duration determination result; The confidence level of each candidate permanent residence being judged as the permanent residence is determined based on the effective stay time of the candidate permanent residence and the total effective stay time, where the total effective stay time is the sum of the effective stay times of all the candidate permanent residences.

[0012] According to a method for determining a permanent residence provided by the present invention, the method of determining the permanent residence of the passenger based on the multi-dimensional determination result of the permanent residence includes: According to the multi-dimensional determination results of each permanent residence and its corresponding probability weight, the confidence of each candidate permanent residence is weighted and summed to obtain the comprehensive confidence of each candidate permanent residence; the probability weight is determined according to the product of the weight coefficient and the gradient coefficient; Determining the permanent residence of the passenger, the permanent residence being the candidate permanent residence with the greatest comprehensive confidence; The weight coefficient is obtained by training based on the multi-dimensional determination results of the permanent residence of multiple samples and their corresponding permanent residence labels.

[0013] According to a method for determining a permanent residence provided by the present invention, the gradient coefficients corresponding to the flight number determination result and the holiday determination result are both set to the first gradient coefficient; the gradient coefficients corresponding to the business trip determination result and the stay duration determination result are set to the second gradient coefficient; the gradient coefficient corresponding to the ID card determination result is set to the third gradient coefficient; The first gradient coefficient is greater than the second gradient coefficient, and the second gradient coefficient is greater than the third gradient coefficient.

[0014] According to a method for determining a permanent residence provided by the present invention, after weighted summing up the confidence of each candidate permanent residence according to the multi-dimensional determination result of each permanent residence and its corresponding probability weight, the method further includes: Calculating a total confidence, where the total confidence is the sum of the comprehensive confidences of all candidate permanent residences; For any candidate permanent residence, a new comprehensive confidence of the candidate permanent residence is determined based on the ratio of the comprehensive confidence of the candidate permanent residence to the total confidence, so as to achieve confidence normalization processing.

[0015] In a second aspect, the present invention further provides a device for determining a permanent residence, comprising: A multi-dimensional determination unit, configured to analyze the passenger's permanent residence from multiple dimensions according to the passenger's flight travel record, and obtain a multi-dimensional determination result of the passenger's permanent residence; A permanent residence determination unit, configured to determine the permanent residence of the passenger based on the multi-dimensional determination result of the permanent residence; The multi-dimensional determination result of the permanent residence includes multiple items of the flight number determination result, holiday determination result, business trip determination result, stay duration determination result or ID card determination result.

[0016] The method and device for determining the permanent residence provided by the present invention comprehensively consider various factors and business characteristics, and can adapt to different application scenarios by performing at least two of the five dimensional analyses of the passenger's flight travel records, namely, flight number determination, holiday determination, business trip determination, stay duration determination, or ID card determination. The permanent residence of the passenger is finally determined by using the city that may be the passenger's permanent residence and the corresponding confidence results, and the permanent residence of the passenger can be accurately determined without obtaining the passenger's full location information. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 This is one of the flow charts of the method for determining a permanent residence provided by the present invention.

[0019] Figure 2 This is the second flow chart of the method for determining the permanent residence provided by the present invention.

[0020] Figure 3 It is a structural schematic diagram of the device for determining a permanent residence provided by the present invention.

[0021] Figure 4 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] It should be noted that, in the description of the present invention, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0024] The terms "first", "second", etc. in the present invention are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present invention can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are generally of the same type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" means at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0025] Combine the following Figure 1-Figure 4 The method and device for determining the permanent residence provided by the present invention are described. It should be noted that all actions of obtaining signals, information or data in the present invention are carried out under the premise of complying with the corresponding data protection laws and policies of the country of location and with the authorization given by the owner of the corresponding device.

[0026] It should be noted that the executor of the method for determining the permanent residence provided by the present invention is the corresponding permanent residence determination device, specifically including but not limited to any one of servers, computer equipment, mobile phones, tablet computers, laptop computers, PDAs, vehicle-mounted electronic devices, wearable devices, ultra-mobile personal computers (Ultra-Mobile Personal Computer, UMPC), netbooks or personal digital assistants (Personal Digital Assistant, PDA) and other devices.

[0027] After statistical processing and information mining, civil aviation passenger information can obtain passenger group label information, that is, the distribution of different characteristics of a certain part of passengers. The contemporary market emphasizes targeted services and personalized product design. Passenger group labels, as statistical data, can better reveal passenger needs and market trends, which is conducive to formulating more accurate marketing strategies and product optimization and improvement plans. At the same time, it can also improve the level of passenger travel safety management and provide strong support for the sustainable development of the tourism industry.

[0028] The group label product first collects and summarizes user attribute information such as gender, age, cabin, and fare; user-generated information such as ticket issuance, take-off, and arrival time; flight schedule information such as flight schedule changes, aircraft model changes, and flight cancellations; flight dynamic information such as check-in opening, flight take-off, flight diversion, and flight mergers. Then, the data is processed using the set logic to form passenger group labels such as passenger age distribution, number of companions distribution, and source distribution. Passenger group labels analyze passenger transportation travel behavior, classify and count passengers in layers, and define them according to overall market information, basic passenger information, traffic flow, consumption level, travel preferences, and transit dimensions. It accurately predicts passenger flow based on different needs, helping companies to gain an in-depth understanding of customer characteristics and identify source markets.

[0029] The passenger's permanent residence is an important field in the group label. Accurately determining the passenger's permanent residence is very important for civil aviation companies to provide targeted services and personalized product design to passengers.

[0030] At present, the common method to determine the passenger's permanent residence is to use the buried point analysis method to obtain the passenger's longitude and latitude information when the passenger performs the event operation of the flight travel, and use the preset map library to match the longitude and latitude information to the address, and then use the obtained address information as the passenger's permanent residence.

[0031] However, when determining the permanent residence of a passenger, the above method first needs to obtain the full location information of the passenger, such as longitude and latitude information. If the full location information of the passenger cannot be obtained, the permanent residence of the passenger cannot be determined. In addition, determining the permanent residence of a passenger based on longitude and latitude information is a single-threaded logic that does not comprehensively consider various factors, lacks business support, and cannot adapt to different application scenarios. As a result, the address determined based on longitude and latitude information may not be the permanent residence of the passenger, but the passenger's travel destination, which in turn leads to inaccurate results of the determined permanent residence of the passenger.

[0032] In view of this, the present invention provides a method and device for accurately determining the permanent residence of a passenger through multi-dimensional analysis by comprehensively considering various factors and business characteristics without obtaining the full location information of the passenger.

[0033] Figure 1 is one of the flow charts of the method for determining the permanent residence provided by the present invention, such as Figure 1 As shown, the steps include but are not limited to the following steps 101 and 102.

[0034] Step 101: Analyze the passenger's permanent residence from multiple dimensions based on the passenger's flight travel record to obtain a multi-dimensional determination result of the passenger's permanent residence.

[0035] The multi-dimensional determination results of the permanent residence include, but are not limited to, multiple results such as the number of flights, holidays, business trips, length of stay, or ID card determination results.

[0036] Among them, the flight number determination result includes but is not limited to the city of the passenger's permanent residence determined based on the passenger's flight number and the corresponding confidence level; the holiday determination result includes but is not limited to the city of the passenger's permanent residence determined based on the passenger's flight travel records during the holiday period and the corresponding confidence level; the business trip determination result includes but is not limited to the city of the passenger's permanent residence determined based on the passenger's business trip itinerary during non-holiday period and the corresponding confidence level; the stay duration determination result includes but is not limited to the city of the passenger's permanent residence determined based on the length of time the passenger stays in different cities and the corresponding confidence level; the ID card determination result includes but is not limited to the city of the passenger's permanent residence determined based on the passenger's ID card number and the corresponding confidence level.

[0037] Flight travel records are a type of civil aviation passenger information, which includes but is not limited to one or more of passenger information, flight information, ticket information, itinerary-related information, reservation and confirmation information, payment information, etc.

[0038] Among them, passenger information includes but is not limited to one or more of the passenger's name, ID number, contact information, etc.; flight information includes but is not limited to one or more of the flight number, airline name, departure city and arrival city, departure time and arrival time, etc.; air ticket information includes but is not limited to one or more of the ticket number, ticket discount rate, fare, seat number, cabin class, etc.; itinerary-related information includes but is not limited to one or more of the transfer information, baggage check-in information, passenger dining information, etc.; reservation and confirmation information includes but is not limited to one or more of the reservation date, confirmation number, reservation status, etc.; payment information includes but is not limited to one or more of the payment method, transaction amount and payment date, etc.

[0039] Specifically, after obtaining the passenger's flight travel record from the database, the passenger's permanent residence is analyzed from at least two of the five dimensions of flight number determination, holiday determination, business trip determination, length of stay determination or ID card determination based on the passenger's flight travel record, and a multi-dimensional determination result of the permanent residence is obtained, which includes at least two of the flight number determination results, holiday determination results, business trip determination results, length of stay determination results or ID card determination results, thereby determining several cities that may be the passenger's permanent residence and their corresponding confidence levels in each dimension.

[0040] It should be noted that the passenger's flight travel records used as the basis for analyzing the passenger's permanent residence from multiple dimensions can be the passenger's flight travel records from the time the passenger registered his / her account to the date of permanent residence analysis, or can be the passenger's permanent residence in a specific time period, and the passenger's permanent residence can be analyzed based on the passenger's flight travel records in the specific time period.

[0041] In one embodiment, after obtaining the flight travel records of passengers from the database, data cleaning is performed to fill in empty values ​​and correct abnormal values ​​for abnormal flight travel records.

[0042] Take the flight travel record as an example, which includes the passenger name, flight number, departure time, arrival time, departure city, and arrival city. When the departure time in a flight travel record of passenger A is obtained from the database as an abnormal value of null or garbled, the flight travel records with the same arrival time and departure city are searched from the flight travel records of other passengers in the database according to the arrival time and departure city in the flight travel record, as the flight travel records of the same itinerary. Then, the departure time value in the flight travel record of the same itinerary is used to complete or correct the departure time value in the flight travel record of passenger A.

[0043] Step 102: Determine the permanent residence of the passenger based on the multi-dimensional determination result of the permanent residence.

[0044] Specifically, after conducting a multi-dimensional analysis of the passenger's flight travel records and obtaining a multi-dimensional determination result of the permanent residence, the passenger's permanent residence is finally determined based on a preliminary analysis of several cities that may be the passenger's permanent residence and their corresponding confidence levels in each dimension.

[0045] The method for determining the permanent residence provided by the present invention comprehensively considers various factors and business characteristics, and can adapt to different application scenarios by performing at least two of the five dimensional analyses of the passenger's flight travel records, namely, flight number determination, holiday determination, business trip determination, length of stay determination, or ID card determination. The method can also use the cities that may be the passenger's permanent residence and the corresponding confidence results to ultimately determine the passenger's permanent residence, and can accurately determine the passenger's permanent residence without obtaining the passenger's full location information.

[0046] In addition, the method for determining the permanent residence provided by the present invention does not require the collection of a large amount of spatiotemporal information of the user, but rather makes a comprehensive judgment based on a small number of civil aviation travel records combined with business experience. It is a low-cost method that guarantees accuracy and is suitable for operation by civil aviation data owners.

[0047] Based on the above embodiment, as an optional embodiment, when the multi-dimensional determination result of the permanent residence includes the flight number determination result, the flight number determination result is determined based on the following method: Determine each departure city and its corresponding departure times, and each arrival city and its corresponding arrival times according to the flight travel records; Taking all the departure cities and all the arrival cities as candidate permanent residences, a first candidate permanent residence set is constructed; determining a confidence level of each of the candidate permanent residences in the first candidate permanent residence set being judged as a permanent residence as the flight number determination result; The confidence level of each candidate permanent residence being judged as a permanent residence is determined based on the number of flights and the total number of flights of the candidate permanent residence. The number of flights is determined based on the number of departures and arrivals of the candidate permanent residence. The total number of flights is determined based on the total number of departures and the total number of arrivals.

[0048] Specifically, when determining the number of flights, the passenger's flight travel records are first obtained from the database, and tool software is used to determine each departure city and its corresponding number of departures in the passenger's flight travel records, as well as each arrival city and its corresponding number of arrivals. The total number of departures is determined based on the number of departures of all departure cities in the flight travel records, and the total number of arrivals is determined based on the number of arrivals of all arrival cities. Then, the total number of flights of the passenger is determined based on the total number of departures and the total number of arrivals.

[0049] Furthermore, all departure cities and all arrival cities in the flight travel record are taken as candidate permanent residences of the passenger to construct a first candidate permanent residence set.

[0050] For each candidate permanent residence in the first candidate permanent residence set, the number of flights of the candidate permanent residence is determined according to the sum of its corresponding number of departures and arrivals, and then the confidence of the candidate permanent residence being judged as the permanent residence is determined according to the number of flights of the candidate permanent residence and the total number of flights. After the confidence of all candidate permanent residences in the first candidate permanent residence set being judged as permanent residences is determined, the flight number determination result can be obtained.

[0051] Optionally, for each candidate permanent residence in the first candidate permanent residence set, the confidence level of being judged as a permanent residence is the quotient of the number of flights to the candidate permanent residence and the total number of flights.

[0052] Optionally, the total number of departures is the sum of the number of departures from all departure cities in the flight travel record, the total number of arrivals is the sum of the number of arrivals from all arrival cities in the flight travel record, and the total number of flights is the sum of the total number of departures and the total number of arrivals.

[0053] In one embodiment, the aggregate functions and clauses in the SQL (Structured Query Language) tool software (such as MySQL, Navicat, SSMS, etc.) are used to collect statistics on each departure city and its corresponding departure times, and each arrival city and its corresponding arrival times in the flight travel record.

[0054] Taking a passenger's 10 flight travel records as an example, Table 1 shows the data for determining the number of flights of the passenger's flight travel records. As shown in Table 1, after statistics by the tool software, the number of departures in City B is 6 and the number of arrivals is 4, the number of departures in City S is 2 and the number of arrivals is 3, the number of departures in City N is 1 and the number of arrivals is 3, and the number of departures in City G is 1. In addition, the total number of departures is 10, the total number of arrivals is 10, and the total number of flights is 20.

[0055] Table 1

[0056] Then, City B, City S, City N and City G are taken as the first candidate permanent residence set, and the number of flights to City B is 10, the number of flights to City S is 5, the number of flights to City N is 4 and the number of flights to City G is 1.

[0057] Furthermore, according to the quotient of the number of flights of each candidate permanent residence in the first candidate permanent residence set and the total number of flights, the confidence of each candidate permanent residence being determined as a permanent residence in the flight number determination is determined, and the confidence of City B being determined as a permanent residence is obtained. is 0.5, the confidence level that City S is considered as the permanent residence is 0.25, the confidence level that city N is considered as the permanent residence is 0.2, the confidence level that City G is considered as the permanent residence is 0.05.

[0058] Finally, cities B, S, N, and G and their corresponding confidence levels that are judged as permanent residences are , , and As the result of the flight number determination.

[0059] The method for determining the permanent residence provided by the present invention analyzes the permanent residence of the passenger as a whole, and according to the passenger's complete flight travel record, all the en route cities (including the departure city and the arrival city) are regarded as possible candidate permanent residences, to ensure that no permanent residence city is omitted, and according to the number of flights of each candidate permanent residence and the total number of flights of the flight travel record, the confidence of each candidate permanent residence being judged as the passenger's permanent residence is determined based on the determination dimension of the number of flights, thereby providing a stable data basis for the subsequent determination of the passenger's permanent residence based on the multi-dimensional determination results of the permanent residence including the determination results of the number of flights.

[0060] Based on the above embodiment, as an optional embodiment, when the multi-dimensional determination result of the permanent residence includes the holiday determination result, the holiday determination result is determined based on the following method: Obtaining the first flight travel record of the passenger in each holiday time period, wherein each holiday time period is determined according to a holiday time period data set pre-retrieved from a third-party database; According to the first flight travel record, determine the departure city of each first flight and its corresponding first flight departure number, and determine the arrival city of each last flight and its corresponding last flight arrival number; Taking all the first flight departure cities and all the last flight arrival cities as candidate permanent residences, a second candidate permanent residence set is constructed; determining a confidence level of each of the candidate permanent residences in the second candidate permanent residence set being judged as a permanent residence as the holiday determination result; The confidence level of each candidate permanent residence being judged as the permanent residence is determined based on the number of holiday flights and the total number of holiday flights of the candidate permanent residence, the number of holiday flights is determined based on the number of first flight departures and the number of last flight arrivals of the candidate permanent residence, the total number of holiday flights is determined based on the total number of first flight departures and the total number of last flight arrivals; The first flight departure city is the departure city in the passenger's first flight record in the first half of the holiday time period, and the last flight arrival city is the arrival city in the passenger's last flight record in the second half of the holiday time period.

[0061] Specifically, when determining holidays, holiday time period information related to the year, month, and day of the passenger's flight travel record is obtained in advance by accessing a special holiday API interface, or using an open source holiday data set, or using a public calendar service, or from a third-party database. Then, flight travel records that travel during each holiday time period are selected from the passenger's flight travel records as the first flight travel record.

[0062] Then, use the tool software to determine the passenger's first flight departure city and its corresponding first flight departure number, and determine the last flight arrival city and its corresponding last flight arrival number from the first flight travel record; the first flight departure city is the departure city in the passenger's first flight record in the first half of the holiday period, and the last flight arrival city is the arrival city in the passenger's last flight record in the second half of the holiday period.

[0063] The total number of first flight departures is determined based on the number of first flight departures from all first flight departure cities in the first flight travel record, and the total number of last flight arrivals is determined based on the number of last flight arrivals from all last flight arrival cities. Then, the total number of holiday flights for the passenger is determined based on the total number of first flight departures and the total number of last flight arrivals.

[0064] Furthermore, all the first flight departure cities and all the last flight arrival cities in the first flight travel record are taken as candidate permanent residences of the passenger to construct a second candidate permanent residence set.

[0065] For each candidate permanent residence in the second candidate permanent residence set, the number of holiday flights of the candidate permanent residence is determined based on the sum of the number of departures of the first flight and the number of arrivals of the last flight, and then the confidence of the candidate permanent residence being judged as the permanent residence is determined based on the number of holiday flights of the candidate permanent residence and the total number of holiday flights. After determining the confidence of all candidate permanent residences in the second candidate permanent residence set being judged as permanent residences, the holiday determination result can be obtained.

[0066] Optionally, for each candidate permanent residence in the second candidate permanent residence set, the confidence level of being judged as a permanent residence is the quotient of the number of holiday flights of the candidate permanent residence and the total number of holiday flights.

[0067] Optionally, the total first flight departures are the sum of the first flight departures of all first flight departure cities in the first flight travel record, the total last flight arrivals are the sum of the last flight arrivals of all last flight arrival cities in the first flight travel record, and the total holiday flights are the sum of the total first flight departures and the total last flight arrivals.

[0068] Optionally, by pre-accessing a dedicated holiday API interface, or using an open source holiday data set, or using a public calendar service, the holiday time period information related to the year, month, and day of the passenger's flight travel record can be obtained.

[0069] Optionally, after obtaining the statutory holiday information from a third-party database, if the statutory holidays are New Year's Day, Qingming Festival and Dragon Boat Festival, the relevant holiday time period is set to New Year's Day, Qingming Festival, Dragon Boat Festival and the 2 days before and after; if the statutory holidays are Spring Festival, May Day and National Day, the relevant holiday time period is set to Spring Festival, May Day, National Day and the 7 days before and after.

[0070] For example, after obtaining statutory holiday information from 2020 to 2022 from a third-party database, the relevant holiday time periods are set as follows: Holidays in 2020: 20191230~20200103 (New Year's Day); 20200117~20200206 (Spring Festival); 20200402~20200408 (Tomb-Sweeping Day); 20200424~20200512 (Labor Day); 20200623~20200629 (Dragon Boat Festival); 20200924~20201015 (Mid-Autumn Festival and National Day); Holidays in 2021: 20201230~20210105 (New Year's Day); 20210204~20210224 (Spring Festival); 20210401~20210407 (Tomb-Sweeping Day); 20210424~20210512 (Labor Day); 20210610~20210616 (Dragon Boat Festival); 20210917~202010923 (Mid-Autumn Festival); 20210924~20211015 (National Day); Holidays in 2022: 20211230~20220105 (New Year's Day); 20220124~20220213 (Spring Festival); 20220401~20220407 (Tomb-Sweeping Day); 20220423~20220511 (May Day); 20220601~20220607 (Dragon Boat Festival); 20220908~20220914 (Mid-Autumn Festival); 20220924~20221015 (National Day).

[0071] As an optional embodiment, whether the passenger is a student is determined based on the passenger's age when traveling on the flight; If the passenger is a student, the holiday time period is determined as the student holiday time period.

[0072] For example, if the passenger is between 18 and 24 years old, the passenger is determined to be a student, and when determining holidays, the holiday time period is determined as the student holiday time period.

[0073] Optionally, after obtaining statutory holiday information from a third-party database, if the statutory holidays are May 1, National Day, New Year's Day, Qingming Festival and Dragon Boat Festival, the relevant student holiday time period is set to May 1, National Day, New Year's Day, Qingming Festival, Dragon Boat Festival and the day before and after.

[0074] Optionally, the student holiday period also includes the winter vacation period (i.e., during the Spring Festival and 20 days before and after) and the summer vacation period (i.e., from June 20 to September 10).

[0075] For example, after obtaining the statutory holidays from 2020 to 2022 from the third-party database, the relevant student holiday time periods are set as follows: Student holidays in 2020: 20191231~20200102 (New Year's Day); 20200104~20200219 (Winter vacation); 20200403~20200407 (Tomb-Sweeping Day); 20200430~20200506 (Labor Day); 20200620~20200910 (Dragon Boat Festival and summer vacation); 20200930~20201009 (Mid-Autumn Festival and National Day); Student holidays in 2021: 20201231~20210104 (New Year's Day); 20210122~20210309 (Winter vacation); 20210402~20210406 ​​(Tomb-Sweeping Day); 20210430~20210506 (May Day); 20210611~20210615 (Dragon Boat Festival); 20210620~20200910 (Summer vacation); 20210918~202010922 (Mid-Autumn Festival); 20210930~20211008 (National Day); Student holidays in 2022: 20211231~20220104 (New Year's Day); 20220111~20220226 (Winter vacation); 20220402~20220406 (Tomb-Sweeping Day); 20220430~20220505 (May Day); 20220602~20220606 (Dragon Boat Festival); 20220620~20220913 (Summer vacation and Mid-Autumn Festival); 20220930~20221008 (National Day).

[0076] In one embodiment, when a passenger has only one flight travel record during a holiday period, if the travel time is in the first half of the holiday period, only the first flight departure city is counted once, that is, the departure city of the flight travel record is used as the first flight departure city of the holiday period; if the travel time is in the second half of the holiday period, only the last flight arrival city is counted once, that is, the arrival city of the flight travel record is used as the last flight arrival city of the holiday period.

[0077] For example, when a passenger has only one flight travel record from City B to City S during the Dragon Boat Festival holiday period in 2022 (20220602~20220606), if the travel time is before 12:00 noon on June 4, 2022, then for this holiday period, City B will be used as the departure city of the first flight only; if the travel time is after 12:00 noon on June 4, 2022, then for this holiday period, City S will be used as the arrival city of the last flight only.

[0078] If the passenger's flight travel records during the Dragon Boat Festival holiday in 2022 include from City B to City S on June 2, 2022, and from City S to City B on June 5, 2022, then for that holiday period, City B will be used as both the departure city of the first flight and the arrival city of the last flight.

[0079] That is to say, during a holiday period, if a passenger has two or more flight travel records, the first flight departure city and the last flight arrival city will be counted once respectively, that is, the departure city of the first flight travel record will be determined as the first flight departure city, and the arrival city of the last flight travel record will be determined as the last flight arrival city.

[0080] If a passenger has only one flight travel record during a holiday period, the first flight departure city or the last flight arrival city will be counted based on whether the flight travel time is in the first half or the second half of the holiday period.

[0081] In order to better illustrate the holiday determination process provided by the present invention, taking all flight travel records of a passenger (non-student) during the holiday period from 2022 to 2022 as an example, the passenger traveled from City B to City S on January 18, 2020, and from City K to City B on February 5, 2020 during the 2020 Spring Festival holiday period; during the 2021 National Day holiday period, from City S to City B on September 28, 2021 (the first half); during the 2022 Dragon Boat Festival holiday period, from City G to City B on June 6, 2022 (the second half).

[0082] The passenger has two flight travel records during the 2020 Spring Festival holiday period. City B will be determined as the first flight departure city and the last flight arrival city during the 2020 Spring Festival holiday period, with the number of times being 1; there is only one flight travel record in the first half of the 2021 National Day holiday period. City S will be determined as the first flight departure city during the 2021 National Day holiday period, with the number of times being 1; there is only one flight travel record in the second half of the 2022 Dragon Boat Festival holiday period. City B will be determined as the last flight arrival city during the 2022 Dragon Boat Festival holiday period, with the number of times being 1.

[0083] The second candidate permanent residence set includes candidate permanent residence cities B and S. The number of holiday flights in city B is 3, the number of holiday flights in city S is 1, and the total number of holiday flights is 4.

[0084] Furthermore, according to the quotient of the number of holiday flights of each candidate permanent residence in the second candidate permanent residence set and the total number of holiday flights, the confidence of each candidate permanent residence being determined as a permanent residence in the holiday determination is determined, and the confidence of City B being determined as a permanent residence is obtained. is 0.75, the confidence level that City S is considered as the permanent residence is 0.25.

[0085] Finally, City B and City S and their corresponding confidence levels that are judged as permanent residences are and As a result of holiday determination.

[0086] The method for determining a permanent residence provided by the present invention performs holiday determination for holiday travel scenarios of passengers. According to the characteristics of passengers traveling on holidays, that is, "departing from the permanent residence and returning to the permanent residence", the passenger's first flight departure city and the last flight arrival city are counted at most once during any holiday time period. Then, all the counted first flight departure cities and last flight arrival cities are taken as candidate permanent residences. The confidence level of each candidate permanent residence being judged as the passenger's permanent residence is determined based on the holiday determination dimension. The method is highly adapted to the holiday travel application scenarios of passengers and improves the accuracy of determining the passenger's permanent residence in the absence of full location information.

[0087] Based on the above embodiment, as an optional embodiment, when the multi-dimensional determination result of the permanent residence includes a business trip determination result, the business trip determination result is determined based on the following method: Obtain the passenger's second flight travel record during non-holiday periods; Determine, from the second flight travel record, a continuous business trip itinerary whose first business trip flight and / or last business trip flight meets the discount rate condition; the continuous business trip itinerary is a continuous trip in which the passenger stays at the travel destination for less than a preset time; According to the continuous business trip itinerary, determine the departure city of each first business trip and its corresponding first business trip departure number, and determine the arrival city of each last business trip and its corresponding last business trip arrival number; Taking all the departure cities of the first business trip and all the arrival cities of the last business trip as candidate permanent residences, a third candidate permanent residence set is constructed; Determining the confidence level of each candidate permanent residence in the third candidate permanent residence set being judged as a permanent residence as the business trip determination result; The confidence level of each candidate permanent residence being judged as the permanent residence is determined based on the number of business travel flights and the total number of business travel flights of the candidate permanent residence. The number of business travel flights is determined based on the number of first business travel departures and the number of last business travel arrivals of the candidate permanent residence. The total number of business travel flights is determined based on the total number of first business travel departures and the total number of last business travel arrivals.

[0088] Specifically, when making a business trip determination, after obtaining the passenger's flight travel records from the database, first exclude the passenger's first flight travel record during the holiday period to obtain the second flight travel record, and then filter out the continuous itineraries in which the passenger stays at the travel destination for less than a preset time from the second flight travel records, that is, the continuous business trip itinerary. In addition, the first business trip flight and / or the last business trip flight in the continuous business trip itinerary used for business trip determination meet the preset discount rate conditions.

[0089] After determining each continuous business trip itinerary from the second flight travel record, for any continuous business trip itinerary, after the flight travel records therein are sorted from earliest to latest according to travel time, the departure city in the first flight travel record is the departure city of the first business trip flight, that is, the first business trip departure city, and the arrival city in the last flight travel record is the arrival city of the last business trip flight, that is, the last business trip arrival city.

[0090] For any continuous business trip, determine the first business trip departure city and the last business trip arrival city. According to all the continuous business trips, determine the first business trip departure city and the last business trip arrival city, and then get the corresponding first business trip departure number and last business trip arrival number.

[0091] Then, all the departure cities for the first business trip and all the arrival cities for the last business trip are taken as candidate permanent residences to construct the third candidate permanent residence set.

[0092] For each candidate permanent residence in the third candidate permanent residence set, the number of business trip flights of the candidate permanent residence is determined based on the sum of the corresponding first business trip departure number and the last business trip arrival number, and then the confidence of the candidate permanent residence being judged as the permanent residence is determined based on the business trip flight number and the total business trip flight number of the candidate permanent residence. After determining the confidence of all candidate permanent residences in the third candidate permanent residence set being judged as permanent residences, the business trip determination result can be obtained.

[0093] Optionally, for each candidate permanent residence in the third candidate permanent residence set, the confidence level of being judged as a permanent residence is the quotient of the number of business travel flights of the candidate permanent residence and the total number of business travel flights.

[0094] Optionally, the total first business trip departures are the sum of the first business trip departures in all first business trip departure cities in all consecutive business trip itineraries, the total last business trip arrivals are the sum of all last business trip arrivals in all consecutive business trip itineraries, and the total business trip flights are the sum of the total first business trip departures and the total last business trip arrivals.

[0095] It is understandable that each continuous business trip includes at least two flight travel records. When the flight travel records in any continuous business trip are sorted from earliest to latest according to travel time, the arrival city in the previous flight travel record is always the same as the arrival city in the next flight travel record.

[0096] Optionally, the preset duration is 10 days, 20 days, 30 days or 45 days, etc.

[0097] Optionally, the discount rate condition refers to the number of times that the discount rate of the first business trip flight and / or the last business trip flight in the continuous business trip itinerary is 80% to 10% is greater than or equal to 1.

[0098] In one embodiment, a continuous business travel itinerary is defined as a number of consecutive itineraries in which the discount rate of the first business travel flight and / or the last business travel flight is 80% to 10% or greater and the passenger stays at the travel destination for less than 30 days.

[0099] Table 2 is a continuous itinerary of a passenger in a non-holiday period provided by this embodiment. As shown in Table 2, it includes 3 continuous itineraries, a total of 7 flight travel records, and the passenger's stay in the travel destination is less than 30 days. Among them, since the discount rate of the two travel flights in continuous itinerary 2 is not greater than 20%, that is, it does not meet the discount rate condition that the number of times the discount rate of the first travel flight and / or the last travel flight is 80% to 10% is greater than or equal to 1, continuous itinerary 2 is not a continuous travel itinerary, and the passenger's continuous travel itinerary only includes continuous itinerary 1 and continuous itinerary 2.

[0100] Table 2

[0101] In continuous trip 1, the departure city of the first business trip is determined to be City B, and the corresponding number of departures for the first business trip is 1. The arrival city of the last business trip is determined to be City B, and the corresponding number of arrivals for the last business trip is 1. In continuous trip 3, the departure city of the first business trip is determined to be City G, and the corresponding number of departures for the first business trip is 1. The arrival city of the last business trip is determined to be City B, and the corresponding number of arrivals for the last business trip is 1.

[0102] Then, the third candidate permanent residence set constructed includes City B and City G, and the number of business travel flights in City B is 3 times, the number of business travel flights in City G is 1 time, and the total number of business travel flights is 4 times.

[0103] Furthermore, according to the quotient of the number of business trips of each candidate permanent residence in the third candidate permanent residence set and the total number of business trips, the confidence of each candidate permanent residence being determined as a permanent residence in the business trip determination is determined, and the confidence of City B being determined as a permanent residence is obtained. The confidence level of G being judged as the permanent residence is 0.75. is 0.25.

[0104] Finally, City B and City G and their corresponding confidence levels that are judged as permanent residences and As a result of the travel determination.

[0105] The method for determining the permanent residence provided by the present invention performs travel determination for application scenarios of continuous business trips for passengers. First, the flight itineraries of passengers during holidays are excluded, and then the continuous itineraries that do not meet the "short-term" characteristics of the business trips and do not meet the discount rate conditions are excluded. The characteristic of "starting from the permanent residence and returning to the permanent residence" of the passengers when traveling for business is utilized, and the first departure city and the last arrival city in the passenger's continuous business trip itinerary are determined as candidate permanent residences. The confidence level of each candidate permanent residence being judged as the passenger's permanent residence is determined in the determination dimension of travel determination. The method is highly adapted to the application scenarios of passengers' business trips and improves the accuracy of determining the passenger's permanent residence in the absence of full location information.

[0106] Based on the above embodiment, as an optional embodiment, when the multi-dimensional determination result of the permanent residence includes a stay duration determination result, the stay duration determination result is determined based on the following method: Determine the effective length of stay of the passenger in each arrival city based on the flight travel record; Taking all the arrival cities as candidate permanent residences, a fourth candidate permanent residence set is constructed; Determining the confidence level of each candidate permanent residence in the fourth candidate permanent residence set being judged as a permanent residence as the stay duration determination result; The confidence level of each candidate permanent residence being judged as the permanent residence is determined based on the effective stay time of the candidate permanent residence and the total effective stay time, where the total effective stay time is the sum of the effective stay times of all the candidate permanent residences.

[0107] Specifically, when determining the length of stay, first retrieve all the flight travel records of the passenger from the database, and sort them from the earliest to the last according to the travel time, and calculate the effective stay time of the passenger in each arrival city starting from the travel time of the first flight travel record, and the effective stay time of the passenger in the arrival city in the first flight travel record is the time interval between the two travel times in the first and second flight travel records, and so on. All the arrival cities in all flight travel records are taken as candidate permanent residences, and the fourth candidate permanent residence set is constructed, and the sum of the effective stay times of all candidate permanent residences in the fourth candidate permanent residence set is taken as the total effective stay time.

[0108] For any candidate permanent residence, the confidence level of the candidate permanent residence being judged as the permanent residence is determined based on the quotient of the effective stay time of the candidate permanent residence and the total effective stay time. After calculating the confidence levels of all candidate permanent residences being judged as permanent residences, all candidate permanent residences and their corresponding confidence levels are used as the stay time determination result.

[0109] Table 3 is the flight travel record of a passenger provided by the present invention from January 1, 2021 to January 1, 2024. As shown in Table 3, the arrival cities S, B, and N are all selected as candidate permanent residences in the fourth candidate permanent residence set.

[0110] Table 3

[0111] The effective duration of the passenger's stay in City S is calculated from the travel time in the first flight travel record, that is, from March 1, 2021. As of June 1, 2021, the effective duration of the passenger's stay in City S is 92 days. After the passenger arrives in City B on June 1, 2021, the travel time of the next flight travel record is June 1, 2022, so the effective duration of the passenger's stay in City B is 365 days.

[0112] Since the passenger arrived in City C on June 1, 2022, there is no record of the passenger's next flight in the database as of January 1, 2024. At this time, based on the period from June 1, 2022 to January 1, 2024, the passenger's effective stay in City N is calculated to be 579 days. The total effective stay is the sum of the passenger's effective stay in the candidate permanent residence cities S, B, and N, which is 1036 days in total.

[0113] Furthermore, according to the quotient of the effective stay time of each candidate permanent residence and the total effective stay time, the confidence level of S city being judged as the permanent residence in the stay time determination dimension is determined with two decimal places retained. The confidence level of City B being judged as the permanent residence is 0.09. is 0.35, and the confidence level that city N is considered as the permanent residence is It is 0.56.

[0114] Finally, the confidence scores of S, B, N and their corresponding permanent residences are calculated. , , As the result of the stay duration determination.

[0115] The method for determining the permanent residence provided by the present invention calculates the effective stay time of each arrival city in the user's flight travel record, and determines the confidence of each arrival city being judged as the permanent residence from the dimension of the stay time, thereby improving the accuracy of determining the passenger's permanent residence in the absence of full location information; and considering the diversity of modern transportation modes, the passenger's travel mode is not limited to civil aviation. Therefore, when the stay time is determined, the effective stay time is calculated only based on the arrival city, avoiding the problem of discontinuity in the calculation of the effective stay time caused by the passenger changing his / her location through other travel modes.

[0116] Based on the above embodiment, as an optional embodiment, when the multi-dimensional determination result of the permanent residence includes an ID card determination result, the ID card determination result is determined based on the following method: Obtain the passenger ID number from the flight travel record; Parse the passenger's ID number based on the ID number coding rule to obtain a unique city corresponding to the passenger's ID number; The unique city is used as a candidate permanent residence to obtain the ID card determination result.

[0117] Specifically, after retrieving the passenger's flight travel record from the database, the passenger's ID number is extracted from the record, and then the ID number is parsed according to the ID number encoding rules to obtain the district string representing the administrative division in the passenger's ID number, and based on the string, the only city corresponding to the district string is matched from the administrative division database, and then the only city is used as a candidate permanent residence, and the confidence that the only city is judged as the permanent residence in the ID card determination dimension is 1.

[0118] Optionally, when the ID card number encoding rules include that the complete ID card number is 18 digits and the first 6 digits represent the administrative division, use the string processing function in programming languages ​​such as Python, Java, C++, C4, JavaScript, etc. to extract the first 6 digits from the ID card number.

[0119] Optionally, the string processing functions provided by the programming language include but are not limited to slice(), substring(), substr() and other string processing functions.

[0120] Taking the ID card number encoding rule including the complete ID card number being 18 digits, and the first 6 digits representing the administrative division as an example, when the first 6 digits are "110000", the corresponding only city is City B. Then, if the first 6 digits of the passenger's ID card number are "110000", then City B is selected as a candidate permanent residence, and the confidence P_(B_5) that City B is determined as the permanent residence in the ID card determination dimension is 1.

[0121] The method for determining permanent residence provided by the present invention parses the passenger's ID number and directly takes the city corresponding to the division character string representing the administrative division in the passenger's ID number as a candidate permanent residence. By combining the ID card determination dimension, the accuracy of determining the passenger's permanent residence in the absence of full location information is improved.

[0122] Based on the above embodiment, as an optional embodiment, determining the passenger's permanent residence based on the multi-dimensional determination result of the permanent residence includes: According to the multi-dimensional determination results of each permanent residence and its corresponding probability weight, the confidence of each candidate permanent residence is weighted and summed to obtain the comprehensive confidence of each candidate permanent residence; the probability weight is determined according to the product of the weight coefficient and the gradient coefficient; Determining the permanent residence of the passenger, the permanent residence being the candidate permanent residence with the greatest comprehensive confidence; The weight coefficient is obtained by training based on the multi-dimensional determination results of the permanent residence of multiple samples and their corresponding permanent residence labels.

[0123] Specifically, each multi-dimensional determination result of the permanent residence includes several candidate permanent residences and their corresponding confidence levels. When performing a multi-dimensional analysis of the permanent residence based on flight travel records, the factors considered in different dimensional analyses are different, and the determination results of different dimensions have different probability weights for the final determination of the passenger's permanent residence. For each candidate permanent residence and its corresponding confidence level determined in the same dimensional determination, they have the same probability weight when calculating the comprehensive confidence level, while the candidate permanent residence and its corresponding confidence level determined in different dimensional determinations have different probability weights when performing weighted summation to calculate the comprehensive confidence level. After calculating the comprehensive confidence level of each candidate permanent residence, the candidate permanent residence with the largest comprehensive confidence level is determined as the passenger's permanent residence.

[0124] Table 4 is a table of determination results obtained after performing multi-dimensional analysis of a passenger's flight travel record based on the permanent residence provided by the present invention. As shown in Table 4, the probability weight of the flight number determination dimension is , the probability weight of the holiday determination dimension is , the probability weight of the travel judgment dimension is , the probability weight of the duration determination dimension is , and the probability weight of the ID card determination dimension is .

[0125] Table 4

[0126] After calculation, the comprehensive confidence of candidate permanent residence city B is , the comprehensive confidence of candidate permanent residence city S is , the comprehensive confidence of candidate permanent residence city N is , the comprehensive confidence of candidate permanent residence city G is Finally, the candidate permanent residence with the largest comprehensive confidence is finally determined as the passenger's permanent residence.

[0127] Among them, the probability weight is determined according to the product of the weight coefficient and the gradient coefficient; the weight coefficient is obtained by training based on the multi-dimensional determination results of the permanent residence of multiple samples and their corresponding permanent residence labels.

[0128] Optionally, the gradient coefficient is determined according to the importance of the judgment results of each dimension.

[0129] Optionally, the weight coefficient of each dimension determination result is determined according to the following steps: First, the actual number of days that passengers stay within the itinerary calculation time range is obtained through surveys and questionnaires, and the real permanent residence city and its corresponding real permanent residence probability are determined based on the actual number of days of stay.

[0130] For example, a passenger stays in City B for 30 days and in City S for 70 days within the itinerary calculation time range. The passenger's actual permanent residence cities include City B and City S. The probability of the actual permanent residence corresponding to City B is 0.3, and the probability of the actual permanent residence corresponding to City S is 0.7.

[0131] Next, the expression of the weight coefficient of the judgment result of each dimension is set to the following formula: ; in, For the The confidence level of the permanent residence determined by the determination results of each dimension, is the number of flights, For the The overall parameters of the dimension determination results, , , , For the The multivariate polynomial coefficients corresponding to the dimension judgment results.

[0132] Then, the weight coefficient corresponding to each dimension judgment result is trained separately, and the correlation between the dimension judgment result and the number of flights is established by separately training the multivariate polynomial coefficients in the weight coefficient expression corresponding to each dimension judgment result.

[0133] Specifically, according to the above weight coefficient The expression of is used to construct a neural network model corresponding to the judgment result of a certain dimension. Each candidate permanent residence city and its corresponding confidence in the judgment result of this dimension are input into the neural network model as input data, and the confidence adjusted by the weight coefficient is output.

[0134] For example, if a passenger stays in City B for 30 days and in City S for 70 days during the travel calculation time range, the probability of the actual permanent residence corresponding to City B is is 0.3, the number of flights is 10.

[0135] If the confidence level of City B in the holiday determination result is 0.2, ×0.2 should be equal to the probability of actual permanent residence . In the number of flights When the value is 10, 0.2 is input into the neural network model, and the output is weighted by the weight coefficient. Adjusted confidence .

[0136] Furthermore, based on the least squares method, The multivariate coefficients in the neural network model constructed by the expression , , , Adjust so that the adjusted confidence The probability of actual permanent residence The error between them is minimal.

[0137] Furthermore, the overall parameters is determined according to the following expression: ; in, For the The overall parameters of the dimension determination results, is the probability of actual permanent residence, is the adjusted confidence level.

[0138] Finally, when the neural network model training is completed, the first The weight coefficient of the result of the dimension judgment .

[0139] Optionally, the neural network model is built based on any one of machine learning algorithms such as linear regression, logistic regression, decision tree, clustering algorithm, etc.

[0140] Optionally, the optimization algorithm of the neural network model is a stochastic gradient descent algorithm.

[0141] The method for determining the permanent residence provided by the present invention analyzes the flight travel records according to multi-threaded logic to determine the permanent residence. However, since the flight travel records do not have full location information, the candidate permanent residence determined from the five dimensions of flight number determination, holiday determination, business trip determination, length of stay determination, and ID card determination and its corresponding confidence result are only a fuzzy detection and require additional weighted summation. Therefore, by using the multi-dimensional determination results of the sample permanent residence and its corresponding permanent residence label, and combining the weight coefficients obtained by big data technology and machine learning algorithm training, the robustness of the method for determining the permanent residence can be improved, and the accuracy of the determination of the passenger's permanent residence can be guaranteed without obtaining the full location information.

[0142] In addition, the method for determining the permanent residence provided by the present invention integrates the results of logical judgments of different dimensions, thereby improving the judgment ability of the model. Using multiple models for comprehensive judgment can reduce the deviation and error of a single model, thereby improving the accuracy of the judgment of the permanent residence; the conditions required for judgment under different logics are different. When a certain logic has problems or fails, other models can still provide effective judgments, thereby improving the robustness of the judgment of this method.

[0143] In addition, multi-model comprehensive judgment can increase the credibility of the judgment, because the judgment results of multiple independent logics can verify and confirm each other; in complex decision-making scenarios, multi-model comprehensive judgment can better deal with uncertainty and complexity and improve the accuracy of permanent residence determination.

[0144] Based on the above embodiment, as an optional embodiment, the gradient coefficients corresponding to the flight number determination result and the holiday determination result are both set to the first gradient coefficient; the gradient coefficients corresponding to the business trip determination result and the stay duration determination result are set to the second gradient coefficient; the gradient coefficient corresponding to the ID card determination result is set to the third gradient coefficient; The first gradient coefficient is greater than the second gradient coefficient, and the second gradient coefficient is greater than the third gradient coefficient.

[0145] In one embodiment, the first gradient coefficient , the second gradient coefficient , the third gradient coefficient ,in, It is the total number of flights determined based on the passenger's flight travel records.

[0146] Accordingly, the probability weight of the flight number determination dimension , the probability weight of the holiday determination dimension is , the probability weight of the travel judgment dimension is , the probability weight of the duration determination dimension is , and the probability weight of the ID card determination dimension is ,in, , , , , It is the weight coefficient of each dimension judgment obtained by training based on the multi-dimensional judgment results of multiple sample permanent residences and their corresponding permanent residence labels.

[0147] The method for determining the permanent residence provided by the present invention takes into account the different considerations for determining each dimension. Specifically, considering that the flight number determination dimension can provide a stable data basis for determining the permanent residence, its gradient coefficient is determined to be the largest gradient coefficient, thereby increasing the probability weight when calculating the comprehensive confidence. It is also considered that the ID card determination is often more suitable for determining the permanent residence of passengers with few or no travel records, so its gradient coefficient is set to the smallest gradient coefficient, thereby reducing the probability weight when calculating the comprehensive confidence, avoiding affecting the determination of the permanent residence of passengers with many travel times.

[0148] Furthermore, the underlying determination of the permanent residence determination method provided by the present invention is composed of different multi-threaded logics. In addition to making the reasons for determining the permanent residence of the passenger explainable, it is also possible to adjust the weight of the summary of each determination result according to actual conditions by setting an adjustable gradient coefficient, so that the model is adapted to actual application scenarios and has strong adaptability.

[0149] Figure 2 This is the second flow chart of the method for determining the permanent residence provided by the present invention, such as Figure 2 As shown, after weighted summing of the confidence of each candidate permanent residence according to the multi-dimensional determination results of each permanent residence and its corresponding probability weight, it also includes: Calculating a total confidence, where the total confidence is the sum of the comprehensive confidences of all candidate permanent residences; For any candidate permanent residence, a new comprehensive confidence of the candidate permanent residence is determined based on the ratio of the comprehensive confidence of the candidate permanent residence to the total confidence, so as to achieve confidence normalization processing.

[0150] The method for determining permanent residence provided by the present invention ensures that the sum of the new comprehensive confidences of each candidate permanent residence is 1 by normalizing each comprehensive confidence, and can intuitively show the possibility of each candidate permanent residence of the passenger being determined as the permanent residence.

[0151] Figure 3 : is a schematic diagram of the structure of the device for determining the permanent residence provided by the present invention, such as Figure 3 As shown, including but not limited to: The multi-dimensional determination unit 301 is used to analyze the passenger's permanent residence from multiple dimensions according to the passenger's flight travel record, and obtain a multi-dimensional determination result of the passenger's permanent residence; A permanent residence determination unit 302, configured to determine the permanent residence of the passenger based on the multi-dimensional determination result of the permanent residence; The multi-dimensional determination result of the permanent residence includes multiple items of the flight number determination result, holiday determination result, business trip determination result, stay duration determination result or ID card determination result.

[0152] It should be noted that the device for determining a permanent residence provided by the present invention can execute the method for determining a permanent residence described in any of the above embodiments during specific operation, which will not be elaborated in this embodiment.

[0153] The device for determining permanent residence provided by the present invention comprehensively considers various factors and business characteristics, and can adapt to different application scenarios by performing at least two of the five dimensional analyses of the passenger's flight travel records, namely, flight number determination, holiday determination, business trip determination, length of stay determination, or ID card determination. The permanent residence of the passenger is finally determined by using the city that may be the passenger's permanent residence and the corresponding confidence results, and the permanent residence of the passenger can be accurately determined without obtaining the passenger's full location information.

[0154] Figure 4 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 4 As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430 and a communication bus 440, wherein the processor 410, the communication interface 420 and the memory 430 communicate with each other through the communication bus 440. The processor 410 may call the logic instructions in the memory 430 to execute the method described in any of the above embodiments, the method comprising: analyzing the passenger's permanent residence from multiple dimensions according to the passenger's flight travel record, and obtaining a multi-dimensional determination result of the passenger's permanent residence; determining the passenger's permanent residence based on the multi-dimensional determination result of the permanent residence; the multi-dimensional determination result of the permanent residence includes multiple items of a flight number determination result, a holiday determination result, a business trip determination result, a stay duration determination result or an ID card determination result.

[0155] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0156] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the method for determining a permanent residence provided in the above-mentioned embodiments, the method comprising: analyzing the permanent residence of the passenger from multiple dimensions according to the flight travel record of the passenger, to obtain a multi-dimensional determination result of the permanent residence of the passenger; determining the permanent residence of the passenger based on the multi-dimensional determination result of the permanent residence; the multi-dimensional determination result of the permanent residence includes multiple items of a flight number determination result, a holiday determination result, a business trip determination result, a stay duration determination result or an ID card determination result.

[0157] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0158] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0159] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for determining a permanent residence, characterized in that: include: According to the flight travel record of the passenger, the passenger's permanent residence is analyzed from multiple dimensions to obtain a multi-dimensional determination result of the passenger's permanent residence; Determine the passenger's permanent residence based on the multi-dimensional determination result of the permanent residence; The multi-dimensional determination result of the permanent residence includes multiple items of the flight number determination result, holiday determination result, business trip determination result, stay duration determination result or ID card determination result.

2. The method for determining a permanent residence according to claim 1, characterized in that: In the case where the multi-dimensional determination result of the permanent residence includes the flight number determination result, the flight number determination result is determined based on the following method: Determine each departure city and its corresponding departure times, and each arrival city and its corresponding arrival times according to the flight travel records; Taking all the departure cities and all the arrival cities as candidate permanent residences, a first candidate permanent residence set is constructed; determining a confidence level of each of the candidate permanent residences in the first candidate permanent residence set being judged as a permanent residence as the flight number determination result; The confidence level of each candidate permanent residence being judged as a permanent residence is determined based on the number of flights and the total number of flights of the candidate permanent residence. The number of flights is determined based on the number of departures and arrivals of the candidate permanent residence. The total number of flights is determined based on the total number of departures and the total number of arrivals.

3. The method for determining a permanent residence according to claim 1, characterized in that: In the case where the multi-dimensional determination result of the permanent residence includes the holiday determination result, the holiday determination result is determined based on the following method: Obtaining the first flight travel record of the passenger in each holiday time period, wherein each holiday time period is determined according to a holiday time period data set pre-retrieved from a third-party database; According to the first flight travel record, determine the departure city of each first flight and its corresponding first flight departure number, and determine the arrival city of each last flight and its corresponding last flight arrival number; Taking all the first flight departure cities and all the last flight arrival cities as candidate permanent residences, a second candidate permanent residence set is constructed; determining a confidence level of each of the candidate permanent residences in the second candidate permanent residence set being judged as a permanent residence as the holiday determination result; The confidence level of each candidate permanent residence being judged as the permanent residence is determined based on the number of holiday flights and the total number of holiday flights of the candidate permanent residence, the number of holiday flights is determined based on the number of first flight departures and the number of last flight arrivals of the candidate permanent residence, the total number of holiday flights is determined based on the total number of first flight departures and the total number of last flight arrivals; The first flight departure city is the departure city in the passenger's first flight record in the first half of the holiday time period, and the last flight arrival city is the arrival city in the passenger's last flight record in the second half of the holiday time period.

4. The method for determining a permanent residence according to claim 3, characterized in that: Determine whether the passenger is a student based on the passenger's age at the time of flight travel; If the passenger is a student, the holiday time period is determined as the student holiday time period.

5. The method for determining a permanent residence according to claim 3, characterized in that: In the case where the multi-dimensional determination result of the permanent residence includes a business trip determination result, the business trip determination result is determined based on the following method: Obtain the passenger's second flight travel record during non-holiday periods; Determine, from the second flight travel record, a continuous business trip itinerary whose first business trip flight and / or last business trip flight meets the discount rate condition; the continuous business trip itinerary is a continuous trip in which the passenger stays at the travel destination for less than a preset time; According to the continuous business trip itinerary, determine the departure city of each first business trip and its corresponding first business trip departure number, and determine the arrival city of each last business trip and its corresponding last business trip arrival number; Taking all the departure cities of the first business trip and all the arrival cities of the last business trip as candidate permanent residences, a third candidate permanent residence set is constructed; Determining the confidence level of each candidate permanent residence in the third candidate permanent residence set being judged as a permanent residence as the business trip determination result; The confidence level of each candidate permanent residence being judged as the permanent residence is determined based on the number of business travel flights and the total number of business travel flights of the candidate permanent residence. The number of business travel flights is determined based on the number of first business travel departures and the number of last business travel arrivals of the candidate permanent residence. The total number of business travel flights is determined based on the total number of first business travel departures and the total number of last business travel arrivals.

6. The method for determining a permanent residence according to claim 1, characterized in that: In the case where the multi-dimensional determination result of the permanent residence includes a determination result of the length of stay, the determination result of the length of stay is determined based on the following method: Determine the effective length of stay of the passenger in each arrival city based on the flight travel record; Taking all the arrival cities as candidate permanent residences, a fourth candidate permanent residence set is constructed; Determining the confidence level of each candidate permanent residence in the fourth candidate permanent residence set being judged as a permanent residence as the stay duration determination result; The confidence level of each candidate permanent residence being judged as the permanent residence is determined based on the effective stay time of the candidate permanent residence and the total effective stay time, where the total effective stay time is the sum of the effective stay times of all the candidate permanent residences.

7. The method for determining a permanent residence according to any one of claims 2 to 6, characterized in that: The determining the passenger's permanent residence based on the multi-dimensional determination result of the permanent residence includes: According to the multi-dimensional determination results of each permanent residence and its corresponding probability weight, the confidence of each candidate permanent residence is weighted and summed to obtain the comprehensive confidence of each candidate permanent residence; the probability weight is determined according to the product of the weight coefficient and the gradient coefficient; Determining the permanent residence of the passenger, the permanent residence being the candidate permanent residence with the greatest comprehensive confidence; The weight coefficient is obtained by training based on the multi-dimensional determination results of the permanent residence of multiple samples and their corresponding permanent residence labels.

8. The method for determining a permanent residence according to claim 7, characterized in that: The gradient coefficients corresponding to the flight number determination result and the holiday determination result are both set to the first gradient coefficient; the gradient coefficients corresponding to the business trip determination result and the stay duration determination result are set to the second gradient coefficient; the gradient coefficient corresponding to the ID card determination result is set to the third gradient coefficient; The first gradient coefficient is greater than the second gradient coefficient, and the second gradient coefficient is greater than the third gradient coefficient.

9. The method for determining a permanent residence according to claim 8, characterized in that: After weighted summing of the confidence of each candidate permanent residence according to the multi-dimensional determination results of each permanent residence and its corresponding probability weight, it also includes: Calculating a total confidence, where the total confidence is the sum of the comprehensive confidences of all candidate permanent residences; For any candidate permanent residence, a new comprehensive confidence of the candidate permanent residence is determined based on the ratio of the comprehensive confidence of the candidate permanent residence to the total confidence, so as to achieve confidence normalization processing.

10. A device for determining a permanent place, characterized in that: include: A multi-dimensional determination unit, configured to analyze the passenger's permanent residence from multiple dimensions according to the passenger's flight travel record, and obtain a multi-dimensional determination result of the passenger's permanent residence; A permanent residence determination unit, configured to determine the permanent residence of the passenger based on the multi-dimensional determination result of the permanent residence; The multi-dimensional determination result of the permanent residence includes multiple items of the flight number determination result, holiday determination result, business trip determination result, stay duration determination result or ID card determination result.

Images