Data correlation clustering analysis-based intentional passenger information determination method and device

Abstract

The invention belongs to the technical field of data processing, and particularly discloses an intentional passenger information determination method and device based on data correlation clustering analysis, and the method comprises the steps: locating a target passenger based on the island ticket information in a target time period, and obtaining the known passenger characteristics of the target passenger in different dimensions, presetting different values of unknown passenger characteristics of the target passenger; the known passenger features and the unknown passenger features are subjected to standard quantization processing and then input to a passenger feature contribution degree model, and car renting duration estimation vectors of target passengers under different values of the unknown passenger features are output; and respectively determining relevancy between the car rental time length estimation vector and the sample standard vectors of different car rental time length categories, and classifying the target passenger to an intention passenger data set corresponding to the car rental time length category with the maximum relevancy. According to the invention, the positioning of the target passenger and the classification of the possible car rental duration are completed, and a data information basis is provided for the follow-up active development of accurate tourism car rental service pushing.

Description

Technical Field

[0001] This application belongs to the technical field of data processing, and more specifically, relates to a method and device for determining intention passenger information based on data correlation clustering analysis. Background Art

[0002] Due to its excellent natural geographical environment, island cities have natural tourism attributes. The opening of the circum-island tourist highway attracts many tourists to drive within the island by themselves, and for this reason, the tourism car rental service has emerged. The current tourism car rental service model mainly relies on users placing orders online automatically or consulting nearby car rental stores after arriving at the port area. Individual online push models are also scatter-shot pushes based on limited data information, mainly relying on traditional market research and simple user portrait construction, lacking precise active push based on big data analysis and prediction, and it is difficult to cope with complex and changeable market demands and passenger behavior patterns. The realization of precise active push for tourism car rental services depends on the accurate positioning of intention passenger information. Summary of the Invention

[0003] In view of the above-mentioned defects existing in the prior art, this application provides a method and device for determining intention passenger information based on data correlation clustering analysis, aiming to solve the problem of intention passenger information positioning.

[0004] In a first aspect, this application provides a method for determining intention passenger information based on data correlation clustering analysis, including: Based on the inbound ticket information within a target time period, locate target passengers and obtain known passenger characteristics of target passengers in different dimensions, and preset different values of unknown passenger characteristics of target passengers; After standard quantization processing of the known passenger characteristics and unknown passenger characteristics, input them into a passenger characteristic contribution model, and output an estimated rental duration vector of the target passenger under different values of the unknown passenger characteristics; Respectively determine the correlation between the estimated rental duration vector and the sample standard vectors of different rental duration categories, and classify the target passengers into the intention passenger data set corresponding to the rental duration category with the maximum correlation; Among them, the passenger characteristic contribution model is used to represent the contribution of various passenger characteristics to the rental duration, and the sample standard vector is a reference vector statistically generated based on the passenger characteristic contribution model for the rental duration sub-data sets corresponding to different rental duration categories within a historical time period.

[0005] In a second aspect, this application further provides a device for determining intention passenger information based on data correlation clustering analysis, including: An acquisition module, configured to locate target passengers based on the inbound ticket information within a target time period and obtain known passenger characteristics of target passengers in different dimensions, and preset different values of unknown passenger characteristics of target passengers; An estimation module, configured to perform standard quantification processing on known passenger characteristics and unknown passenger characteristics and then input them into a passenger characteristic contribution degree model, and output an estimated rental duration vector of the target passenger under different values of the unknown passenger characteristics; A classification module, configured to respectively determine the correlation between the estimated rental duration vector and the sample standard vectors of different rental duration categories, and classify the target passenger into the intention passenger dataset corresponding to the rental duration category with the largest correlation; Wherein, the passenger characteristic contribution degree model is used to represent the contribution degree of various passenger characteristics to the rental duration, and the sample standard vector is a reference vector statistically generated based on the passenger characteristic contribution degree model for the rental duration sub-datasets corresponding to different rental duration categories within a historical time period.

[0006] In a third aspect, the present application further provides an electronic device, including: at least one memory for storing a program; at least one processor for executing the program stored in the memory, and when the program stored in the memory is executed, the processor is configured to execute the method described in the first aspect or any one of the possible implementation manners of the first aspect.

[0007] In a fourth aspect, the present application further provides a computer-readable storage medium, which stores a computer program, and when the computer program runs on a processor, it causes the processor to execute the method described in the first aspect or any one of the possible implementation manners of the first aspect.

[0008] In a fifth aspect, the present application further provides a computer program product, and when the computer program product runs on a processor, it causes the processor to execute the method described in the first aspect or any one of the possible implementation manners of the first aspect.

[0009] The method and device for determining intention passenger information based on data correlation clustering analysis provided by the present application perform statistical analysis on the rental duration sub-datasets corresponding to different rental duration categories within a historical time period based on the passenger characteristic contribution degree model to generate sample standard vectors of different rental duration categories; locate the target passenger within the target time period, obtain the known passenger characteristics of different dimensions that affect the rental duration of the target passenger, and preset the unknown passenger characteristics that affect the rental duration, and jointly use them as the input of the passenger characteristic contribution degree model to obtain an estimated rental duration vector under different values of the unknown passenger characteristics; further calculate the correlation between the estimated rental duration vector and the sample standard vectors of different rental duration categories, and use the correlation to divide the target passenger into the intention passenger dataset corresponding to the possible rental duration category; complete the positioning of the target passenger and the division of the possible rental duration category, providing a data information basis for subsequent actively launching accurate tourist car rental service push. Description of the Drawings

[0010] To more clearly illustrate the technical solutions in this application or related technologies, the following will briefly introduce the accompanying drawings required for use in the embodiments or related technology descriptions. Obviously, the accompanying drawings in the following descriptions are some embodiments of this application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0011] Figure 1 is a schematic flowchart of a method for determining intended passenger information based on data correlation clustering analysis provided by an embodiment of this application; Figure 2 is a histogram of the TraH value fluctuations of different rental duration sub-datasets provided by an embodiment of this application; Figure 3 is a schematic structural diagram of a device for determining intended passenger information based on data correlation clustering analysis provided by an embodiment of this application; Figure 4 is a schematic structural diagram of an electronic device provided by an embodiment of this application. Detailed implementation manners

[0012] In order to make the objectives, technical solutions and advantages of this application more clear and understandable, the following further details this application in combination with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not used to limit this application.

[0013] Figure 1 is a schematic flowchart of a method for determining intended passenger information based on data correlation clustering analysis provided by an embodiment of this application. As Figure 1 shown, the method at least includes the following steps: S101. Locate target passengers based on the inbound ticket information within the target time period, obtain known passenger characteristics of the target passengers in different dimensions, and preset different values of the unknown passenger characteristics of the target passengers; S102. After standard quantization processing of the known passenger characteristics and unknown passenger characteristics, input them into the passenger characteristic contribution degree model, and output the rental duration estimation vectors of the target passengers under different values of the unknown passenger characteristics; S103. Respectively determine the correlation degrees between the rental duration estimation vectors and the sample standard vectors of different rental duration categories, and classify the target passengers into the intended passenger dataset corresponding to the rental duration category with the largest correlation degree.

[0014] The method and device for determining intention passenger information based on data correlation clustering analysis provided by this application statistically analyze the rental duration sub-datasets corresponding to different rental duration categories within a historical time period based on a passenger feature contribution model, and set sample standard vectors for different rental duration categories; locate target passengers within a target time period, obtain known passenger features of different dimensions that affect the rental duration of the target passengers, and preset unknown passenger features that affect the rental duration, and jointly use them as the input of the passenger feature contribution model to obtain rental duration estimation vectors under different values of the unknown passenger features; further calculate the correlation between the rental duration estimation vectors and the sample standard vectors of different rental duration categories, and use the correlation to divide the target passengers into the intention passenger datasets corresponding to the possible rental duration categories, providing a data information basis for subsequent active and accurate tourism car rental services; complete the positioning of the target passengers and the division of the possible rental duration categories, providing a data information basis for subsequent active and accurate tourism car rental service push.

[0015] For S101: Locate target passengers based on the inbound ticket information within the target time period, obtain known passenger features of different dimensions of the target passengers, and preset different values of the unknown passenger features of the target passengers.

[0016] Specifically, the ways of entering and leaving an island city are relatively single, and passengers can only enter and leave the island by ticket. The inbound ticket information includes the unique identifier, age, gender, inbound time, etc. of the passenger. At the same time, tourism trips often have a group nature, such as couples traveling, families traveling, group traveling, etc. Passengers traveling in groups often purchase multiple inbound tickets through a single user terminal. Analyze the data of multiple inbound tickets ordered by the same user terminal as associated tickets. The passengers corresponding to the associated tickets are all fellow travelers. Therefore, the ticket information of fellow travelers can also be associated through the inbound ticket information of a certain passenger. Further, the historical inbound ticket information of the passenger can be retrieved based on the unique identifier of the passenger in the inbound ticket information.

[0017] Using the inbound ticket information as the basis for judging the subsequent car rental tendency of users has stronger data timeliness and scenario relevance compared to judging based on the historical car rental information of users on the car rental platform, and can complete demand prediction more accurately.

[0018] Current passengers who have already entered the island and those who are about to enter the island are all target passengers for the accurate push of tourism car rental services. Locate the target passengers within the target time period through the inbound ticket information within the target time period. The target time period can include a short historical time period and / or a short future time period close to the current moment, such as the previous 3 days and the next 3 days of the current moment, or for another example, the previous week and the next week of the current moment. The target time period can be selected according to seasonal tourism demands.

[0019] From the inbound ticket information, known passenger characteristics of target passengers can be obtained in different dimensions. These known passenger characteristics are associated with the car rental duration and affect the car rental duration. Optionally, the known passenger characteristics include age characteristic TraE, personnel structure characteristic TraS, historical inbound frequency characteristic TraT, etc.

[0020] In some embodiments, obtaining the known passenger characteristics of target passengers in different dimensions in S101 specifically includes: Obtain the age information of the target passenger and the accompanying persons based on the inbound ticket information, determine the age characteristic TraE based on the median age of the target passenger and the accompanying persons, and determine the personnel structure characteristic TraS in combination with the stage in which the ages of the target passenger and the accompanying persons are located.

[0021] Specifically, both the age characteristic and the accompanying personnel structure characteristic reflect the influence of the passenger age on the car rental demand.

[0022] Optionally, use the age median as the age characteristic TraE of all accompanying passengers. Through analysis and comparison, it is found that compared with using the age mean as the age characteristic TraE of all accompanying passengers, the median can better determine the car rental service reservation tendency of all accompanying passengers, and the final prediction result is more accurate.

[0023] Exemplarily, if the age median of all accompanying passengers > 50, then the age characteristic TraE = high; if 30 < the age median of all accompanying passengers ≤ 50, then the age characteristic TraE = medium; if the age median of all accompanying passengers ≤ 30, then the age characteristic TraE = low.

[0024] Optionally, the mode of the stage in which the ages of all accompanying passengers are located can also be used as the age characteristic TraE.

[0025] Furthermore, in order to correct the possible accidental deviation of using the age median or the mode of the age stage as the age characteristic TraE of all accompanying passengers, consider the stage in which the ages of all accompanying passengers are located to determine the personnel structure characteristic TraS. Generally speaking, among all accompanying passengers, the car rental demands of the following types of personnel compositions gradually increase: only middle-aged and young men, only middle-aged and young including women, only the elderly or children, and the elderly and children at the same time.

[0026] Exemplarily, if all accompanying passengers have the elderly and children at the same time, then the personnel structure characteristic TraS = high; if all accompanying passengers only have the elderly or children, then the personnel structure characteristic TraS = relatively high; if all accompanying passengers only have middle-aged and young including women, then the personnel structure characteristic TraS = medium; if all accompanying passengers only have middle-aged and young men, then the personnel structure characteristic TraS = low.

[0027] In some embodiments, the known passenger characteristics of different dimensions obtained in S101 specifically include: Determine the historical island entry frequency characteristic TraT of the target passenger based on the island entry ticket information of the target passenger in the adjacent historical time period.

[0028] Specifically, there is a certain positive correlation between the car rental demand of the target passenger and the historical island entry frequency. The higher the historical island entry frequency, the higher the possibility of developing into a potential car rental customer. The historical island entry frequency characteristic TraT of the target passenger is determined by the island entry ticket information of the target passenger in the adjacent historical time period (such as in the past two years).

[0029] Exemplarily, in the past two years, if the historical island entry frequency of the target passenger > 30 times, then the historical island entry frequency characteristic TraT = many; if 10 times < the historical island entry frequency of the target passenger ≤ 30 times, then the historical island entry frequency characteristic TraT = medium; if the historical island entry frequency of the target passenger ≤ 10 times, then the historical island entry frequency characteristic TraT = few.

[0030] In addition to the known passenger characteristics that can be determined by using the current island entry ticket information and historical island entry ticket information of the target passenger, there are unknown passenger characteristics that affect the car rental duration, such as the travel destination characteristic TraP. Due to reasons such as many passengers not necessarily booking all hotels in advance, it is difficult to obtain such characteristics through relevant data. Therefore, the extraction of unknown passenger characteristics is completed by presetting different possible values.

[0031] For S102: After standard quantization processing of the known passenger characteristics and unknown passenger characteristics, input them into the passenger characteristic contribution degree model, and output the car rental duration estimation vector of the target passenger under different values of the unknown passenger characteristics.

[0032] Specifically, the passenger characteristic contribution degree model is used to characterize the contribution degree of various passenger characteristics to the car rental duration. In some embodiments, the passenger characteristic contribution degree model is specifically pre-constructed through the following steps: Construct a car rental duration data set of island entry passengers in the historical time period. Each sample in the car rental duration data set includes the car rental duration and various passenger characteristics; Determine the information gain of various passenger characteristics to the car rental duration data set, and construct a passenger characteristic contribution degree model based on the determined information gain.

[0033] Specifically, the car rental duration dataset is obtained by statistically analyzing the relevant data of inbound passengers within a historical time period. Each sample includes the car rental duration and various passenger characteristics. The car rental duration includes multiple different categories, and various passenger characteristics such as the aforementioned age characteristic TraE, personnel structure characteristic TraS, inbound frequency characteristic TraT, and tourist destination characteristic TraP, etc. Among them, the tourist destination characteristic TraP can be obtained through information such as the hotel orders and scenic spot tickets of passengers.

[0034] Exemplarily, the car rental duration TraCT is divided into four categories: long, medium, short, and none. When the car rental duration > 10 days, TraCT = long; when 10 days ≥ car rental duration > 3 days, TraCT = medium; when 3 days ≥ car rental duration ≥ 1 day, TraCT = short; when the car rental duration = 0 days, TraCT = none. An example of the car rental duration dataset is shown in Table 1 specifically.

[0035] Table 1

[0036] Information entropy is an indicator to measure the uncertainty of a random variable. In the embodiments of the present application, the information gain of various passenger characteristics for the composition duration dataset is calculated by means of information entropy, and further the contribution ratio of various passenger characteristics to the car rental duration is calculated to construct a passenger characteristic contribution model. The specific process is as follows: Step a: Determine the empirical entropy of the car rental duration dataset.

[0037] The empirical entropy is used to measure the uncertainty or chaos degree of a random variable. The probability distribution of the discrete random variable "car rental duration" Satisfies:

[0038] Among them, Is the probability of the i-th type of car rental duration, Represents the number of samples of the i-th type of car rental duration in the car rental duration dataset, Represents the total number of samples, and m represents the number of car rental duration categories. The empirical entropy of the car rental duration is determined using the probability distribution Satisfies:

[0039] In the car rental duration dataset exemplified in Table 1, the car rental duration includes four categories: long, medium, short, and none, that is, the value of m is 4.

[0040] Step b: Determine the empirical conditional entropy of various passenger characteristics for the car rental duration.

[0041] The empirical conditional entropy refers to the empirical conditional entropy under a certain condition constraint. The empirical conditional entropy specifically satisfies:

[0042] Among them, represents the empirical conditional entropy of the rental duration with respect to the j-th type of passenger feature, represents the sample size of the i-th type of rental duration in the rental duration dataset, represents the total number of samples, represents the sample size of the k-th value in the j-th type of passenger feature, and K represents the number of values of the j-th type of passenger feature. Taking Table 1 as an example, for the age feature TraE, it includes three categories: high, medium, and low, that is, K = 3; for the personnel structure feature TraS, it includes four categories: high, relatively high, medium, and low, that is, K = 4.

[0043] Step c: Determine the information gain of each type of passenger feature with respect to the rental duration dataset.

[0044] The information gain of a certain type of passenger feature with respect to the rental duration represents the degree to which the uncertainty of the rental duration category is reduced given the value of this type of passenger feature. Using the obtained empirical entropy of the rental duration and the empirical conditional entropy of each type of passenger feature with respect to the rental duration, calculate the information gain of each type of passenger feature with respect to the rental duration, specifically satisfying:

[0045] Among them, represents the information gain of the j-th type of passenger feature with respect to the rental duration.

[0046] After obtaining the information gain of each type of passenger feature with respect to the rental duration, construct a passenger feature contribution model. The information gain of each type of passenger feature with respect to the rental duration will ultimately be reflected in the weights of the quantified values of each type of passenger feature in the passenger feature contribution model. Therefore, it is necessary to normalize the information gain of each type of passenger feature with respect to the rental duration to determine the contribution degree ratio of each type of passenger feature with respect to the rental duration.

[0047] After normalization, construct the passenger feature contribution model as follows:

[0048] Among them, represents the quantified analysis result for characterizing the rental duration, , , and respectively represent the quantified values of the historical island entry frequency feature TraT, age feature TraE, personnel structure feature TraS, and tourist destination feature TraP, , , and respectively represent the weights of the quantified values of each type of passenger feature.

[0049] The passenger feature contribution model provided by the embodiments of the present application accurately quantifies the influence degree of various passenger features in historical data on the car rental duration through statistical analysis of the historical data, providing a basis for determining the information of potential passengers in the subsequent stage. After standard quantization processing of the foregoing known passenger features and unknown passenger features, they are input into the passenger feature contribution model, and an n-dimensional car rental duration estimation vector of the target passenger under different value ranges of the unknown passenger features is output. where n is the number of value ranges of the unknown passenger features.

[0050] Optionally, performing standard quantization processing on the known passenger features and unknown passenger features specifically means performing value assignment processing on various passenger features within the same value assignment range.

[0051] For S103: Determine the correlation degrees between the car rental duration estimation vector and the sample standard vectors of different car rental duration categories respectively, and classify the target passenger into the intention passenger data set corresponding to the car rental duration category with the largest correlation degree.

[0052] Specifically, the sample standard vector is a reference vector statistically generated based on the passenger feature contribution model for the car rental duration sub-data sets corresponding to different car rental duration categories within a historical time period.

[0053] In some embodiments, the sample standard vectors of different car rental duration categories are obtained based on the following steps: Construct a car rental duration data set of inbound passengers within a historical time period, and each sample in the car rental duration data set includes the car rental duration and various passenger features; Divide into multiple car rental duration sub-data sets according to the car rental duration category; After performing standard quantization processing on various passenger features of the samples in the car rental duration sub-data set, input them into the passenger feature contribution model, and output the quantitative analysis results of the samples; Perform interval statistics on the quantitative analysis results of the samples in the car rental duration sub-data, screen multiple intervals with the largest amount of interval data, and construct the sample standard vectors of the corresponding car rental duration categories with the average values of the quantitative analysis results within the selected intervals.

[0054] Specifically, after constructing the passenger feature contribution model using the car rental duration data set, further analyze and process the car rental duration data set. Divide the car rental duration data set into multiple car rental duration sub-data sets according to the car rental duration category. For each car rental duration sub-data set, perform standard quantization processing on various passenger features of the samples and then input them into the constructed passenger feature contribution model to obtain the quantitative analysis results of each sample.

[0055] Exemplarily, taking the rental duration sub-dataset with a long rental duration as an example, for any sample, if the age feature TraE = high, it is assigned a value of 3; if the age feature TraE = medium, it is assigned a value of 2; if the age feature TraE = low, it is assigned a value of 1.

[0056] If the frequency of entering the island feature TraT = many, it is assigned a value of 3; if the frequency of entering the island feature TraT = medium, it is assigned a value of 2; if the frequency of entering the island feature TraT = few, it is assigned a value of 1.

[0057] If the personnel structure feature TraS = high, it is assigned a value of 3; if the personnel structure feature TraS = relatively high, it is assigned a value of 2.25; if the personnel structure feature TraS = medium, it is assigned a value of 1.5; if the personnel structure feature TraS = low, it is assigned a value of 0.75. Alternatively, for the elderly (over 55 years old) and children (under 12 years old) among the fellow travelers, they are respectively assigned a value of 0.5, for middle-aged and young women, it is assigned a value of 0.3, and for middle-aged and young men, it is assigned a value of 0.2. The sum of the assignment results of all fellow travelers is used as the standard quantization value of the personnel structure feature.

[0058] For the tourism destination feature TraP, there are only two categories, namely multiple and single, which are respectively assigned values of 3 and 1.5. Alternatively, according to the statistical value of the actual number of tourism destinations, it is assigned a value within the range of 0 to 3.

[0059] The above are all examples of the standard quantization processing of various passenger characteristics. In the embodiments of the present application, the standard quantization processing methods for various passenger characteristics are not limited. After the standard quantization processing of various passenger characteristics, the standard quantization values of each sample are obtained { , , , }. The standard quantization values of each sample { , , , } are input into the passenger characteristic contribution degree model, and the quantitative analysis result TraH of the sample rental duration estimation is output.

[0060] Exemplarily, all samples in each rental duration sub-dataset are subjected to the aforementioned quantitative analysis, and the following quantitative analysis results are obtained:

[0061] Among them, a, b, c, and d respectively represent the number of samples in four types of rental duration sub-datasets.

[0062] Furthermore, an interval statistics is performed on the quantitative analysis results of the samples in the rental duration sub-data, multiple intervals with the largest amount of interval data are selected, and the average value of the quantitative analysis results within the selected intervals is used to construct the sample standard vector corresponding to the rental duration category.

[0063] In some embodiments, interval statistics are performed on the quantization analysis results of the samples in the rental duration sub-dataset in the form of a histogram. Figure 2 It is a histogram of the TraH value fluctuations of different rental duration sub-datasets provided by the embodiments of the present application. As Figure 2 shown, the abscissa is the interval of the TraH value, the ordinate is the data volume, and different colors correspond to different rental duration categories. The fluctuation histogram of each rental duration sub-dataset is drawn according to the data volume of the points falling within the unit interval.

[0064] For the TraH values of the quantization analysis results of the samples in each rental duration sub-dataset, multiple intervals with the largest data volume in the interval are selected, that is, multiple intervals with the highest ordinate values, and the average value of the TraH values of the quantization analysis results within the selected intervals is used to construct the sample standard vector corresponding to the rental duration category. The number of selected intervals is n, which is the same as the preset number of values of the unknown passenger characteristics of the target passenger, so as to ensure that the rental duration estimation vector and the sample standard vector have the same dimension.

[0065] Exemplarily, for the rental duration sub-dataset with the rental duration category of, after constructing a fluctuation histogram for the quantization analysis results of a samples in the sub-dataset, n intervals with the most concentrated data volume are selected. For each of the 4 selected intervals, the average value of the TraH value is obtained as the sample standard value of the interval, and the sample standard vector with the rental duration category of is constructed with the sample standard values of the 4 intervals. The sample standard vectors for 4 rental durations are finally obtained as follows:

[0066] The sample standard vector represents the maximum possible quantization value of each rental duration and serves as the basis for subsequent clustering processing.

[0067] In the process of determining the intention passenger information, a passenger characteristic contribution model and sample standard vectors corresponding to various rental durations are constructed through statistical analysis of historical data; in the process of analyzing and processing the real-time information of passengers, the known passenger characteristics and unknown passenger characteristics of the target passenger are subjected to standard quantization processing and then input into the passenger characteristic contribution model to obtain the rental duration estimation vector of the target passenger under different values of the unknown passenger characteristics, and the correlation between the rental duration estimation vector and the sample standard vectors of various rental durations is calculated to complete the classification of the target passenger.

[0068] The correlation between the rental duration estimation vector and the sample standard vectors of various rental durations can be measured by the Pearson correlation coefficient, Spearman correlation coefficient, Kendall correlation coefficient, etc.

[0069] In some embodiments, the correlation between the estimated rental duration vector and the sample standard vectors of various rental durations is represented by the Pearson correlation coefficient. The Pearson correlation coefficient between the estimated rental duration vector and the sample standard vectors of various rental durations satisfies:

[0070] Where, represents the Pearson correlation coefficient between the estimated rental duration vector and the sample standard vector corresponding to the rental duration category x; represents the i-th element in the estimated rental duration vector, which is the quantification result of the rental duration of the target passenger under the i-th unknown passenger feature value; represents the mean value of each element in the estimated rental duration vector, and satisfies: ; represents the i-th element in the sample standard vector of the rental duration category x; represents the mean value of each element in the sample standard vector of the rental duration category x, and satisfies: . The rental duration categories x can be L, M, S, and N respectively. can be the aforementioned L, M, S, and N respectively, corresponding to long, medium, short, and no rental durations.

[0071] After calculating the Pearson correlation coefficients between the estimated rental duration vector of the target passenger and the sample standard vectors of various rental durations , the Pearson correlation coefficients are arranged in ascending / descending order, and the target passenger is classified into the intention passenger dataset corresponding to the rental duration category with the largest Pearson correlation coefficient.

[0072] Optionally, if other correlation degrees except the maximum correlation degree meet the preset conditions, the target passenger is classified into the intention passenger dataset corresponding to the rental duration category corresponding to the correlation degree that meets the preset conditions. For example, according to the degree of closeness between each Pearson correlation coefficient and 1, a certain threshold is preset. Even if the Pearson correlation coefficient obtained under a certain rental duration category is not the maximum value, but this Pearson correlation coefficient is greater than the preset threshold, the target passenger can also be clustered into the intention passenger dataset corresponding to the corresponding rental duration category.

[0073] After classifying all target passengers within the target time period, the dataset of potential passengers with long rental duration TraLT{TraV1, TraV2,...}, the dataset of potential passengers with medium rental duration TraMT{TraV1, TraV2,...}, the dataset of potential passengers with short rental duration TraST{TraV1, TraV2,...}, and the dataset of potential passengers with no rental duration TraNT{TraV1, TraV2,...} are obtained, where TraV can be used to represent the identifier of the target passenger. Finally, the dataset of potential passengers TMsg{TraLT, TraMT, TraST} within the target time period is output as the basis for accurate and proactive push of car rental services.

[0074] The following describes the apparatus for determining potential passenger information based on data correlation clustering analysis provided in this application. The apparatus for determining potential passenger information based on data correlation clustering analysis described below can be correspondingly referred to with the method for determining potential passenger information based on data correlation clustering analysis described above.

[0075] Figure 3 It is a schematic structural diagram of the apparatus for determining potential passenger information based on data correlation clustering analysis provided in an embodiment of this application. As Figure 3 shown, the apparatus at least includes: An acquisition module 301, configured to locate target passengers based on the inbound ticket information within the target time period, obtain known passenger characteristics of the target passengers in different dimensions, and preset different values of the unknown passenger characteristics of the target passengers; An estimation module 302, configured to perform standard quantization processing on the known passenger characteristics and the unknown passenger characteristics and then input them into a passenger characteristic contribution degree model, and output an estimated rental duration vector of the target passengers under different values of the unknown passenger characteristics; A classification module 303, configured to respectively determine the Pearson correlation coefficients between the estimated rental duration vector and the sample standard vectors of different rental duration categories, and classify the target passengers into the dataset of potential passengers corresponding to the rental duration category with the largest Pearson correlation coefficient; Among them, the passenger characteristic contribution degree model is used to represent the contribution degree of various passenger characteristics to the rental duration, and the sample standard vector is a reference vector statistically generated based on the passenger characteristic contribution degree model for the rental duration sub-datasets corresponding to different rental duration categories within the historical time period.

[0076] It can be understood that the detailed function implementation of the above-mentioned each unit / module can refer to the introduction in the foregoing method embodiments, and will not be elaborated here.

[0077] It should be understood that the above device is used to execute the method in the above embodiment. For the corresponding program modules in the device, their implementation principles and technical effects are similar to those described in the above method. The working process of the device can refer to the corresponding process in the above method, which will not be elaborated here.

[0078] Based on the method in the above embodiment, an embodiment of the present application provides an electronic device. The device may include: at least one memory for storing programs and at least one processor for executing the programs stored in the memory. Among them, when the program stored in the memory is executed, the processor is used to execute the method described in the above embodiment.

[0079] Figure 4 is a schematic structural diagram of the electronic device provided by the embodiment of the present application. As Figure 4 shown, the electronic device may include: a processor (Processor) 401, a communication interface (Communications Interface) 402, a memory (Memory) 403, and a communication bus 404. Among them, the processor 401, the communication interface 402, and the memory 403 complete mutual communication through the communication bus 404. The processor 401 may call software instructions in the memory 403 to execute the method described in the above embodiment.

[0080] In addition, when the logical instructions in the above memory 403 are implemented in the form of software functional units and sold or used as an independent product, they may be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, may be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in the various embodiments of the present application.

[0081] Based on the method in the above embodiment, an embodiment of the present application provides a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program runs on the processor, the processor is caused to execute the method in the above embodiment.

[0082] Based on the method in the above embodiment, an embodiment of the present application provides a computer program product. When the computer program product runs on the processor, the processor is caused to execute the method in the above embodiment.

[0083] It can be understood that the processor in the embodiments of the present application may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0084] The method steps in the embodiments of the present application may be implemented in a hardware manner or by a processor executing software instructions. The software instructions may be composed of corresponding software modules, and the software modules may be stored in a random access memory (RAM), flash memory, read-only memory (ROM), programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), registers, hard disks, removable hard disks, CD-ROMs, or any other form of storage medium well known in the art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from the storage medium and write information to the storage medium. Of course, the storage medium may also be a component of the processor. The processor and the storage medium may be located in an ASIC.

[0085] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk (SSD)), etc.

[0086] It can be understood that the various numerical numbers involved in the embodiments of the present application are only for the convenience of description and are not used to limit the scope of the embodiments of the present application.

[0087] Those skilled in the art can easily understand that the above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for determining information of intended passengers based on data correlation clustering analysis, characterized in that, Including: Locate target passengers based on the ticket information of entering the island within the target time period, obtain the known passenger characteristics of the target passengers in different dimensions, and preset different values of the unknown passenger characteristics of the target passengers; After performing standard quantization processing on the known passenger characteristics and the unknown passenger characteristics, input them into the passenger characteristic contribution degree model, and output the rental duration estimation vector of the target passenger under different values of the unknown passenger characteristics; Respectively determine the correlation degrees between the rental duration estimation vector and the sample standard vectors of different rental duration categories, and classify the target passenger into the intention passenger dataset corresponding to the rental duration category with the maximum correlation degree; Among them, the passenger characteristic contribution degree model is used to characterize the contribution degrees of various passenger characteristics to the rental duration, and the sample standard vector is a reference vector statistically generated based on the passenger characteristic contribution degree model for the rental duration sub-datasets corresponding to different rental duration categories within the historical time period.

2. The method for determining the information of intended passengers according to claim 1, wherein The sample standard vectors of different rental duration categories are obtained based on the following steps: Construct a rental duration dataset of passengers entering the island within the historical time period, where each sample in the rental duration dataset includes the rental duration and various passenger characteristics; Divide into multiple rental duration sub-datasets according to the rental duration category; After performing standard quantization processing on various passenger characteristics of the samples in the rental duration sub-dataset, input them into the passenger characteristic contribution degree model, and output the quantitative analysis results of the samples; Perform interval statistics on the quantitative analysis results of the samples in the rental duration sub-dataset, screen multiple intervals with the largest amount of interval data, and construct the sample standard vector corresponding to the rental duration category with the average value of the quantitative analysis results within the selected intervals.

3. The method for determining the information of intended passengers according to claim 1 or 2, characterized in that, The passenger characteristic contribution degree model is pre-constructed based on the following steps: Construct a rental duration dataset of passengers entering the island within the historical time period, where each sample in the rental duration dataset includes the rental duration and various passenger characteristics; Determine the information gain of various passenger characteristics to the rental duration dataset, and construct the passenger characteristic contribution degree model based on the determined information gain.

4. The method for determining intention passenger information according to claim 1, characterized in that, The method further includes: If other correlation degrees except the maximum correlation degree meet the preset conditions, classify the target passenger into the intention passenger dataset corresponding to the rental duration category corresponding to the correlation degree that meets the preset conditions.

5. The method for determining intended passenger information according to claim 1, characterized in that The known passenger characteristics in different dimensions include age characteristics, personnel structure characteristics, and historical island entry frequency characteristics, and the unknown passenger characteristics include tourism destination characteristics.

6. The method for determining intended passenger information according to claim 5, wherein The obtaining of the known passenger characteristics of the target passenger in different dimensions includes: Obtain the age information of the target passenger and the accompanying personnel based on the ticket information of entering the island, determine the age characteristics based on the median of the ages of the target passenger and the accompanying personnel, and determine the personnel structure characteristics in combination with the stages where the ages of the target passenger and the accompanying personnel are located; Determine the historical island entry frequency characteristics of the target passenger based on the ticket information of the target passenger entering the island in the adjacent historical time period.

7. The method for determining intended passenger information according to claim 1, wherein The correlation degrees between the rental duration estimation vector and the sample standard vectors of different rental duration categories are represented by Pearson correlation coefficients.

8. The method for determining the information of intended passengers according to claim 7, characterized in that, The Pearson correlation coefficient between the estimated rental duration vector and the sample standard vectors of different rental duration categories satisfies: Among them, represents the Pearson correlation coefficient between the rental duration estimation vector and the sample standard vector corresponding to the rental duration category x; represents the i-th element in the rental duration estimation vector, corresponding to the rental duration quantization result of the target passenger under the i-th unknown passenger feature value; represents the mean value of each element in the rental duration estimation vector; represents the i-th element in the sample standard vector of the rental duration category x, represents the mean value of each element in the sample standard vector of the rental duration category x.

9. An apparatus for determining information of intended passengers based on data correlation clustering analysis, characterized in that, including: an acquisition module, configured to locate target passengers based on the inbound ticket information within a target time period, acquire known passenger characteristics of the target passengers in different dimensions, and preset different values of unknown passenger characteristics of the target passengers; an estimation module, configured to perform standard quantization processing on the known passenger characteristics and the unknown passenger characteristics and then input them into a passenger characteristic contribution degree model, and output an estimated rental duration vector of the target passengers under different values of the unknown passenger characteristics; a classification module, configured to respectively determine the correlation degrees between the estimated rental duration vector and the sample standard vectors of different rental duration categories, and classify the target passengers into the intention passenger data set corresponding to the rental duration category with the maximum correlation degree; wherein, the passenger characteristic contribution degree model is used to represent the contribution degrees of various passenger characteristics to the rental duration, and the sample standard vector is a reference vector statistically generated based on the passenger characteristic contribution degree model for rental duration sub-data sets corresponding to different rental duration categories within a historical time period.

10. An electronic device / image signal generator / network device / transmitter / terminal / base station / industrial control computer, characterized in that, including: at least one memory, configured to store a computer program; at least one processor, configured to execute the program stored in the memory, and when the program stored in the memory is executed, the processor is configured to execute the method according to any one of claims 1-8.

Images