GB2573351A - Machine learning system and medium for calculating passenger values of airline - Google Patents
Machine learning system and medium for calculating passenger values of airline Download PDFInfo
- Publication number
- GB2573351A GB2573351A GB1815344.5A GB201815344A GB2573351A GB 2573351 A GB2573351 A GB 2573351A GB 201815344 A GB201815344 A GB 201815344A GB 2573351 A GB2573351 A GB 2573351A
- Authority
- GB
- United Kingdom
- Prior art keywords
- passenger
- values
- value
- feature
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
- 238000010801 machine learning Methods 0.000 title claims abstract description 11
- 238000012549 training Methods 0.000 claims abstract description 54
- 238000012360 testing method Methods 0.000 claims abstract description 47
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 30
- 238000002790 cross-validation Methods 0.000 claims description 17
- 238000000034 method Methods 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 10
- 230000001105 regulatory effect Effects 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 238000009825 accumulation Methods 0.000 description 3
- 230000004075 alteration Effects 0.000 description 2
- 238000009412 basement excavation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
- G06N5/025—Extracting rules from data
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Finance (AREA)
- Entrepreneurship & Innovation (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Computing Systems (AREA)
- Tourism & Hospitality (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Educational Administration (AREA)
- Primary Health Care (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Medical Informatics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A machine learning system for calculating customer values is disclosed. A feature database is preset with feature algorithms in one-to-one correspondence with feature parameters forming a parameter set 101. Historical customer information is combined with the feature database to generate two data sets: a training set and a test set 102. The training set is input into the XGBoost algorithm engine to generate a reference model 103, which is cross validated with the test set to generate a value assessment model 104. Customer information is input into the value assessment model to generate customer value scores 105. The customers may be airline passengers, and the customer values may comprise passenger value scores. The system may employ a passenger value table comprising passenger value sections and estimated passenger values, enabling passenger value scores to be associated with estimated passenger values. Potential passenger values may be judged against a potential value threshold to determine if a passenger can be classified into a high-end passenger database. The feature parameters may comprise information relating to a customer’s travels, bookings, bad experiences, social influence, social age, interests, or membership level. A model hyper-parameter database and error index may be used to iteratively train the model.
Description
MACHINE LEARNING SYSTEM AND MEDIUM FOR CALCULATING PASSENGER VALUES OF AIRLINE
Technical Field
The present invention relates to the technical field of information, and particularly relates to a system and medium for calculating passenger values of an airline.
Background
With the increasing competition among airlines, analysis of passenger values becomes the most concerned issue among the airlines. At present, for the calculating modes of passenger values, the airlines often conduct simple quantitative analysis on the history of passengers in the airlines so as to obtain value scores of the passengers. For example, the history of the passenger is used to calculate the total historical consumption amount or historical flight miles of the passenger in the airline, and then the total consumption amount and the historical flight miles are converted into value scores. This value calculation system allows the airlines to roughly screen out high-end consumer groups. However, this value calculation system is simple in structure, which leads to the inaccuracy of the passenger value score generated finally and cannot meet the needs of the existing airlines for the analysis of the passenger value.
Summary
The present invention aims to solve one of technical problems in the above background at least to a certain extent. To this end, the first purpose of the present invention is to propose a system for calculating passenger values of an airline. The multidimensional information of the passengers of the airline is collected, and then the passenger value scores are generated in a machine learning mode. Accurate passenger value analysis data can be provided for marketing decisions and cost decisions of the airline.
The second purpose of the present invention is to propose a non-temporary computer readable storage medium.
To achieve the above purposes, the embodiment of the first aspect of the present invention proposes a method for calculating passenger values of an airline. The system comprises:
i presetting a feature database, wherein the feature database comprises a feature parameter set and feature algorithms in one-to-one correspondence with feature parameters in the feature parameter set; acquiring historical information of a plurality of passengers, and generating training data according to the feature database and the historical information of the plurality of passengers, wherein the training data comprises a training set and a test set; inputting the training set into an Xgboost algorithm engine to generate a reference model; conducting cross validation on the reference model according to the test set to generate a passenger value assessment model; andinputting the passenger information into the passenger value assessment model to generate passenger value scores, and associating the passenger value scores with the passenger information.
According to the system for calculating passenger values of the airline in the embodiments of the present invention, the feature database is preset, wherein the feature database comprises a feature parameter set and feature algorithms in one-to-one correspondence with feature parameters in the feature parameter set; then the historical information of the plurality of passengers is acquired, and the training data is generated according to the feature database and the historical information of the plurality of passengers, wherein the training data comprises a training set and a test set; next the training set is inputted into an Xgboost algorithm engine to generate the reference model; then cross validation is conducted on the reference model according to the test set to generate the passenger value assessment model; and finally, the passenger information is inputted into the passenger value assessment model to generate the passenger value scores, and the passenger value scores are associated with the passenger information. Thus, the multidimensional information of the passengers of the airline is collected, and then the passenger value scores are generated in a machine learning mode. Accurate passenger value analysis data can be provided for marketing decisions and cost decisions of the airline, in order to provide better services for the passengers.
In addition, the system for calculating passenger values of the airline according to the above embodiment of the present invention can also have the following appended technical features:
Optionally, the system also comprises: presetting a passenger value table, wherein the passenger value table comprises a plurality of passenger value sections and estimated passenger values in one-to-one correspondence with the passenger value sections; and acquiring the estimated passenger values corresponding to the passenger value sections to which the passenger value scores belong according to the passenger value scores, and associating the estimated passenger values with the passenger value scores.
Optionally, the system also comprises: presetting a potential value threshold and a high-end passenger database; acquiring actual passenger values, generating potential passenger values according to the actual passenger values and the estimated passenger values, and associating the potential passenger values with the estimated passenger values; judging whether the potential passenger values reach the potential value threshold; and if so, storing the passenger information associated with the potential passenger values into the high-end passenger database.
Optionally, the step of conducting cross validation on the reference model according to the test set to generate a passenger value assessment model comprises: presetting model hyper-parameters and a model error index; inputting a test sample of the test set into the reference model to generate an estimated result; judging whether an error between a test result of the test set and an estimated result is smaller than the model error index; if so, using the reference model as the passenger value assessment model; and if not, regulating the model hyper-parameters and conducting iterative training until the error between the generated estimated result and the test result is smaller than the model error index.
Optionally, the feature parameters comprise: one or more of a variation tendency parameter, a bad experience parameter, a travel parameter, a booking parameter, an integral parameter, an attribute parameter and a state parameter.
Optionally, the attribute parameter comprises number of joint pedestrians, social influence information and interest-related information.
Optionally, the state parameter comprises a passenger social age, a passenger member level and member social information.
To achieve the above purposes, the embodiment of the second aspect of the present invention provides a non-temporary readable storage medium, storing computer programs. The programs, when executed by a processor, realize the method for calculating passenger values of the airline in the embodiment of the first aspect.
To achieve the above purposes, the embodiment of the third aspect of the present invention provides a system for calculating passenger values of an airline. The system comprises: a presetting module used for presetting a feature database, wherein the feature database comprises a feature parameter set and feature algorithms in one-to-one correspondence with feature parameters in the feature parameter set; an acquisition module used for acquiring historical information of a plurality of passengers; a first generation module used for generating training data according to the feature database and the historical information of the plurality of passengers, wherein the training data comprises a training set and a test set; a second generation module used for inputting the training set into an Xgboost algorithm engine to generate a reference model; a third generation module used for conducting cross validation on the reference model according to the test set to generate a passenger value assessment model; anda value calculation module used for inputting the passenger information into the passenger value assessment model to generate passenger value scores, and associating the passenger value scores with the passenger information.
According to the system for calculating passenger information of the airline in the embodiments of the present invention, the historical information of the plurality of passengers is acquired through the acquisition module; the feature database is preset through the presetting module according to the feature algorithms in one-to-one correspondence with the feature parameters in the feature database; the first generation module generates the training data according to the feature database and the historical information of the plurality of passengers; the second generation module inputs the training set into the Xgboost algorithm engine to generate the reference model; the third generation module conducts cross validation on the reference model according to the test set to generate the passenger value assessment model; and finally, the passenger information is inputted into the passenger value assessment model through the value calculation module to generate the passenger value scores, and the passenger value scores are associated with the passenger information. Thus, the multidimensional information of the passengers of the airline is collected, and then the passenger value scores are generated in a machine learning mode. Accurate passenger value analysis data can be provided for marketing decisions and cost decisions of the airline, in order to provide better services for the passengers.
Appended aspects and advantages of the present invention will be partially supplied in the following description. Parts will become apparent from the following description, or will be known through the practice of the present invention.
Brief Description of the Drawings
Fig. 1 is a flow chart of a method for calculating passenger values of an airline provided in embodiments of the present invention.
Fig. 2 is a flow chart of another method embodiment for calculating passenger values of an airline provided in embodiments of the present invention.
Detailed Description
Embodiments of the present invention will be described below in detail. Examples of the embodiments are shown in drawings, wherein same or similar reference signs refer to same or similar elements or elements having same or similar functions from beginning to end. Embodiments described below by reference to the drawings are exemplary embodiments, and are used for explaining the present invention, and shall not be understood as a limitation to the present invention.
The system and medium for calculating passenger values of the airline in the embodiments of the present invention will be described below by reference to the drawings.
In the existing method for calculating passenger values, the mode of directly converting the total consumption amount and/or the historical flight miles into the value scores is often adopted. The calculation method is simple in structure, leading to the inaccuracy of the generated score result and low reference value. In view of this, in the method for calculating passenger values of the airline proposed in the embodiments of the present invention, the historical information of the passengers is acquired, and the multidimensional feature database is generated according to the preset feature algorithm; and then the passenger value assessment model is generated according to the feature database through machine learning. Accurate passenger value analysis data can be provided for marketing decisions and cost decisions of the airline.
To better understand the above technical solution, exemplary embodiments of the present invention will be described below in more detail with reference to the drawings. Although the exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be realized in various forms, and shall not be limited by the embodiments elaborated herein. On the contrary, the purpose of providing the embodiments is to understand the present invention more thoroughly and to completely communicate the scope of the present invention to those skilled in the art.
To better understand the above technical solution, the above technical solution will be detailed below in combination with drawings of the description and specific implementation.
Fig. 1 is a flow chart of a system for calculating passenger values of an airline provided in embodiments of the present invention. As shown in Fig. 1, the system for calculating passenger values of the airline comprises:
step 101: presetting a feature database, wherein the feature database comprises a feature parameter set and feature algorithms in one-to-one correspondence with feature parameters in the feature parameter set.
Specifically, the feature parameter set can be obtained by analyzing and extracting passenger data accumulated in the airline for a long time. To ensure the accuracy of the finally generated passenger value scores and the reference value of the data, a large number of feature parameters are extracted through analysis and verification, and corresponding feature algorithms are formulated with respect to the features of the feature parameters.
As an example, the feature parameter set includes but not limited to one or more of a variation tendency parameter, a bad experience parameter, a travel parameter, a booking parameter, an integral parameter, an attribute parameter and a state parameter. Further, the variation tendency parameter also comprises a maximum booking time interval parameter, an average booking time interval parameter, a maximum advance booking time parameter, an average advance booking time parameter, a frequent booking time point parameter, a buying channel parameter, a booking frequency parameter, a return parameter, a change frequency parameter, a booking amount parameter, a delay time parameter, etc.
The integral parameter includes but not limited to an integral accumulation channel, an accumulated integral number, an integral exchange score, an integral exchange frequency and an integral exchange channel. It can be understood that the integral accumulation channel is an acquisition mode of the integral, such as an integral acquired by using a credit card, an integral acquired by using a cash consumption mode, and an integral acquired through onboard consumption/purchase of duty-free goods; the accumulation integral number is the integral numerical values owned by passengers, such as a credit card integral of the passenger and an accumulated integral of account numbers of the passenger of the airline; the integral exchange scores mean the quantity of the integrals which have been exchanged with other commodities or services by the passengers; and the integral exchange frequency is the frequency of exchanging commodities or services with the integrals by the passengers. It can be understood that the activity degree of using the integrals by corresponding passengers can be further calculated through the integral exchange scores and the integral exchange frequency; and the integral exchange channel is the kind of specific commodities or services exchanged by the passengers using the integrals.
Optionally, many feature algorithms are formulated according to the characteristics of the feature parameters and needs of the finally outputted data. The feature algorithm with the feature parameter the maximum booking time interval parameter is illustrated below:
(1) booking time in a preset time threshold corresponding to each of the passenger information is filtered out according to a booking time field in the historical information of the passenger by using a Filter algorithm;
(2) the booking time of the previous record in Partition field is acquired by using Lag function by taking passenger ID as the Partition field and taking the booking time as an ascending sort field, and is recorded as Lag (booking time); and a variable booking time difference=booking time -Lag (booking time) is further generated; and (3) Max function is used for the variable booking time difference by taking passenger ID as Group field to obtain a maximum booking time interval.
As another example, the attribute parameter includes but not limited to number of joint pedestrians, social influence information and interest-related information.
The number of joint pedestrians refers to the information of number of people which walk together with the passengers when walking together with others. It can be understood that, when the passenger walks together with others, the value generated by the passenger shall be more than the value generated by the passenger self, and shall include the value generated by the joint pedestrians.
The social influence information includes but not limited to the attention of the account number of each social medium of the passenger and the social influence of people in close relationship with the passenger. The attention of the account number of each social medium means the number of fans of the account number in each social medium, the number of friends and response number of others in state update information.
The interest-related information means whether the passenger and the airline sign a contract or have other dealings which generate the actual commercial value. Namely, it is judged whether the passenger brings other gains for the airline besides the shipping service through the interest-related information. For example, the passenger and the airline sign a freight contract, and the contract brings considerable gains for the airline each year. At this moment, the value of the passenger shall be not limited to the specific value generated by using the aviation service in the airline.
As another example, the state parameter includes but not limited to a passenger social age, a passenger member level and member social information.
The passenger social age and the passenger member level are generated by a member system established in the airline, and can represent the consumption loyalty and the consumption level of the passenger for the airline to a certain extent. The member social information means the information generated by the passenger in a social process. For example, the passenger purchases a house and sells land in a certain place recently. This kind of informational is acquired through many modes,such as state update information of the corresponding accounts of the passenger in social media, network news and social platform information. Thus, the consumption type to be generated by the passenger can be further judged according to the social information of the passenger.
Step 102: acquiring historical information of a plurality of passengers, and generating training data according to the feature database and the historical information of the plurality of passengers, wherein the training data comprises a training set and a test set.
Specifically, the historical information of the plurality of passengers may be the historical information of the passengers in one airline. It can be understood that, the passenger value assessment model, which is finally obtained by generating training data according to the historical information of the passengers in one airline and training the passenger value assessment model subsequently according to the training data, outputs the passenger value scores, matched with the passenger values of multiple airlines, according to the inputted passenger information.
As an example, the historical information of the passengers in one airline is acquired;
training data is generated according to the feature data and the historical information of the passengers in one airline; and the passenger value assessment model is generated according to the training data. Then, the passenger information is inputted into the passenger value assessment model to generate passenger value scores, and the passenger value scores are associated with the passenger information. The passenger information may refer to Intranet actual information of the passenger, i.e., passenger information counted through the Intranet of one airline, and is an accurate actual result. The passenger value scores are network-wide value scores of the passengers, i.e., the passenger value scores of the historical information of the passengers corresponding to multiple airlines. Namely, after the training data is generated according to the historical information of the passengers in multiple airlines and the passenger value assessment model is generated according to the training data,when the passenger value assessment model is used, the passenger information counted through the Intranet of one airline can be inputted to generate the network-wide value scores of the passengers.
Step 103: inputting the training set into an Xgboost algorithm engine to generate a reference model.
The full name of Xgboost is eXtremeGradient Boosting which is a machine learning function library that focuses on a gradient boosting algorithm. As a supervising model, the Xgboost can establish a multilayer node.
Step 104: conducting cross validation on the reference model according to the test set to generate a passenger value assessment model.
Many modes can be used for cross validation. For example, K-fold cross validation mode divides the training data into K subsamples; one individual subsample is reserved as data for validating the model; and other K-l samples are used for training. Cross validation is repeated for K times. Each subsample is validated once. A single estimation is finally obtained through the average result of K times; or a validation mode is reserved.
As an example, the cross validation mode specifically comprises presetting model hyper-parameters and a model error index; inputting a test sample of the test set into the reference model to generate an estimated result; further judging whether an error between a test result of the test set and an estimated result is smaller than the model error index; if so, using the reference model as the passenger value assessment model; andif not, regulating the model hyper-parameters and conducting iterative training until the error between the generated estimated result and the test result is smaller than the model error index.
The step of judging whether an error between a test result of the test set and an estimated result is smaller than the model error index specifically includes: comparing the test result with the estimated result and recording the number of inconsistent items in the test result and the estimated result to generate a ratio of the number of inconsistent items to the total number of items of the test result. Further, it is judged whether the ratio is smaller than the model error index.
It can be understood that the purpose of cross validation is to judge the accuracy of the estimation result generated by the reference model by inputting the test sample of the test set into the reference model and comparing the estimated result generated by the reference model with the test result of the test set. The reference model with satisfactory accuracy is determined as the passenger value assessment model.
Step 105: inputting the passenger information into the passenger value assessment model to generate passenger value scores, and associating the passenger value scores with the passenger information.
In conclusion, the system for calculating passenger values of an airline in the embodiments of the present invention comprises: presetting the feature database, wherein the feature database comprises a feature parameter set and feature algorithms in one-to-one correspondence with feature parameters in the feature parameter set; acquiring historical information of a plurality of passengers, and generating training data according to the feature database and the historical information of the plurality of passengers, wherein the training data comprises a training set and a test set; inputting the training set into an Xgboost algorithm engine to generate a reference model; conducting cross validation on the reference model according to the test set to generate a passenger value assessment model; andinputting the passenger information into the passenger value assessment model to generate passenger value scores, and associating the passenger value scores with the passenger information. Thus, the multidimensional information of the passengers of the airline is collected, and then the passenger value scores are generated in a machine learning mode. Accurate passenger value analysis data can be provided for marketing decisions and cost decisions of the airline.
Fig. 2 is a flow chart of another system for calculating passenger values of an airline provided in embodiments of the present invention. As shown in Fig. 2, the method for calculating passenger values of the airline comprises:
step 201: presetting a feature database, wherein the feature database comprises a feature parameter set and feature algorithms in one-to-one correspondence with feature parameters in the feature parameter set.
Step 202: acquiring historical information of a plurality of passengers, and generating training data according to the feature database and the historical information of the plurality of passengers, wherein the training data comprises a training set and a test set.
Step 203: inputting the training set into an Xgboost algorithm engine to generate a reference model.
Step 204: conducting cross validation on the reference model according to the test set to generate a passenger value assessment model.
Step 205: inputting the passenger information into the passenger value assessment model to generate passenger value scores, and associating the passenger value scores with the passenger information.
It should be noted that the descriptions of steps 201-205 correspond to the descriptions of above steps 101-105. Therefore, see the descriptions of steps 101-105 for the descriptions of steps 201-205, and will not be repeated herein.
Step 206: presetting a passenger value table, wherein the passenger value table comprises a plurality of passenger value sections and estimated passenger values in one-to-one correspondence with the passenger value sections.
As an example, a better presetting mode of the passenger value table is described below in a tabular form:
| Passenger Value Score Sections | 0 | 0~10 | 10-20 | 20-30 | 30-40 |
| Estimated Passenger Value | 0 | 0-500 | 500-700 | 700-1000 | 1000-1500 |
| 40-50 | 50-60 | 60-70 | 70-80 | 80-90 | 90-99 | 99-100 |
| 1500-2100 | 2100-2500 | 2500-3000 | 3000-3500 | 3500-4500 | 4500-25000 | 25000+ |
Thus, by presetting the passenger value table, the passenger score values are further converted into estimated passenger values, so as to know passenger value information more intuitively.
Step 207: acquiring the estimated passenger values corresponding to the passenger value sections to which the passenger value scores belong according to the passenger value scores, and associating the estimated passenger values with the passenger value scores.
It can be understood that, after the passenger value scores are acquired, the estimated passenger values corresponding to the passenger value scores can be acquired according to the passenger value scores and the preset passenger value score table. Moreover, after the estimated passenger values are associated with the passenger value scores, a decision maker can conveniently screen passengers of different values according to the information based on the association relationship.
As another example, the method for calculating passenger values of the airline further comprises:
step 208: presetting a potential value threshold and a high-end passenger database.
Step 209: acquiring actual passenger values, generating potential passenger values according to the actual passenger values and the estimated passenger values, and associating the potential passenger values with the estimated passenger values.
The actual passenger values include but not limited to actual consumption generated by the passengers in the airline within certain time. The consumption includes aviation service consumption, onboard consumption or duty-free goods consumption. Further, the difference or ratio between the actual passenger values and the estimated passenger values is calculated to generate the potential passenger values. The potential passenger values are associated with the estimated passenger values. Thus, the decision maker of the airline can judge whether the corresponding passenger shall take an appropriate policy according to the potential passenger values to excavate the potential consumption value.
As an example, the Intranet actual values of the passengers (i.e., the actual passenger values) are acquired; the potential passenger values (i.e., estimation of potential passenger values which can be further excavated) are generated according to the Intranet actual values of the passengers and network-wide estimated values of the passengers (i.e., estimated passenger values); and the potential passenger values are associated with the network-wide estimated values of the passengers. The Intranet actual values of the passengers are actual consumption generated by the passengers in one airline within certain time; the network-wide estimated values of the passengers are the passenger value scores of the historical information of the passengers corresponding to multiple airlines; and the potential passenger values refer to the difference between the Intranet actual values of the passengers and the network-wide estimated values of the passengers. Namely, after the Intranet actual values of the passengers are acquired, the difference between the Intranet actual values of the passengers and the network-wide estimated values of the passengers generated by the passenger value estimation model is further calculated, so as to generate the potential passenger values; and the potential passenger values are associated with the network-wide estimated values of the passengers. Thus, under the condition of only acquiring the Intranet actual values of the passengers in one airline, the potential passenger values can be calculated according to the Intranet actual values of the passengers of one airline and the network-wide estimated values of the passengers generated by the passenger value estimation model, so as to judge the excavation space of the actual consumption generated by the passenger in one airline compared with the estimated consumption generated by the passenger in multiple airlines.
Step 210: judging whether the potential passenger values reach the potential value threshold; if so, entering step 211.
Step 211: storing the passenger information associated with the potential passenger values into the high-end passenger database.
Multiple modes can be used for judging whether the potential passenger values reach the potential value threshold. For example, a numerical value is set as a potential value threshold so as to further judge whether the potential passenger values are higher than the numerical value; or a ratio is set as a potential value threshold so as to further judge whether the ratio between the actual passenger values and the estimated passenger values is lower than the ratio. It can be understood that, when the numerical value of the potential value of a passenger is higher, the excavation value of the passenger is higher and the passenger is a passenger to be relatively valued. However, when the ratio between the actual passenger value of a passenger and the estimated passenger value is low, the passenger is a passenger with higher loyalty for the airline and is also a passenger to be extremely maintained.
The embodiment of the present invention provides a non-temporary readable storage medium, storing computer programs. The programs, when executed by a processor, enable the processor to execute the steps of the method for calculating passenger values of the airline.
It should be indicated that the explanation for the above method for calculating the passenger values of the airline in the embodiments of Fig. 1 and Fig. 2 is also applicable to the non-temporary readable storage medium of the embodiments, and will not be repeated herein.
The embodiment of the present invention provides a system for calculating passenger values of an airline. The system for calculating the passenger values of the airline comprises: a presetting module, an acquisition module, a first generation module, a second generation module, a third generation module, and a value calculation module.
The presetting module is used for presetting a feature database, wherein the feature database comprises a feature parameter set and feature algorithms in one-to-one correspondence with feature parameters in the feature parameter set.
The acquisition module is used for acquiring historical information of a plurality of passengers.
The first generation module is used for generating training data according to the feature database and the historical information of the plurality of passengers, wherein the training data comprises a training set and a test set.
The second generation module is used for inputting the training set into an Xgboost algorithm engine to generate a reference model.
The third generation module is used for conducting cross validation on the reference model according to the test set to generate a passenger value assessment model.
The value calculation module is used for inputting the passenger information into the passenger value assessment model to generate passenger value scores, and associating the passenger value scores with the passenger information.
It should be indicated that the explanation for the above method for calculating the passenger values of the airline in the embodiments of Fig. 1 and Fig. 2 is also applicable to the system for calculating the passenger values of the airline in the embodiment, and will not be repeated herein.
In conclusion, the system for calculating the passenger values of the airline in the embodiments of the present invention acquires the historical information of the plurality of passengers through the acquisition module,and presets the feature database through the presetting module according to the feature algorithms in one-to-one correspondence with feature parameters in the feature database. Further, the first generation module generates the training data according to the feature database and the historical information of the plurality of passengers. Further, the second generation module inputs the training set into the Xgboost algorithm engine to generate a reference model. Further, the third generation module conducts cross validation on the reference model according to the test set to generate a passenger value assessment model. Finally, the passenger information is inputted into the passenger value assessment model through the value calculation module to generate the passenger value scores, and the passenger value scores are associated with the passenger information. Thus, the multidimensional information of the passengers of the airline is collected, and then the passenger value scores are generated in a machine learning mode. Accurate passenger value analysis data can be provided for marketing decisions and cost decisions of the airline.
Those skilled in the art should understand that the embodiments of the present invention can provide a method, system or computer program product. Therefore, the present invention can adopt a form of a full hardware embodiment, a full software embodiment or an embodiment combining software and hardware. Moreover, the present invention can adopt a form of a computer program product capable of being implemented on one or more computer available storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer available program codes.
The present invention is described with reference to flow charts and/or block diagrams according to the method, device (system) and computer program product in the embodiments of the present invention. It should be understood that each flow and/or block in the flow charts and/or block diagrams and a combination of flows and/or blocks in the flow charts and/or block diagrams can be realized through computer program instructions. The computer program instructions can be provided for a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that a device for realizing designated functions in one or more flows of the flow charts and/or one or more blocks of the block diagrams is generated through the instructions executed by the processor of the computer or other programmable data processing devices.
The computer program instructions can also be stored in a computer readable memory which can guide the computer or other programmable data processing devices to operate in a special mode, so that the instructions stored in the computer readable memory generate a manufactured product including an instruction device, the instruction device realizing designated functions in one or more flows of the flow charts and/or one or more blocks of the block diagrams.
The computer program instructions can also be loaded on the computer or other programmable data processing devices, so that a series of operation steps are executed on the computer or other programmable devices to generate processing realized by the computer. Therefore, the instructions executed on the computer or other programmable devices provide steps for realizing designated functions in one or more flows of the flow charts and/or one or more blocks of the block diagrams.
It should be noted that in claims, any reference mark between brackets shall not form limitations to the claims. The word include does not exclude the existence of components or steps not listed in the claims. Words a or one in the front of components do not exclude the existence of a plurality of the components. The present invention can be realized by means of hardware including a plurality of different components and by means of a computer for appropriate programming. In unit claims which list a plurality of apparatuses, a plurality of the apparatuses may be specifically embodied through the same hardware item. The use of the words of first, second, third, etc. does not indicate any sequence. These words can be interpreted as names.
Although preferred embodiments of the present invention are described, those skilled in the art can make additional alterations and amendments to the embodiments once knowing basic creative concepts. Therefore, the appended claims are interpreted to include the preferred embodiments and all the alterations and amendments which fall into the scope of the present invention.
Obviously, those skilled in the art could implement various modifications to and variations of the present invention without departing from the spirit and scope of the present invention. So, the present invention is intended to include the modifications and variations if the amendments and variations of the present invention belong to claims of the present invention and the equivalent technical scope.
In the illustration of the present invention, it should be understood that the terms such as first and second are only used for the purpose of description, rather than being understood to indicate or imply relative importance or hint the number of indicated technical features. Thus, the feature limited by first and second can explicitly or impliedly comprise one or more features. In the illustration of the present invention, the meaning of a plurality of is two or more unless otherwise clearly specified.
In the present invention, unless otherwise specifically regulated and defined, terms such as installation, connected, connecting, fixation and the like shall be understood in broad sense, and for example, may refer to fixed connection or detachable connection or integral connection,may refer to mechanical connection or electrical connection,and may refer to direct connection or indirect connection through an intermediate medium or inner communication of two elements or interaction relationship of two elements. For those ordinary skilled in the art, the specific meanings of the above terms in the present invention may be understood according to concrete conditions.
In the present invention, unless otherwise clearly specified and limited, a first feature is above or below a second feature may mean that the first feature and the second feature come into direct contact or the first feature and the second feature come into indirect contact through an intermediary. Moreover, the first feature is on, above and over the second feature may mean that the first feature is directly above or slightly above the second feature, or may just indicate that the horizontal height of the first feature is higher than that of the second feature. The first feature is under, below and beneath the second feature may mean that the first feature is directly below or slightly below the second feature, or may just indicate that the horizontal height of the first feature is lower than that of the second feature.
In the illustration of this description, the illustration of reference terms one embodiment, some embodiments, example, specific example or some examples, etc. means that specific features, structures, materials or characteristics illustrated in combination with the embodiment or example are included in at least one embodiment or example of the present invention. In this description, exemplary statements for the above terms shall not be interpreted to aim at the same embodiment or example. Moreover, the described specific features, structures, materials or characteristics can be combined appropriately in any one or more embodiments or examples. In addition, those skilled in the art can combine and integrate different embodiments or examples and features of different embodiments or examples illustrated in this description without conflict.
Although the embodiments of the present invention have been shown and described above, it will be appreciated that the above embodiments are exemplary and shall not be understood as limitations to the present invention. Those ordinary skilled in the art can make changes, amendments, replacements and variations to the above embodiments within the scope of the present invention.
Claims (10)
1. A machine-learning system for calculating customer values, characterized by comprising:
a presetting module used for presetting a feature database, wherein the feature database comprises a feature parameter set and feature algorithms in one-to-one correspondence with feature parameters in the feature parameter set;
an acquisition module used for acquiring historical information of a plurality of customers;
a first generation module used for generating training data according to the feature database and the historical information of the plurality of customers, wherein the training data comprises a training set and a test set;
a second generation module used for inputting the training set into an Xgboost algorithm engine to generate a reference model;
a third generation module used for conducting cross validation on the reference model according to the test set to generate a value assessment model; and a value calculation module used for inputting the customers information into the value assessment model to generate customers value scores, and associating the customers value scores with the customers information.
2. The system for calculating customer values of claim 1, wherein the customer values comprise passenger values of an airplane, and the customers comprise passenger of an airplane, and the customer information comprise passenger information, and the customers values scores comprise passenger value score.
3. The system for calculating customer values of claim 2, wherein the presetting module further comprises a passenger value table comprising a plurality of passenger value sections and estimated passenger values in one-to-one correspondence with the passenger value sections, thereby the value calculation module acquires the estimated passenger values corresponding to the passenger value sections to which the passenger value scores belong according to the passenger value scores, and associating the estimated passenger values with the passenger value scores.
4. The system for calculating customer values of claim 3, wherein the presetting module further comprises a potential value threshold and a high-end passenger database, and the acquisition module acquires the actual passenger values, and the value calculation module generates potential passenger values according to the actual passenger values and the estimated passenger values, and associating the potential passenger values with the estimated passenger values, and the system further comprises a first judging module used for judging whether the potential passenger values reach the potential value threshold, if so, storing the passenger information associated with the potential passenger values into the high-end passenger database.
5. The system for calculating customer values of claim 4, wherein the presetting module further comprises a model hyper-parameters database and a model error index, and the test set comprises a test sample inputted into the reference model to generate an estimated result, and the system further comprises a second judging module used for judging whether an error between a test result of the test set and an estimated result is smaller than the model error index, if so, using the reference model as the passenger value assessment model, or if not, regulating the model hyper-parameters and conducting iterative training until the error between the generated estimated result and the test result is smaller than the model error index.
6. The system for calculating customer values of any preceding claim, characterized in that the feature parameters comprise one or more of a variation tendency parameter, a bad experience parameter, a travel parameter, a booking parameter, an integral parameter, an attribute parameter and a state parameter.
7. The system for calculating customer values of claim 6, characterized in that the attribute parameter comprises number of joint pedestrians, social influence information and interest-related information.
8. The system for calculating customer values of claim 6 or 7, characterized in that the state parameter comprises a customer social age, a customer member level and member social information.
9. A non-temporary computer readable storage medium, storing computer programs, characterized in that the programs, when executed by a processor, realize the system for calculating customers values of any one of claims 1-7.
10. A method for calculating passenger values of an airline, characterized by comprising:
presetting a feature database, wherein the feature database comprises a feature parameter set and feature algorithms in one-to-one correspondence with feature parameters in the feature parameter set;
acquiring historical information of a plurality of passengers, and generating training data according to the feature database and the historical information of the plurality of passengers, wherein the training data comprises a training set and a test set;
inputting the training set into an Xgboost algorithm engine to generate a reference model;
conducting cross validation on the reference model according to the test set to generate a passenger value assessment model; and inputting the passenger information into the passenger value assessment model to generate passenger value scores, and associating the passenger value scores with the passenger information.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810408752.XA CN108596678A (en) | 2018-05-02 | 2018-05-02 | A kind of airline passenger value calculation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| GB201815344D0 GB201815344D0 (en) | 2018-11-07 |
| GB2573351A true GB2573351A (en) | 2019-11-06 |
Family
ID=63620421
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| GB1815344.5A Withdrawn GB2573351A (en) | 2018-05-02 | 2018-09-20 | Machine learning system and medium for calculating passenger values of airline |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN108596678A (en) |
| GB (1) | GB2573351A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110288142A (en) * | 2019-06-18 | 2019-09-27 | 国网上海市电力公司 | A kind of engineering based on XGBoost algorithm is exceeded the time limit prediction technique |
| CN112396342A (en) * | 2020-11-30 | 2021-02-23 | 中国民航信息网络股份有限公司 | Passenger value calculation method, device and server |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111898786A (en) * | 2019-05-05 | 2020-11-06 | 顺丰科技有限公司 | Electronic lock demand prediction method, system, equipment and storage medium |
| CN110751523A (en) * | 2019-10-21 | 2020-02-04 | 中国民航信息网络股份有限公司 | Method and device for discovering potential high-value passengers |
| CN110751395B (en) * | 2019-10-21 | 2023-07-04 | 中国民航信息网络股份有限公司 | Passenger journey state determining method, device and server |
| CN112215451A (en) * | 2020-07-21 | 2021-01-12 | 中国人民公安大学 | Differentiation security check method and system based on civil aviation passenger classification |
| CN111950623B (en) * | 2020-08-10 | 2023-11-14 | 中国平安人寿保险股份有限公司 | Data stability monitoring method, device, computer equipment and medium |
| CN112329543A (en) * | 2020-10-12 | 2021-02-05 | 易显智能科技有限责任公司 | Method and related device for evaluating driving ability process |
| CN112163785A (en) * | 2020-10-19 | 2021-01-01 | 科技谷（厦门）信息技术有限公司 | Civil aviation passenger personal influence assessment method based on big data and neural network |
| CN116166735B (en) * | 2023-04-21 | 2023-07-04 | 民航成都信息技术有限公司 | Aviation data processing method and device, electronic equipment and storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107194803A (en) * | 2017-05-19 | 2017-09-22 | 南京工业大学 | A kind of P2P nets borrow the device of borrower's assessing credit risks |
| CN107240005A (en) * | 2017-06-13 | 2017-10-10 | 携程旅游网络技术（上海）有限公司 | The commending system and method for air ticket addition product |
| CN107403325B (en) * | 2017-08-10 | 2021-05-11 | 中国民航信息网络股份有限公司 | Method and device for evaluating reliability of air ticket order |
-
2018
- 2018-05-02 CN CN201810408752.XA patent/CN108596678A/en active Pending
- 2018-09-20 GB GB1815344.5A patent/GB2573351A/en not_active Withdrawn
Non-Patent Citations (1)
| Title |
|---|
| None * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110288142A (en) * | 2019-06-18 | 2019-09-27 | 国网上海市电力公司 | A kind of engineering based on XGBoost algorithm is exceeded the time limit prediction technique |
| CN112396342A (en) * | 2020-11-30 | 2021-02-23 | 中国民航信息网络股份有限公司 | Passenger value calculation method, device and server |
| CN112396342B (en) * | 2020-11-30 | 2023-12-26 | 中国民航信息网络股份有限公司 | Passenger value calculation method, device and server |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108596678A (en) | 2018-09-28 |
| GB201815344D0 (en) | 2018-11-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| GB2573351A (en) | Machine learning system and medium for calculating passenger values of airline | |
| Heinrich et al. | Requirements for data quality metrics | |
| Morar | An overview of the consumer value literature–perceived value, desired value | |
| AU2020202909A1 (en) | Machine learning classification and prediction system | |
| Bai et al. | Supply-chain performance-measurement system management using neighbourhood rough sets | |
| CN102654909B (en) | SLA pro-jobs level establishing method and system | |
| US20120278091A1 (en) | Sales prediction and recommendation system | |
| US20130204822A1 (en) | Tools and methods for determining relationship values | |
| US20150134401A1 (en) | In-memory end-to-end process of predictive analytics | |
| WO2017205463A1 (en) | Systems and methods for providing user-specific results based on test-drive of product or service | |
| US20050131754A1 (en) | System and method for estimating the feasibility of outsourcing information technology services | |
| Ayu Nariswari et al. | Testing an AHP model for aircraft spare parts | |
| CN108921702A (en) | Garden trade and investment promotion method and device based on big data | |
| US9785731B1 (en) | Prescriptive reservoir asset management | |
| US20210012363A1 (en) | Device, method and computer-readable medium for analyzing customer attribute information | |
| US20210109906A1 (en) | Clustering model analysis for big data environments | |
| KR102515439B1 (en) | Method, device and system for matching beauty product manufacturers using influencer information and brand maker information based on artificial intelligence | |
| Besbes et al. | Pricing analytics for rotable spare parts | |
| Zhu et al. | Consumer preference analysis based on text comments and ratings: A multi-attribute decision-making perspective | |
| Esmaeili et al. | Personalizing group recommendation to social network users | |
| Lee et al. | Nonprofits’ online social capital and charitable support | |
| CN107992879A (en) | The method of the potential potential loss risk of automatic decision vehicle maintenance user | |
| Ha et al. | Social filtering using social relationship for movie recommendation | |
| CN109766440B (en) | Method and system for determining default classification information for object text description | |
| Franceschini et al. | Standardisation of quality and reliability tests in the auto-parts industry: a structured approach concerning thermal systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| WAP | Application withdrawn, taken to be withdrawn or refused ** after publication under section 16(1) |