CN105512773A - Passenger travel destination prediction method and device - Google Patents
Passenger travel destination prediction method and device Download PDFInfo
- Publication number
- CN105512773A CN105512773A CN201510994069.5A CN201510994069A CN105512773A CN 105512773 A CN105512773 A CN 105512773A CN 201510994069 A CN201510994069 A CN 201510994069A CN 105512773 A CN105512773 A CN 105512773A
- Authority
- CN
- China
- Prior art keywords
- eigenwert
- forecast model
- passenger
- training
- travelling
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 67
- 238000012549 training Methods 0.000 claims description 64
- 241001269238 Data Species 0.000 claims description 16
- 238000003066 decision tree Methods 0.000 claims description 14
- 238000012706 support-vector machine Methods 0.000 claims description 7
- 238000004321 preservation Methods 0.000 claims description 2
- 230000006399 behavior Effects 0.000 description 15
- 238000004590 computer program Methods 0.000 description 7
- 238000012795 verification Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 230000003542 behavioural effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000004140 cleaning Methods 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- VUFNLQXQSDUXKB-DOFZRALJSA-N 2-[4-[4-[bis(2-chloroethyl)amino]phenyl]butanoyloxy]ethyl (5z,8z,11z,14z)-icosa-5,8,11,14-tetraenoate Chemical compound CCCCC\C=C/C\C=C/C\C=C/C\C=C/CCCC(=O)OCCOC(=O)CCCC1=CC=C(N(CCCl)CCCl)C=C1 VUFNLQXQSDUXKB-DOFZRALJSA-N 0.000 description 1
- 206010048669 Terminal state Diseases 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000003860 storage 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Tourism & Hospitality (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a passenger travel destination prediction method. The passenger travel destination prediction method comprises the following steps that: passenger travel behavior records are read from a passenger panoramic view; the eigenvalues of the passenger travel behavior records are extracted and calculated; and a prediction model is called to calculate the eigenvalues, so that prediction results of passenger travel destinations are obtained. Corresponding, the invention also discloses a passenger travel destination prediction device. With the passenger travel destination prediction method and device of the invention adopted, the technical problem that passenger travel destinations cannot be predicted in the current aviation business field can be solved, and an aviation business system can provide more targeted services for the passengers at each contact point of the passengers.
Description
Technical field
The present invention relates to data processing technique, particularly relate to a kind of travelling object Forecasting Methodology and device.
Background technology
Over nearly 20 years, have benefited from the fast development of scientific-technical progress and Internet technology, the technical merit of CAAC experienced by leap large several times, this is reflected in all links that passenger seizes the opportunity, comprise passenger's Reserved seating, buy passenger ticket, value machine from processes such as airport departures from port, occurred the new technologies such as electronic passenger ticket, mobile phone value machine, online duty machine, the behavior of passenger is also more and more diversified simultaneously, these all can be deposited in data warehouse in the form of data, are that excavation waited by huge precious deposits.
Travelling object is a terminal state of travelling, and airplane then just a process for the attainment of one's purpose and means, all travel behaviour details all can reflect trip purpose.
In Civil Aviation Industry to the trip purpose analysis of passenger so far or a blank.In existing passenger data, how not real trip purpose mark, find travelling object trace and sensing from numerous and complicated passenger's behavior data, and predict trip purpose by building model, is urgent problem.
Summary of the invention
For solving the technical matters of existing existence, the embodiment of the present invention provides a kind of travelling object Forecasting Methodology and device.
For achieving the above object, the technical scheme of the embodiment of the present invention is achieved in that
A kind of travelling object Forecasting Methodology, described method comprises:
Travelling behavior record is read from passenger's panoramic view;
Extract and calculate the eigenwert of described travelling record;
Call forecast model to calculate described eigenwert, obtain predicting the outcome of travelling object.
Wherein, before reading travelling record, described method also comprises: set up described forecast model based on support vector machines decision Tree algorithms.
Wherein, before reading travelling behavior record, described method also comprises the step setting up described forecast model, and this step comprises:
Gather many training datas, what training data described in every bar comprised that a passenger once goes on a journey goes out line item and trip purpose;
Extract and calculate the eigenwert of described many training datas respectively;
Based on the eigenwert training forecast model of described many training datas;
Verify the accuracy rate of described forecast model;
Preserve the forecast model by checking.
Wherein, described eigenwert can comprise following any one or combination in any: regular guest's card rank, date type, passenger's age range, sex, type of credential, departure time type, OD destination type, book tickets that sky number interval, value machine mode, advance value machine time, leg are international and domestic in advance, colleague's relationship, ticket booking channel, whether use big customer to encode, whether buy round ticket, freight space type, discount rate interval, team identification.
Wherein, before setting up described forecast model, described method also comprises: structural attitude value computation rule, and selects eigenwert.
Wherein, described training forecast model comprises: the file reading described eigenwert, using every a line variable of described file as input, calls the training that SVM decision Tree algorithms carries out forecast model; If the eigenwert read is not numeric type, then advanced line number value process.
Wherein, described preservation forecast model is: preserve the forecast model by checking with the form of text.
A kind of travelling object prediction unit, described device comprises:
Read module, for reading travelling record from passenger's panoramic view;
Eigenwert module, for extracting and calculating the eigenwert of described travelling record;
Prediction module, calculating described eigenwert for calling forecast model, obtaining predicting the outcome of travelling object.
Wherein, described prediction module, comprising: calculating sub module and set up submodule, wherein,
Set up submodule, for setting up described forecast model based on support vector machines decision Tree algorithms;
Described calculating sub module, calculating described eigenwert for calling described forecast model, obtaining predicting the outcome of travelling object.
Wherein, described read module, also for gathering many training datas, what training data described in every bar comprised that a passenger once goes on a journey goes out line item and trip purpose;
Described eigenwert module, also for extracting and calculate the eigenwert of described many training datas respectively;
Described prediction module set up submodule, specifically for based on described many training datas eigenwert training forecast model, verify the accuracy rate of described forecast model, and preserve by checking forecast model.
Wherein, described eigenwert can comprise following any one or combination in any: regular guest's card rank, date type, passenger's age range, sex, type of credential, departure time type, OD destination type, book tickets that sky number interval, value machine mode, advance value machine time, leg are international and domestic in advance, colleague's relationship, ticket booking channel, whether use big customer to encode, whether buy round ticket, freight space type, discount rate interval, team identification.
Wherein, described eigenwert module, also for structural attitude value computation rule, and selects eigenwert; And, for based on described eigenwert computation rule, extract and calculate eigenwert.
Wherein, described submodule of setting up, for training forecast model, comprising: the file reading described eigenwert, using every a line variable of described file as input, calls the training that SVM decision Tree algorithms carries out forecast model; If the eigenwert read is not numeric type, then advanced line number value process.
Wherein, described submodule of setting up is for preserving the forecast model by checking with the form of text.
The travelling object Forecasting Methodology of the embodiment of the present invention and device, travel behaviour record can be seized the opportunity based on passenger, automatic Prediction goes out the object that passenger goes on a journey at every turn, solve current aerospace business scope and cannot carry out for travelling object the technical matters predicted, analysis for aviation services provides a kind of dimension more going deep into user, make aviation services can support the personalized service of passenger better, be convenient to aviation services system and serve targetedly for it provides to have more at each contact point of passenger.
Accompanying drawing explanation
In accompanying drawing (it is not necessarily drawn in proportion), similar Reference numeral can describe similar parts in different views.The similar reference numerals with different letter suffix can represent the different examples of similar parts.Accompanying drawing generally shows each embodiment discussed herein by way of example and not limitation.
Fig. 1 is the schematic flow sheet of embodiment of the present invention travelling object Forecasting Methodology;
Fig. 2 is embodiment of the present invention forecast model Establishing process schematic diagram;
Fig. 3 is the composition structural representation of embodiment of the present invention travelling object prediction unit.
Embodiment
The embodiment of the present invention provides a kind of travelling object Forecasting Methodology and device, can seize the opportunity travel behaviour record based on civil aviation passenger, the trip purpose that prediction passenger is each.In other words, the embodiment of the present invention can seize the opportunity record from passenger, extract and cleaning basic data, analyze the behavioral characteristic of all links in travelling, utilizing passenger's identity characteristic, colleague's relationship, stroke feature, individual seizes the opportunity the information such as preference to predict the trip purpose of passenger.
Embodiment of the present invention travelling object Forecasting Methodology, as shown in Figure 1, mainly can comprise the steps:
Step 101: read the travelling record in passenger's panoramic view;
In this step, read the travelling record in passenger's panoramic view, obtain some field of passenger's behavior detailed data, for the calculating of eigenwert is prepared.
Here, prediction travelling object needs the whole process behavioral data gathering passenger, from ticket booking all to a series of contact point of drawing a bill, depart from port can be returned by various infosystem and be stored in centring system, these data through cleaning arrangement be stored in passenger's panoramic view.The data of the embodiment of the present invention just derive from this passenger's panoramic view.Wherein, passenger's panoramic view is the Data Integration carried out from the angle of passenger, and data granularity is leg level, includes passenger and draws a bill from making a reservation to, then to the whole process record of departing from port, therefore claim " panoramic view ".
Especially, read passenger's aphorama chart of every day, this passenger's aphorama chart is present in data warehouse, an often leg level information of a capable expression passenger, comprising the most detailed data of all behaviors of passenger, is without finished raw data, and is not suitable for directly as eigenwert.
Step 102: characteristics extraction and calculating;
The embodiment of the present invention the is prespecified computation rule of eigenwert, data step 101 read are as input and calculate in conjunction with pre-configured reference table according to the computation rule of eigenwert, can obtain N number of eigenwert.
Especially, the Data Source of the embodiment of the present invention is the passenger's panoramic view being deployed in Hadoop platform, contains more than 100 passenger's behavior field, and N number of field of, reflection key business point typical according to the feature selection and comparison of navigation business is as eigenwert.The many aspects such as this N number of eigenwert covers booking, draws a bill, departs from port, passenger's feature, can reflect behavioural characteristic and the trip purpose of passenger truly.
Step 103: call forecast model and calculate, obtains travelling object.
In the embodiment of the present invention, after step 102 calculates eigenwert, then call the forecast model preserved in advance and calculate, the result obtained is travelling object.Particularly, call forecast model, be input as the eigenwert file that step 102 obtains, return results, this result is a file, and the often row of this file comprises: current passenger goes out some fields of line item at every turn and predicts the outcome.
The embodiment of the present invention uses support vector machine (SVM, SupportVectorMachine) optimizing decision tree algorithm sets up described forecast model, the support of this algorithm to multidimensional data is relatively good, be applicable to the data with multiple eigenwert, relatively meet the feature of passenger data, passenger's panoramic view comprises more than 100 passenger's behavior field, through abstract calculating, tens passenger's preference dimensions can be obtained as eigenwert, these passenger's preferences more typically can reflect behavior and the feature of a passenger, must consider as the key element setting up forecast model.
Wherein, before step 101, described method also comprises the process building described forecast model, and namely grow out of nothing generation forecast model.The embodiment of the present invention utilizes SVM decision Tree algorithms to set up forecast model, eigenwert comprises N number of eigenwert, this algorithm adopts the thought of cross validation, raw data is divided into groups, a part is as training set, another part, as checking collection, first carries out the training of forecast model with training set, the forecast model of training and obtaining tested by recycling checking collection.The process building described forecast model can comprise Model Selection, training data collection, verification model, the forecast model trained is carried out the process of preserving etc., and building process as shown in Figure 2, mainly comprises the steps:
Step 201: select prediction algorithm;
First select suitable forecast model, and predict the outcome according to default four kinds of Commercial Air Service feature.The embodiment of the present invention, according to business experience for many years and the analysis to business, determines that first presetting four predicts the outcome, i.e. commercial affairs, travel, go to school, be on home leave, and mostly seizes the opportunity object to cover.
In conjunction with the data structure of business characteristic, passenger's panoramic view and the advantage of SVM algorithm, embodiment of the present invention choice for use based on the classification tree algorithm of SVM as forecast model, optimum efficiency can be obtained between the complexity and academic ability of model according to limited eigenwert and sample information, enormously simplify the problems such as common classification and recurrence, pay close attention to the conversion of service logic to algorithm model.
Step 202: structural attitude value computation rule, selects N number of eigenwert;
In fact, can according to Commercial Air Service rule and the analysis experience accumulated, structural attitude value computation rule, and the rule being applicable embodiment of the present invention forecast model.
In this step, eigenwert computation rule is selected based on business experience, and the data basis of this eigenwert computation rule is passenger's panoramic view, and the computing formula of employing must meet and can reflect service logic, and be suitable for carrying out calculating.
Wherein, N number of eigenwert can comprise any one or its combination in any in following information: regular guest's card rank, date type (Spring Festival/Clear and Bright/Dragon Boat Festival/Valentine's Day/winter vacation/summer vacation/Christmas/May Day/National Day/working day/weekend), passenger's age range, sex, type of credential, departure time type, OD destination type, ticket booking sky number interval in advance, value machine mode, the advance value machine time, leg is international and domestic, colleague's relationship, ticket booking channel, big customer whether is used to encode, whether buy round ticket, freight space type, discount rate is interval, team identification etc.Each eigenwert has clear and definite computation process and explanation, is easy to safeguard transplant.
In addition, also need the reference table needed in structural attitude value computation process, calculate N number of eigenwert using passenger's panoramic view each row of data as input.These reference tables are formulated according to service logic, such as date type, be according to annual date range festivals or holidays, several large class will be divided into the date, be respectively the Spring Festival, Clear and Bright, the Dragon Boat Festival, Valentine's Day, winter vacation, summer vacation, Christmas, May Day, National Day, working day, weekend.Such as the departure time is divided into the morning, noon, evening according to definition.
Here, reference table and eigenwert are not relations one to one, and some eigenwert needs to use reference table, does not need reference table when some eigenwert calculates.For above-mentioned eigenwert content, need in the embodiment of the present invention to use referring to table: time interval definition of booking tickets in advance, the definition of advance value machine time interval, value machine type definition, discount section definition, date type definition, departure time type definition, household register reference table, tourist city definition.Above-mentioned reference table can be self-defined according to actual needs, repeats no more.
The eigenwert computation rule of the embodiment of the present invention is as shown in table 1 below.
Table 1
Step 203: gather training data;
Gather training data, this training data is used for setting up forecast model, this training data comprises travelling detailed data and the real trip purpose of passenger, and cover mention in step 201 all to predict the outcome, namely cover commercial affairs, travel, go to school, four aspects of being on home leave, and the quantity of training data is the more more conducive to obtaining more reasonably forecast model, in the embodiment of the present invention, the collection of training data will meet international flight segments and domestic flight segments all covers, four kinds of trip purposes have, the annual condition such as all to cover in 12 month.Such as, this training data can be altogether more than the data of 10000 travel behaviours and real trip purpose thereof for more than 100 true passengers.It should be noted that training data will ensure is real data, and will mark real trip purpose.
Here, every bar training data is for a travel behaviour of a passenger, and except increasing a field representing trip purpose, form and the field of every bar training data are the same with the data in passenger's panoramic view.
Step 204: the eigenwert extracting and calculate described training data;
Here, the eigenwert computation rule of specifying according to step 202 when extracting eigenwert, carries out eigenwert calculating to the training data that step 203 obtains, obtains altogether N number of eigenwert.
Wherein, the computing formula of each eigenwert is very clear and definite, and sky number interval of such as booking tickets in advance, reads the date of embarking on journey of the current trip leg of current passenger, booking date, PNR date created from passenger's panoramic view tables of data.If this leg booking date is not empty, then this leg embark on journey the date deduct this leg booking the date, otherwise embark on journey in this leg, the date deducts PNR date created, the number of days obtained is number of days of booking tickets in advance, based on this section definition of number of days number of days and association is booked tickets in advance of booking tickets in advance, then can be booked tickets in advance a day number interval, be the integer of 1 to 5.Then this eigenwert is exactly an integer.Computation rule for each eigenwert can refer to table 1 above, repeats no more.
This stage needs to formulate some reference tables, such as age range definition list, in advance book tickets number of days interval table, festivals or holidays definition list etc. carry out aided solving, illustrated above, repeated no more herein.
Particularly, when extracting and calculate eigenwert, according to the passenger's behavior data read from passenger's panoramic view, due to data volume too much and can not directly use, resolve its data field and calculate according to pre-configured computation rule, obtaining the value of N number of eigenwert, result is a text, often capable N number of value, represents N number of eigenwert.Such as, certain eigenwert is " date ", span require be " Spring Festival/Clear and Bright/Dragon Boat Festival/Valentine's Day/winter vacation/summer vacation/Christmas/May Day/National Day/working day/weekend " in one.When calculating this eigenwert, the local date of the corresponding travelling read from passenger's panoramic view is such as 20140131, judge that this date is the Spring Festival through pre-configured computation rule, then calculate the value of its eigenwert " date " for " Spring Festival ".
Calculate in the manner described above, can obtain N number of eigenwert for each passenger corresponding training data of going on a journey at every turn, these eigenwerts are passed to following training module as one group of variable and are calculated.In the embodiment of the present invention, send N number of eigenwert of many training datas to training module with the form of text, every a line of text file comprises a stack features variable, the corresponding training data of every a line.
Step 205: training forecast model;
In this step, call algorithmic function, obtain the optimized parameter of forecast model.Call the SVM decision Tree algorithms of Python exploitation and train, process comprises: search for most suitable parameter, matching etc. by net method.
Concrete, read eigenwert file, using one of every a line group of variable as input, call the training that SVM decision Tree algorithms carries out forecast model.If the eigenwert read is not numeric type, need advanced line number value process, N number of eigenwert of final core algorithm function call is all numeric form, be adopt GridSearchCV to apply three (3-fold) cross validations in the function of acquiescence, obtain the optimized parameter of forecast model and preserve.Such as, algorithm is the algorithm bag calling Python, and the parameter of use is: ' rf__n_estimators':[10,20,30,40,50,60,70,80,90,100], cv=10, verbose=0, n_jobs=2.
Step 206: checking forecast model;
This step is used to a kind of statistical analysis technique verifying and check forecast model performance, comprises the checking of accuracy rate and travelling speed.
Gather verification msg, this verification msg is identical with the content that above-mentioned training data comprises, comprise the trip detailed data of passenger and real trip purpose data, and the various aspects contained in predicting the outcome, the eigenwert of verification msg is calculated based on the mode that step 202 is identical with 203, this eigenwert input current predictive model is carried out calculating and being predicted the outcome, predicting the outcome of obtaining and real trip purpose data in verification msg are compared, calculate accuracy rate, if rate of accuracy reached is to the requirement condition preset, so think that forecast model enters step 207 by checking, if accuracy rate does not reach default requirement condition, so think current predictive model not by checking, then repetition training process, namely step 203 is turned back to, re-start the training of forecast model, until rate of accuracy reached is to re-set target.
The once trip of the corresponding passenger of the every bar data of verification msg here, comprise travelling detailed data and the real trip purpose of passenger, its form is identical with above-mentioned training data with field, namely represents the field of trip purpose compared to the many increases of every bar data one in passenger's panoramic view.Such as, the verification msg of this step can select 1000.
Step 207: preserve forecast model.
Here, the forecast model that step 206 is verified is an available forecast model, includes the parameter optimized in this forecast model, reaches the accuracy rate of expection and portable.In this step, this available forecast model is preserved, follow-up carry out the prediction of travelling object time directly can quote this forecast model and predicted the outcome.Here, forecast model is saved as a text, the text calling this forecast model during use can obtain the trip purpose of passenger.
The embodiment of the present invention, through multiple steps such as Model Selection, feature selecting and extraction, the training of forecast model, the checkings of forecast model, establishes a complete travelling object forecast model.This forecast model has simple feature efficiently, may operate on multiple platform.Travelling object forecast model is implemented, in use procedure, to be deployed to by its algorithm routine on server and can to use.
It should be noted that, the algorithm realizing the embodiment of the present invention can adopt MapReduce parallel computation frame, is applicable to operating in Hadoop platform, to support the calculating of super large data volume.In addition, the embodiment of the present invention is applicable to Linux, Windows system, and be applicable to the parallel computation of big data quantity, its input and output are text, well can merge with other programs.
Accordingly, a kind of travelling object prediction unit of the embodiment of the present invention, as shown in Figure 3, this device mainly comprises: read module 31, eigenwert module 32 and prediction module 33, wherein:
Read module 31, for reading travelling record from passenger's panoramic view;
Eigenwert module 32, for extracting and calculating the eigenwert of described travelling record;
Prediction module 33, calculating described eigenwert for calling forecast model, obtaining predicting the outcome of travelling object.
Wherein, described prediction module 33, comprising: calculating sub module and set up submodule, wherein, sets up submodule, for setting up described forecast model based on support vector machines decision Tree algorithms; Described calculating sub module, calculating described eigenwert for calling described forecast model, obtaining predicting the outcome of travelling object.Here, described submodule of setting up, for training forecast model, comprising: the file reading described eigenwert, using every a line variable of described file as input, calls the training that SVM decision Tree algorithms carries out forecast model; If the eigenwert read is not numeric type, then advanced line number value process.Wherein, described submodule of setting up is for preserving the forecast model by checking with the form of text.
Wherein, described read module 31, also for gathering many training datas, what training data described in every bar comprised that a passenger once goes on a journey goes out line item and trip purpose; Described eigenwert module, also for extracting and calculate the eigenwert of described many training datas respectively; Describedly set up submodule, specifically for the eigenwert training forecast model based on described many training datas, verify the accuracy rate of described forecast model, and preserve the forecast model by checking.
Here, described eigenwert can comprise following any one or combination in any: regular guest's card rank, date type, passenger's age range, sex, type of credential, departure time type, OD destination type, book tickets that sky number interval, value machine mode, advance value machine time, leg are international and domestic in advance, colleague's relationship, ticket booking channel, whether use big customer to encode, whether buy round ticket, freight space type, discount rate interval, team identification.
Wherein, described eigenwert module 32, also for structural attitude value computation rule, and selects eigenwert; Based on described eigenwert computation rule, extract and calculate eigenwert.
Those skilled in the art should understand, embodiments of the invention can be provided as method, system or computer program.Therefore, the present invention can adopt the form of hardware embodiment, software implementation or the embodiment in conjunction with software and hardware aspect.And the present invention can adopt in one or more form wherein including the upper computer program implemented of computer-usable storage medium (including but not limited to magnetic disk memory and optical memory etc.) of computer usable program code.
The present invention describes with reference to according to the process flow diagram of the method for the embodiment of the present invention, equipment (system) and computer program and/or block scheme.Should understand can by the combination of the flow process in each flow process in computer program instructions realization flow figure and/or block scheme and/or square frame and process flow diagram and/or block scheme and/or square frame.These computer program instructions can being provided to the processor of multi-purpose computer, special purpose computer, Embedded Processor or other programmable data processing device to produce a machine, making the instruction performed by the processor of computing machine or other programmable data processing device produce device for realizing the function of specifying in process flow diagram flow process or multiple flow process and/or block scheme square frame or multiple square frame.
These computer program instructions also can be stored in can in the computer-readable memory that works in a specific way of vectoring computer or other programmable data processing device, the instruction making to be stored in this computer-readable memory produces the manufacture comprising command device, and this command device realizes the function of specifying in process flow diagram flow process or multiple flow process and/or block scheme square frame or multiple square frame.
These computer program instructions also can be loaded in computing machine or other programmable data processing device, make on computing machine or other programmable devices, to perform sequence of operations step to produce computer implemented process, thus the instruction performed on computing machine or other programmable devices is provided for the step realizing the function of specifying in process flow diagram flow process or multiple flow process and/or block scheme square frame or multiple square frame.
The above, be only preferred embodiment of the present invention, be not intended to limit protection scope of the present invention.
Claims (14)
1. a travelling object Forecasting Methodology, is characterized in that, described method comprises:
Travelling behavior record is read from passenger's panoramic view;
Extract and calculate the eigenwert of described travelling record;
Call forecast model to calculate described eigenwert, obtain predicting the outcome of travelling object.
2. method according to claim 1, is characterized in that, before reading travelling record, described method also comprises: set up described forecast model based on support vector machines decision Tree algorithms.
3. method according to claim 1 and 2, is characterized in that, before reading travelling behavior record, described method also comprises the step setting up described forecast model, and this step comprises:
Gather many training datas, what training data described in every bar comprised that a passenger once goes on a journey goes out line item and trip purpose;
Extract and calculate the eigenwert of described many training datas respectively;
Based on the eigenwert training forecast model of described many training datas;
Verify the accuracy rate of described forecast model;
Preserve the forecast model by checking.
4. the method according to claim 1 or 3, it is characterized in that, described eigenwert can comprise following any one or combination in any: regular guest's card rank, date type, passenger's age range, sex, type of credential, departure time type, OD destination type, book tickets that sky number interval, value machine mode, advance value machine time, leg are international and domestic in advance, colleague's relationship, ticket booking channel, whether use big customer to encode, whether buy round ticket, freight space type, discount rate interval, team identification.
5. the method according to claim 1,2 or 3, is characterized in that, before setting up described forecast model, described method also comprises: structural attitude value computation rule, and selects eigenwert.
6. method according to claim 3, is characterized in that, described training forecast model comprises:
Read the file of described eigenwert, using every a line variable of described file as input, call the training that SVM decision Tree algorithms carries out forecast model; If the eigenwert read is not numeric type, then advanced line number value process.
7. method according to claim 3, is characterized in that, described preservation forecast model is: preserve the forecast model by checking with the form of text.
8. a travelling object prediction unit, is characterized in that, described device comprises:
Read module, for reading travelling record from passenger's panoramic view;
Eigenwert module, for extracting and calculating the eigenwert of described travelling record;
Prediction module, calculating described eigenwert for calling forecast model, obtaining predicting the outcome of travelling object.
9. device according to claim 8, is characterized in that, described prediction module, comprising: calculating sub module and set up submodule, wherein,
Set up submodule, for setting up described forecast model based on support vector machines decision Tree algorithms;
Described calculating sub module, calculating described eigenwert for calling described forecast model, obtaining predicting the outcome of travelling object.
10. device according to claim 8 or claim 9, is characterized in that,
Described read module, also for gathering many training datas, what training data described in every bar comprised that a passenger once goes on a journey goes out line item and trip purpose;
Described eigenwert module, also for extracting and calculate the eigenwert of described many training datas respectively;
Described prediction module set up submodule, specifically for based on described many training datas eigenwert training forecast model, verify the accuracy rate of described forecast model, and preserve by checking forecast model.
Device described in 11. according to Claim 8 or 10, it is characterized in that, described eigenwert can comprise following any one or combination in any: regular guest's card rank, date type, passenger's age range, sex, type of credential, departure time type, OD destination type, book tickets that sky number interval, value machine mode, advance value machine time, leg are international and domestic in advance, colleague's relationship, ticket booking channel, whether use big customer to encode, whether buy round ticket, freight space type, discount rate interval, team identification.
Device described in 12. according to Claim 8 or 10, is characterized in that, described eigenwert module, also for structural attitude value computation rule, and selects eigenwert; And, for based on described eigenwert computation rule, extract and calculate eigenwert.
13. devices according to claim 10, is characterized in that, described submodule of setting up, for training forecast model, comprising:
Read the file of described eigenwert, using every a line variable of described file as input, call the training that SVM decision Tree algorithms carries out forecast model; If the eigenwert read is not numeric type, then advanced line number value process.
14. devices according to claim 10, is characterized in that, described submodule of setting up is for preserving the forecast model by checking with the form of text.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510994069.5A CN105512773A (en) | 2015-12-25 | 2015-12-25 | Passenger travel destination prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510994069.5A CN105512773A (en) | 2015-12-25 | 2015-12-25 | Passenger travel destination prediction method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105512773A true CN105512773A (en) | 2016-04-20 |
Family
ID=55720735
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510994069.5A Pending CN105512773A (en) | 2015-12-25 | 2015-12-25 | Passenger travel destination prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105512773A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106384022A (en) * | 2016-11-23 | 2017-02-08 | 东软集团股份有限公司 | Travel behavior forecasting method and device |
| CN106779214A (en) * | 2016-12-15 | 2017-05-31 | 南开大学 | A kind of multifactor fusion civil aviation passenger travel forecasting approaches based on topic model |
| CN108234596A (en) * | 2017-11-30 | 2018-06-29 | 中国联合网络通信集团有限公司 | Aviation information-pushing method and device |
| CN110378517A (en) * | 2019-06-20 | 2019-10-25 | 深圳壹账通智能科技有限公司 | Trip purpose prediction technique, calculates equipment and readable storage medium storing program for executing at device |
| CN110647929A (en) * | 2019-09-19 | 2020-01-03 | 京东城市(北京)数字科技有限公司 | Method for predicting travel destination and method for training classifier |
| CN110751576A (en) * | 2019-10-21 | 2020-02-04 | 中国民航信息网络股份有限公司 | Passenger travel determining method, device and server |
| CN113946757A (en) * | 2021-12-21 | 2022-01-18 | 深圳市活力天汇科技股份有限公司 | Method and device for identifying travel purpose of user and readable storage medium |
| CN116166735A (en) * | 2023-04-21 | 2023-05-26 | 民航成都信息技术有限公司 | Aviation data processing method and device, electronic equipment and storage medium |
| CN117787527A (en) * | 2024-02-26 | 2024-03-29 | 东莞市城建规划设计院 | Tour route intelligent planning method based on big data analysis technology |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008138837A1 (en) * | 2007-05-16 | 2008-11-20 | Amadeus S.A.S. | Method and system for automatically keeping travel data consistent between passenger reservation records and corresponding electronic tickets |
| CN103077490A (en) * | 2013-01-29 | 2013-05-01 | 阿坝师范高等专科学校 | Data processing method of tourist consumption information |
| CN104021189A (en) * | 2014-06-13 | 2014-09-03 | 中国民航信息网络股份有限公司 | Civil aviation traveler network generation method |
| CN104317789A (en) * | 2014-04-24 | 2015-01-28 | 科技谷（厦门）信息技术有限公司 | Method for building passenger social network |
| CN104361502A (en) * | 2014-04-24 | 2015-02-18 | 科技谷（厦门）信息技术有限公司 | Analysis method of passenger behavior data |
| CN104392337A (en) * | 2014-12-17 | 2015-03-04 | 常熟市虞城国际旅行社有限公司 | Multifunctional passenger information system |
| CN104463480A (en) * | 2014-12-17 | 2015-03-25 | 常熟市虞城国际旅行社有限公司 | Multifunctional passenger information management system |
-
2015
- 2015-12-25 CN CN201510994069.5A patent/CN105512773A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008138837A1 (en) * | 2007-05-16 | 2008-11-20 | Amadeus S.A.S. | Method and system for automatically keeping travel data consistent between passenger reservation records and corresponding electronic tickets |
| CN103077490A (en) * | 2013-01-29 | 2013-05-01 | 阿坝师范高等专科学校 | Data processing method of tourist consumption information |
| CN104317789A (en) * | 2014-04-24 | 2015-01-28 | 科技谷（厦门）信息技术有限公司 | Method for building passenger social network |
| CN104361502A (en) * | 2014-04-24 | 2015-02-18 | 科技谷（厦门）信息技术有限公司 | Analysis method of passenger behavior data |
| CN104021189A (en) * | 2014-06-13 | 2014-09-03 | 中国民航信息网络股份有限公司 | Civil aviation traveler network generation method |
| CN104392337A (en) * | 2014-12-17 | 2015-03-04 | 常熟市虞城国际旅行社有限公司 | Multifunctional passenger information system |
| CN104463480A (en) * | 2014-12-17 | 2015-03-25 | 常熟市虞城国际旅行社有限公司 | Multifunctional passenger information management system |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106384022B (en) * | 2016-11-23 | 2019-03-15 | 东软集团股份有限公司 | Travel behaviour prediction technique and device |
| CN106384022A (en) * | 2016-11-23 | 2017-02-08 | 东软集团股份有限公司 | Travel behavior forecasting method and device |
| CN106779214A (en) * | 2016-12-15 | 2017-05-31 | 南开大学 | A kind of multifactor fusion civil aviation passenger travel forecasting approaches based on topic model |
| CN106779214B (en) * | 2016-12-15 | 2020-08-28 | 南开大学 | Multi-factor fusion civil aviation passenger trip prediction method based on topic model |
| CN108234596A (en) * | 2017-11-30 | 2018-06-29 | 中国联合网络通信集团有限公司 | Aviation information-pushing method and device |
| CN110378517A (en) * | 2019-06-20 | 2019-10-25 | 深圳壹账通智能科技有限公司 | Trip purpose prediction technique, calculates equipment and readable storage medium storing program for executing at device |
| CN110647929B (en) * | 2019-09-19 | 2021-05-04 | 北京京东智能城市大数据研究院 | Method for predicting travel destination and method for training classifier |
| CN110647929A (en) * | 2019-09-19 | 2020-01-03 | 京东城市(北京)数字科技有限公司 | Method for predicting travel destination and method for training classifier |
| CN110751576A (en) * | 2019-10-21 | 2020-02-04 | 中国民航信息网络股份有限公司 | Passenger travel determining method, device and server |
| CN113946757A (en) * | 2021-12-21 | 2022-01-18 | 深圳市活力天汇科技股份有限公司 | Method and device for identifying travel purpose of user and readable storage medium |
| CN116166735A (en) * | 2023-04-21 | 2023-05-26 | 民航成都信息技术有限公司 | Aviation data processing method and device, electronic equipment and storage medium |
| CN117787527A (en) * | 2024-02-26 | 2024-03-29 | 东莞市城建规划设计院 | Tour route intelligent planning method based on big data analysis technology |
| CN117787527B (en) * | 2024-02-26 | 2024-04-26 | 东莞市城建规划设计院 | Tour route intelligent planning method based on big data analysis technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105512773A (en) | Passenger travel destination prediction method and device | |
| US8965754B2 (en) | Text prediction using environment hints | |
| CN110334289B (en) | Travel destination determining method and target user determining method | |
| CN112236787B (en) | System and method for generating personalized destination recommendations | |
| CN109934619A (en) | User's portrait tag modeling method, apparatus, electronic equipment and readable storage medium storing program for executing | |
| CN107767259A (en) | Loan risk control method, electronic installation and readable storage medium storing program for executing | |
| CN110516910A (en) | Declaration form core based on big data protects model training method and core protects methods of risk assessment | |
| CN105320724A (en) | New heuristic for optimizing non-convex function for learning to rank | |
| CN105183870A (en) | Urban functional domain detection method and system by means of microblog position information | |
| Li et al. | Capacity planning for mega container terminals with multi-objective and multi-fidelity simulation optimization | |
| CN113287124A (en) | System and method for ride order dispatch | |
| CN109933699A (en) | A kind of construction method and device of academic portrait model | |
| CN105898762A (en) | Base station optimization and deployment method and device | |
| CN109739965B (en) | Method, device and equipment for migrating cross-domain conversation strategy and readable storage medium | |
| CN106875670A (en) | Taxi concocting method based on gps data under Spark platforms | |
| Li et al. | Knowledge transfer and adaptation for land-use simulation with a logistic cellular automaton | |
| CN113424207B (en) | System and method for efficiently training understandable models | |
| Lin et al. | A stacking model for variation prediction of public bicycle traffic flow | |
| Samaras et al. | A prediction model of passenger demand using AVL and APC data from a bus fleet | |
| Urbanek | Data-driven transport policy in cities: A literature review and implications for future developments | |
| Khadivi et al. | Wikipedia in the tourism industry: Forecasting demand and modeling usage behavior | |
| KR20210065773A (en) | Big data based emotional information analysis and evaluation system and Driving method of the Same | |
| CN110348669B (en) | Intelligent rule generation method, intelligent rule generation device, computer equipment and storage medium | |
| CN107895217A (en) | Test and management method and application server | |
| US11361329B2 (en) | Systems and methods for generating optimized market plans |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination |