CN107240005A - The commending system and method for air ticket addition product - Google Patents
The commending system and method for air ticket addition product Download PDFInfo
- Publication number
- CN107240005A CN107240005A CN201710444047.0A CN201710444047A CN107240005A CN 107240005 A CN107240005 A CN 107240005A CN 201710444047 A CN201710444047 A CN 201710444047A CN 107240005 A CN107240005 A CN 107240005A
- Authority
- CN
- China
- Prior art keywords
- user
- data
- xgboost
- addition product
- air ticket
- 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 29
- 238000012360 testing method Methods 0.000 claims abstract description 44
- 238000012549 training Methods 0.000 claims abstract description 38
- 238000004364 calculation method Methods 0.000 claims abstract description 12
- 238000005259 measurement Methods 0.000 claims abstract description 8
- 238000002790 cross-validation Methods 0.000 claims description 10
- 230000035945 sensitivity Effects 0.000 claims description 10
- 238000009826 distribution Methods 0.000 claims 1
- 235000013399 edible fruits Nutrition 0.000 claims 1
- 230000000694 effects Effects 0.000 description 10
- 238000003066 decision tree Methods 0.000 description 7
- 230000033228 biological regulation Effects 0.000 description 4
- 238000013461 design Methods 0.000 description 2
- 238000005538 encapsulation Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 238000013499 data model Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000010998 test method 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
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G06Q50/40—
Abstract
The invention discloses the commending system of air ticket addition product and method, commending system includes model training module, model measurement authentication module and recommending module;Model training module trains Xgboost models by Xgboost algorithms and combined training data;Model measurement authentication module is tested Xgboost models by test data, and verifies by the corresponding the result of test data the output accuracy rate of Xgboost models;Recommending module is when exporting accuracy rate more than or equal to default accuracy rate, Xgboost models of reaching the standard grade, when receiving recommendation request, recommends air ticket addition product by the user data in Xgboost models and combination request.The present invention accurately meticulously recommends air ticket addition product using Xgboost models, improves user experience, and reduce the model training time using Spark Distributed Calculations.
Description
Technical field
The present invention relates to a kind of commending system of air ticket addition product and method.
Background technology
On-line purchase air ticket has become the main way that user purchases the air ticket, and user is while purchasing the air ticket, often
Can be purchased the air ticket addition product (such as insurance, quick security check passage and air ticket joint vending articles).Also have many consumers by
In purchase can be forgotten hastily, in order to give user more preferable buying experience, user can be likely to purchase when user purchases the air ticket
Product carry out acquiescence choose.Accurately acquiescence, which is chosen, can effectively improve user experience, and high the choosing of error rate can cause to use
Family is dissatisfied and then causes customer loss.The buying behavior of Accurate Prediction user, and carry out acquiescence choose recommendation it is particularly significant.
By development for many years, commending system is widely used in every field, is especially pushed away in personalization
Recommend the inside.For example, film and music etc..The commending system of main flow is mainly divided to two kinds, using record, utilizes feature.Utilize record
It is the recommendation based on collaborative filtering, using the purchasing history of user, finds out the close user of purchasing history, recommend for them similar
Product.Another is the recommendation based on model, and purchase problem is asked as simple classification using the feature of user and commodity
Topic, for example, whether a user likes a certain portion's film.There are many algorithms based on model, for example, decision tree, SVM
(Support Vector Machine, SVMs), Bayes etc..
And recommendation problem on air ticket addition product and conventional recommendation problem are different.First, it is on purchase
The recommendation problem of the later addition product of air ticket, what is recommended in the past is usually the pass that like product either user buys product
Joint product;Secondly, seldom, the product quantity of conventional recommendation is usually hundred to ten thousand or even million to the quantity of product;Finally, due to
Air ticket is not a product often bought (a general user only buys less than 4 for 1 year), and the historical information of user compares
It is few, and there are many new users to occur.
In summary, how to predict that user purchases the air ticket the row of addition product using the air ticket sequence information of relative deficiency
To be a difficult process.For a user, the air ticket frequency of purchase is not high, so unsuitable use cooperateed with
Filter the algorithm of a class.
The content of the invention
The technical problem to be solved in the present invention be in order to overcome in the prior art commending system can not recommend machine exactly
There is provided a kind of commending system of air ticket addition product and method for the defect of ticket addition product.
The present invention is to solve above-mentioned technical problem by following technical proposals:
A kind of commending system of air ticket addition product, its feature is that the commending system includes model training module, mould
Type tests authentication module and recommending module;
The model training module is used to train by Xgboost algorithms (a kind of sorting algorithm) and combined training data
Go out Xgboost models;
The model measurement authentication module is used to test the Xgboost models by test data, and leads to
The corresponding the result of the test data is crossed to verify the output accuracy rate of the Xgboost models;
The recommending module is used to, when the output accuracy rate is more than or equal to default accuracy rate, reach the standard grade described
Xgboost models, the recommending module is additionally operable to when receiving recommendation request, is asked by the Xgboost models and combination
User data in asking recommends air ticket addition product.
It is preferred that the training data includes the first user's history order data and the first user characteristic data, the survey
Trying data includes second user History Order data and second user characteristic, and the user data includes the 3rd user's history
Order data, the 3rd user characteristic data and current flight information.
It is preferred that the first user's history order data, the second user History Order data and the 3rd user go through
History order data includes origin, destination, the departure time, landing time and air ticket addition product purchase information;
First user characteristic data, the second user characteristic and the 3rd user characteristic data include year
Age, sex and price sensitivity;
The current flight information includes the current departure time, current departure city and ticket price.
It is preferred that the model training module is used for using Spark Distributed Calculations (a kind of efficient Distributed Calculation)
To train Xgboost models;And/or,
The model measurement authentication module is used to test the Xgboost models using cross-validation method.
It is preferred that the recommending module is additionally operable to reach the standard grade after the Xgboost models, test that (one kind is by right by A/B
The method of testing according to the facts tested) decision-making value is adjusted out, if the user's purchase probability gone out by the Xgboost model predictions is more than
Or during equal to the decision-making value, acquiescence chooses corresponding air ticket addition product, if user's purchase probability is less than described determine
During plan threshold value, cancellation is chosen to corresponding air ticket addition product.
A kind of recommendation method of air ticket addition product, its feature is that the recommendation method comprises the following steps:
S1, by Xgboost algorithms and combined training data train Xgboost models;
S2, by test data the Xgboost models are tested, and tested by the way that the test data is corresponding
Result is demonstrate,proved to verify the output accuracy rate of the Xgboost models;
S3, when the output accuracy rate is more than or equal to default accuracy rate, the Xgboost models of reaching the standard grade are described to push away
Recommend module to be additionally operable to when receiving recommendation request, pushed away by the user data in the Xgboost models and combination request
Recommend out air ticket addition product.
It is preferred that in step S1In, the training data includes the first user's history order data and the first user characteristics
Data;
In step S2In, the test data includes second user History Order data and second user characteristic;
In step S3In, the user data includes the 3rd user's history order data, the 3rd user characteristic data and worked as
Preceding Flight Information.
It is preferred that the first user's history order data, the second user History Order data and the 3rd user go through
History order data includes origin, destination, the departure time, landing time and air ticket addition product purchase information;
First user characteristic data, the second user characteristic and the 3rd user characteristic data include year
Age, sex and price sensitivity;
The current flight information includes the current departure time, current departure city and ticket price.
It is preferred that in step S1In, Xgboost models are trained using Spark Distributed Calculations;And/or, in step
S2In, the Xgboost models are tested using cross-validation method.
It is preferred that in step S3In, after the Xgboost models of reaching the standard grade, decision-making value is adjusted out by A/B tests, if
When the user's purchase probability gone out by the Xgboost model predictions is more than or equal to the decision-making value, acquiescence is chosen corresponding
Air ticket addition product, if user's purchase probability be less than the decision-making value, cancel to corresponding air ticket addition product
Choose.
On the basis of common sense in the field is met, above-mentioned each optimum condition can be combined, and produce each preferable reality of the present invention
Example.
The positive effect of the present invention is:
The commending system and method for the air ticket addition product that the present invention is provided accurately meticulously are pushed away using Xgboost models
Air ticket addition product is recommended out, accurately acquiescence can be carried out to the air ticket addition product required for user and is chosen, recommendation essence is improved
Degree, saves human cost, so as to improve user experience, and reduces the model training time using Spark Distributed Calculations,
To obtain required model in the short period of time.
Brief description of the drawings
Fig. 1 is the structural representation of the commending system of the air ticket addition product of present pre-ferred embodiments.
Fig. 2 is the flow chart of the recommendation method of the air ticket addition product of present pre-ferred embodiments.
Embodiment
The present invention is further illustrated below by the mode of embodiment, but does not therefore limit the present invention to described reality
Apply among a scope.
As shown in figure 1, the commending system 1 for the air ticket addition product that the present embodiment is provided includes model training module 11, mould
Type tests authentication module 12 and recommending module 13.
Specifically, model training module 11 is used to train Xgboost by Xgboost algorithms and combined training data
Model, suitable data are chosen in historical data base and are used as data for training pattern.
In the present embodiment, the training data includes the first user's history order data and the first user characteristic data,
The first user's history order data include the first origin, the first destination, first departure time, first landing the time and
First air ticket addition product buys information, and first user characteristic data includes the first age, the first sex and the first price
Susceptibility, price sensitivity is used to characterize the information such as low price sensitive users or high price user, does not have in the present embodiment certainly
Body limits the type of training data, can accordingly be selected according to actual conditions.
Xgboost models have many parameters to need regulation, and the quality of parameter regulation has different effects.In the present embodiment
In, the most suitable decision tree quantity of Xgboost models, depth capacity, minimum splitting parameter and are adjusted out by training data
Practise the parameter such as speed, so as to obtain most suitable Xgboost models, how to adjust out suitable parameter need to according to actual conditions and
Required model judges, wherein, the influence of the quantity and depth capacity of decision tree to whole modelling effect is maximum, except above-mentioned 4
There is other specification beyond individual parameter, but it is little to influential effect, therefore no longer repeat one by one in the present embodiment.
In the present embodiment, model training module is used to train Xgboost models using Spark Distributed Calculations.
Due to air ticket addition product seldom, for each product, whether user can be bought as two classification problems, for two
Classification problem, gradient decline decision tree be a popular algorithm, although but the algorithm accuracy is high, prediction effect is good,
But need substantial amounts of time, for it is this kind of have a large amount of parallel computations the problem of, just can by by can be parallel part point
It is fitted on different servers to reduce the calculating time.Xgboost algorithms are selected according to recommended requirements and feature, and are passed through
Spark Distributed Calculations reduce the model training time so that obtain required model in the short period of time, and line is supplied after encapsulation
Upper system is called.
Model measurement authentication module 12 is used to test the Xgboost models by test data, and passes through
The test data corresponding the result verifies the output accuracy rate of the Xgboost models, is selected in historical data base
Suitable data are taken to be used as the data for test model.
In the present embodiment, the test data includes second user History Order data and second user characteristic,
The second user History Order data include the second origin, the second destination, second departure time, second landing the time and
Second air ticket addition product buys information, and the second user characteristic includes the second age, second sex and the second price
Susceptibility, the type of test data is not limited specifically in the present embodiment certainly, can be carried out according to actual conditions corresponding
Selection.
In the present embodiment, model measurement authentication module is used to carry out the Xgboost models using cross-validation method
Test.Cross-validation method is the basic skills of testing model quality, divides the data into K parts, in turn 1 part of work K numbers in
For test set, it is left K-1 parts as training set, selects the best model of average test effect.Now, can using cross-validation method
Further adjust out most suitable parameters.
Recommending module 13 is used for when the output accuracy rate is more than or equal to default accuracy rate, and reach the standard grade the Xgboost
Model, recommending module 13 is additionally operable to when receiving recommendation request, passes through the user in the Xgboost models and combination request
Data recommend air ticket addition product.
In the present embodiment, the user data include the 3rd user's history order data, the 3rd user characteristic data and
Current flight information, the 3rd user's history order data include the 3rd origin, the 3rd destination, the 3rd departure time,
3rd landing time and the 3rd air ticket addition product purchase information, the 3rd user characteristic data include the 3rd age, the 3rd
Sex and the 3rd price sensitivity, the current flight information include the current departure time, current departure city and ticket price,
Certainly do not limit the type of user data specifically in the present embodiment, can accordingly be selected according to actual conditions.
In the present embodiment, the Xgboost models are reached the standard grade for a period of time, during this, are adjusted out certainly by A/B tests
Plan threshold value, i.e., carry out control experiment, decision-making value, which is determined, is by using the Xgboost models and using other models
It is no to give tacit consent to the datum line chosen.Xgboost models can predict user's purchase probability, if pre- by the Xgboost models
When the user's purchase probability measured is more than or equal to the decision-making value, acquiescence chooses corresponding air ticket addition product, if described
When user's purchase probability is less than the decision-making value, cancels and the acquiescence of corresponding air ticket addition product is chosen, so as to be user
Recommend suitable air ticket addition product.
Relative to the existing suggested design using rule, Xgboost models can be by special to user and air ticket scene
More preferable study is levied, more accurate careful prediction is done, so as to improve user experience.Rule-based suggested design is mainly base
In business experience, human cost is big and artifical influence factor is big, tends to ignore some important informations.Use Xgboost moulds
Type does not need excessive human cost, can be learnt using the initial data of user and air ticket.And Distributed Calculation is utilized, parallel
Training pattern, so as to shorten the training time and obtain high-precision model.
As shown in Fig. 2 the present embodiment also provides a kind of recommendation method of air ticket addition product, the recommendation method include with
Lower step:
Step 101, by Xgboost algorithms and combined training data train Xgboost models.
In this step, suitable data are chosen in historical data base and are used as data for training pattern.
In this step, the training data includes the first user's history order data and the first user characteristic data, institute
Stating the first user's history order data includes the first origin, the first destination, first departure time, the first landing time and the
One air ticket addition product buys information, and it is quick that first user characteristic data includes the first age, the first sex and the first price
Sensitivity, price sensitivity is used to characterize the information such as low price sensitive users or high price user, certainly not specific in this step to limit
Determine the type of training data, can accordingly be selected according to actual conditions.
Xgboost models have many parameters to need regulation, and the quality of parameter regulation has different effects.In this step
In, the most suitable decision tree quantity of Xgboost models, depth capacity, minimum splitting parameter and are adjusted out by training data
Practise the parameter such as speed, so as to obtain most suitable Xgboost models, how to adjust out suitable parameter need to according to actual conditions and
Required model judges, wherein, the influence of the quantity and depth capacity of decision tree to whole modelling effect is maximum, except above-mentioned 4
There is other specification beyond individual parameter, but it is little to influential effect, therefore no longer repeat one by one in this step.
In this step, Xgboost models are trained using Spark Distributed Calculations.Due to air ticket addition product very
It is few, for each product, whether user can be bought as two classification problems, for two classification problems, gradient declines
Decision tree is a popular algorithm, although but the algorithm accuracy is high, prediction effect is good, need it is substantial amounts of when
Between, for it is this kind of have a large amount of parallel computations the problem of, just can be by can will parallel be partially distributed to different servers
On reduce the calculating time.Select Xgboost algorithms according to recommended requirements and feature, and by Spark Distributed Calculations come
Reduce the model training time so that obtain required model in the short period of time, called after encapsulation for inline system.
Step 102, by test data Xgboost models are tested and verified.
In this step, the Xgboost models are tested by test data, and passes through the test data
Corresponding the result verifies the output accuracy rate of the Xgboost models, and suitable data are chosen in historical data base
It is used as the data for test model.
In this step, the test data includes second user History Order data and second user characteristic, institute
Stating second user History Order data includes the second origin, the second destination, second departure time, the second landing time and the
Two air ticket addition products buy information, and it is quick that the second user characteristic includes the second age, second sex and the second price
Sensitivity, the type of test data is not limited specifically, can accordingly be selected according to actual conditions in this step certainly
Select.
In this step, the Xgboost models are tested using cross-validation method.Cross-validation method is to examine mould
The basic skills of type quality, divides the data into K parts, in turn using 1 part in K numbers evidence as test set, is left K-1 parts as instruction
Practice collection, select the best model of average test effect.Now, most suitable items can further be adjusted out using cross-validation method
Parameter.
Step 103, Xgboost models of reaching the standard grade, adjust out decision-making value, recommend machine according to recommendation request and decision-making value
Ticket addition product.
In this step, when the output accuracy rate is more than or equal to default accuracy rate, the Xgboost moulds of reaching the standard grade
Type, when receiving recommendation request, by the Xgboost models and to combine the user data in request attached to recommend air ticket
Plus product.
In this step, the user data includes the 3rd user's history order data, the 3rd user characteristic data and worked as
Preceding Flight Information, the 3rd user's history order data includes the 3rd origin, the 3rd destination, the 3rd departure time, the
Three landing times and the 3rd air ticket addition product purchase information, the 3rd user characteristic data include the 3rd age, the 3rd property
Other and the 3rd price sensitivity, the current flight information includes the current departure time, current departure city and ticket price, when
Do not limit the type of user data specifically in this step so, can accordingly be selected according to actual conditions.
In this step, the Xgboost models are reached the standard grade for a period of time, during this, decision-making is adjusted out by A/B tests
Threshold value, i.e., carry out control experiment, decision-making value has decided on whether by using the Xgboost models and using other models
Give tacit consent to the datum line chosen.Xgboost models can predict user's purchase probability, if passing through the Xgboost model predictions
When the user's purchase probability gone out is more than or equal to the decision-making value, acquiescence chooses corresponding air ticket addition product, if described use
When family purchase probability is less than the decision-making value, cancels and the acquiescence of corresponding air ticket addition product is chosen, so as to be pushed away for user
Recommend out suitable air ticket addition product.
The present embodiment provide air ticket addition product commending system and method using Xgboost models come accurately meticulously
Recommend air ticket addition product, accurately acquiescence can be carried out to the air ticket addition product required for user and is chosen, recommendation is improved
Precision, saves human cost, so as to improve user experience, and is reduced using Spark Distributed Calculations during model training
Between, to obtain required model in the short period of time.
Although the embodiment of the present invention is the foregoing described, it will be appreciated by those of skill in the art that this is only
For example, protection scope of the present invention is to be defined by the appended claims.Those skilled in the art without departing substantially from
On the premise of the principle and essence of the present invention, various changes or modifications can be made to these embodiments, but these changes and
Modification each falls within protection scope of the present invention.
Claims (10)
1. a kind of commending system of air ticket addition product, it is characterised in that the commending system includes model training module, model
Test authentication module and recommending module;
The model training module is used to train Xgboost models by Xgboost algorithms and combined training data;
The model measurement authentication module is used to test the Xgboost models by test data, and passes through institute
The corresponding the result of test data is stated to verify the output accuracy rate of the Xgboost models;
The recommending module is used for when the output accuracy rate is more than or equal to default accuracy rate, the Xgboost moulds of reaching the standard grade
Type, the recommending module is additionally operable to when receiving recommendation request, passes through the user in the Xgboost models and combination request
Data recommend air ticket addition product.
2. commending system as claimed in claim 1, it is characterised in that the training data includes the first user's history order numbers
According to and the first user characteristic data, the test data include second user History Order data and second user characteristic,
The user data includes the 3rd user's history order data, the 3rd user characteristic data and current flight information.
3. commending system as claimed in claim 2, it is characterised in that the first user's history order data, described second
User's history order data and the 3rd user's history order data include origin, destination, the departure time, landing the time and
Air ticket addition product buys information;
First user characteristic data, the second user characteristic and the 3rd user characteristic data include age, property
Other and price sensitivity;
The current flight information includes the current departure time, current departure city and ticket price.
4. commending system as claimed in claim 1, it is characterised in that the model training module is used for using Spark distributions
Formula calculates to train Xgboost models;And/or,
The model measurement authentication module is used to test the Xgboost models using cross-validation method.
5. commending system as claimed in claim 1, it is characterised in that the recommending module is additionally operable to the Xgboost that reaches the standard grade
After model, decision-making value is adjusted out by A/B tests, if the user's purchase probability gone out by the Xgboost model predictions is big
When the decision-making value, acquiescence chooses corresponding air ticket addition product, if user's purchase probability is less than described
During decision-making value, cancellation is chosen to corresponding air ticket addition product.
6. a kind of recommendation method of air ticket addition product, it is characterised in that the recommendation method comprises the following steps:
S1, by Xgboost algorithms and combined training data train Xgboost models;
S2, by test data the Xgboost models are tested, and pass through the corresponding checking knot of the test data
Fruit verifies the output accuracys rate of the Xgboost models;
S3, when the output accuracy rate is more than or equal to default accuracy rate, reach the standard grade the Xgboost models, the recommending module
It is additionally operable to when receiving recommendation request, recommends machine by the user data in the Xgboost models and combination request
Ticket addition product.
7. recommend method as claimed in claim 6, it is characterised in that in step S1In, the training data includes the first user
History Order data and the first user characteristic data;
In step S2In, the test data includes second user History Order data and second user characteristic;
In step S3In, the user data includes the 3rd user's history order data, the 3rd user characteristic data and current flight
Information.
8. recommend method as claimed in claim 7, it is characterised in that the first user's history order data, described second
User's history order data and the 3rd user's history order data include origin, destination, the departure time, landing the time and
Air ticket addition product buys information;
First user characteristic data, the second user characteristic and the 3rd user characteristic data include age, property
Other and price sensitivity;
The current flight information includes the current departure time, current departure city and ticket price.
9. recommend method as claimed in claim 6, it is characterised in that in step S1In, instructed using Spark Distributed Calculations
Practise Xgboost models;And/or, in step S2In, the Xgboost models are tested using cross-validation method.
10. recommend method as claimed in claim 6, it is characterised in that in step S3In, after the Xgboost models of reaching the standard grade,
Decision-making value is adjusted out by A/B tests, if the user's purchase probability gone out by the Xgboost model predictions is more than or equal to
During the decision-making value, acquiescence chooses corresponding air ticket addition product, if user's purchase probability is less than the decision-making value
When, cancellation is chosen to corresponding air ticket addition product.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710444047.0A CN107240005A (en) | 2017-06-13 | 2017-06-13 | The commending system and method for air ticket addition product |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710444047.0A CN107240005A (en) | 2017-06-13 | 2017-06-13 | The commending system and method for air ticket addition product |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107240005A true CN107240005A (en) | 2017-10-10 |
Family
ID=59987541
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710444047.0A Pending CN107240005A (en) | 2017-06-13 | 2017-06-13 | The commending system and method for air ticket addition product |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107240005A (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108304720A (en) * | 2018-02-06 | 2018-07-20 | 恒安嘉新(北京)科技股份公司 | A kind of Android malware detection methods based on machine learning |
CN108596678A (en) * | 2018-05-02 | 2018-09-28 | 陈思恩 | A kind of airline passenger value calculation method |
CN109166012A (en) * | 2018-09-21 | 2019-01-08 | 苏州创旅天下信息技术有限公司 | The method and apparatus of classification and information push for stroke predetermined class user |
CN109410075A (en) * | 2018-10-23 | 2019-03-01 | 广州市勤思网络科技有限公司 | Intelligence insurance recommended method and system based on Bayes |
CN109493200A (en) * | 2019-01-24 | 2019-03-19 | 深圳市活力天汇科技股份有限公司 | A kind of recommended method of air ticket trip commodity |
CN109492837A (en) * | 2018-12-29 | 2019-03-19 | 携程旅游网络技术(上海)有限公司 | Air ticket order insures method for pushing, device, electronic equipment, storage medium |
CN109886778A (en) * | 2019-01-29 | 2019-06-14 | 上海华程西南国际旅行社有限公司 | The recommended method and system of the tie-in sale product of air ticket |
CN109934512A (en) * | 2019-03-28 | 2019-06-25 | 努比亚技术有限公司 | A kind of training method and system of prediction model |
CN110288461A (en) * | 2019-05-20 | 2019-09-27 | 深圳壹账通智能科技有限公司 | Automatic verification method, electronic device and the storage medium of Products Show |
CN111144946A (en) * | 2019-12-27 | 2020-05-12 | 上海携程商务有限公司 | Revenue management method, system, medium, and electronic device for airline company |
CN111314869A (en) * | 2020-02-18 | 2020-06-19 | 中国联合网络通信集团有限公司 | Flow quota distribution method and device, electronic equipment and storage medium |
WO2022011947A1 (en) * | 2020-10-23 | 2022-01-20 | 平安科技(深圳)有限公司 | Transaction data processing method and apparatus, and computer device and storage medium |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1756336A (en) * | 2004-09-29 | 2006-04-05 | 松下电器产业株式会社 | Television channel commending system and commending method |
US20080046332A1 (en) * | 2006-08-18 | 2008-02-21 | Ben Aaron Rotholtz | System and method for offering complementary products / services |
CN103051730A (en) * | 2013-01-15 | 2013-04-17 | 合肥工业大学 | Multi-source information service-resource allocating system and IA-Min allocating method in cloud-computing business environment |
CN105550207A (en) * | 2015-12-02 | 2016-05-04 | 合一网络技术(北京)有限公司 | Information popularization method and device |
CN105654341A (en) * | 2015-12-28 | 2016-06-08 | 中国民航信息网络股份有限公司 | Aviation product recommendation system and aviation product recommendation method based on cloud service |
CN105893507A (en) * | 2016-03-30 | 2016-08-24 | 乐视控股(北京)有限公司 | Method and device for recommending information of derivative products |
CN106779867A (en) * | 2016-12-30 | 2017-05-31 | 中国民航信息网络股份有限公司 | Support vector regression based on context-aware recommends method and system |
-
2017
- 2017-06-13 CN CN201710444047.0A patent/CN107240005A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1756336A (en) * | 2004-09-29 | 2006-04-05 | 松下电器产业株式会社 | Television channel commending system and commending method |
US20080046332A1 (en) * | 2006-08-18 | 2008-02-21 | Ben Aaron Rotholtz | System and method for offering complementary products / services |
CN103051730A (en) * | 2013-01-15 | 2013-04-17 | 合肥工业大学 | Multi-source information service-resource allocating system and IA-Min allocating method in cloud-computing business environment |
CN105550207A (en) * | 2015-12-02 | 2016-05-04 | 合一网络技术(北京)有限公司 | Information popularization method and device |
CN105654341A (en) * | 2015-12-28 | 2016-06-08 | 中国民航信息网络股份有限公司 | Aviation product recommendation system and aviation product recommendation method based on cloud service |
CN105893507A (en) * | 2016-03-30 | 2016-08-24 | 乐视控股(北京)有限公司 | Method and device for recommending information of derivative products |
CN106779867A (en) * | 2016-12-30 | 2017-05-31 | 中国民航信息网络股份有限公司 | Support vector regression based on context-aware recommends method and system |
Non-Patent Citations (4)
Title |
---|
佚名: ""去哪儿"何去何从?", 《现代企业文化(上旬)》 * |
侯艳: "多家购票平台被曝"帮你消费":产品搭售 消费者不知情", 《广西质量监督导报》 * |
张昊;纪宏超;张红宇: "XGBoost算法在电子商务商品推荐中的应用", 《物联网技术》 * |
杨晓芳; 王喆; 姜海: "基于多项logit模型的在线机票代理商选择行为", 《清华大学学报(自然科学版)》 * |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108304720B (en) * | 2018-02-06 | 2020-12-11 | 恒安嘉新(北京)科技股份公司 | Android malicious program detection method based on machine learning |
CN108304720A (en) * | 2018-02-06 | 2018-07-20 | 恒安嘉新(北京)科技股份公司 | A kind of Android malware detection methods based on machine learning |
CN108596678A (en) * | 2018-05-02 | 2018-09-28 | 陈思恩 | A kind of airline passenger value calculation method |
CN109166012A (en) * | 2018-09-21 | 2019-01-08 | 苏州创旅天下信息技术有限公司 | The method and apparatus of classification and information push for stroke predetermined class user |
CN109166012B (en) * | 2018-09-21 | 2021-05-28 | 苏州创旅天下信息技术有限公司 | Method and device for classifying users in travel reservation class and pushing information |
CN109410075A (en) * | 2018-10-23 | 2019-03-01 | 广州市勤思网络科技有限公司 | Intelligence insurance recommended method and system based on Bayes |
CN109492837A (en) * | 2018-12-29 | 2019-03-19 | 携程旅游网络技术(上海)有限公司 | Air ticket order insures method for pushing, device, electronic equipment, storage medium |
CN109493200A (en) * | 2019-01-24 | 2019-03-19 | 深圳市活力天汇科技股份有限公司 | A kind of recommended method of air ticket trip commodity |
CN109493200B (en) * | 2019-01-24 | 2021-08-10 | 深圳市活力天汇科技股份有限公司 | Method for recommending airline ticket travel commodities |
CN109886778A (en) * | 2019-01-29 | 2019-06-14 | 上海华程西南国际旅行社有限公司 | The recommended method and system of the tie-in sale product of air ticket |
CN109934512A (en) * | 2019-03-28 | 2019-06-25 | 努比亚技术有限公司 | A kind of training method and system of prediction model |
CN109934512B (en) * | 2019-03-28 | 2024-02-09 | 努比亚技术有限公司 | Prediction model training method and training system |
CN110288461A (en) * | 2019-05-20 | 2019-09-27 | 深圳壹账通智能科技有限公司 | Automatic verification method, electronic device and the storage medium of Products Show |
CN111144946A (en) * | 2019-12-27 | 2020-05-12 | 上海携程商务有限公司 | Revenue management method, system, medium, and electronic device for airline company |
CN111314869A (en) * | 2020-02-18 | 2020-06-19 | 中国联合网络通信集团有限公司 | Flow quota distribution method and device, electronic equipment and storage medium |
CN111314869B (en) * | 2020-02-18 | 2021-06-29 | 中国联合网络通信集团有限公司 | Flow quota distribution method and device, electronic equipment and storage medium |
WO2022011947A1 (en) * | 2020-10-23 | 2022-01-20 | 平安科技(深圳)有限公司 | Transaction data processing method and apparatus, and computer device and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107240005A (en) | The commending system and method for air ticket addition product | |
US10783457B2 (en) | Method for determining risk preference of user, information recommendation method, and apparatus | |
Zaid et al. | Impact of customer experience and customer engagement on satisfaction and loyalty: A case study in Indonesia | |
CN108960719B (en) | Method and device for selecting products and computer readable storage medium | |
US20190073335A1 (en) | Using artificial intelligence to determine a size fit prediction | |
CN108985830B (en) | Recommendation scoring method and device based on heterogeneous information network | |
WO2017143919A1 (en) | Method and apparatus for establishing data identification model | |
US9330357B1 (en) | Method, apparatus, and computer program product for determining a provider return rate | |
CN105184618A (en) | Commodity individual recommendation method for new users and system | |
CN107169052A (en) | Recommend method and device | |
CN107368936A (en) | Air control model training method and device | |
CN108846724A (en) | Data analysing method and system | |
CN105468628B (en) | A kind of sort method and device | |
US20210241293A1 (en) | Apparatuses, computer-implemented methods, and computer program products for improved model-based determinations | |
CN107516246A (en) | Determination method, determining device, medium and the electronic equipment of user type | |
CN109345050A (en) | A kind of quantization transaction prediction technique, device and equipment | |
CN106991609A (en) | The recommendation method and apparatus of investment product | |
CN106447384A (en) | Method and apparatus for determining object user | |
CN111709813B (en) | Commodity recommendation method based on big data line | |
CN107742217A (en) | Service control method and device based on businessman | |
CN112559900A (en) | Product recommendation method and device, computer equipment and storage medium | |
CN110109902A (en) | A kind of electric business platform recommender system based on integrated learning approach | |
CN111861679A (en) | Commodity recommendation method based on artificial intelligence | |
KR102571651B1 (en) | Customer-specific control units, systems and methods | |
KR20200020264A (en) | System and method for recommending stock using user similarity |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171010 |
|
RJ01 | Rejection of invention patent application after publication |