CN112948412B - Flight inventory updating method, system, electronic device and storage medium - Google Patents
Flight inventory updating method, system, electronic device and storage medium Download PDFInfo
- Publication number
- CN112948412B CN112948412B CN202110429295.4A CN202110429295A CN112948412B CN 112948412 B CN112948412 B CN 112948412B CN 202110429295 A CN202110429295 A CN 202110429295A CN 112948412 B CN112948412 B CN 112948412B
- Authority
- CN
- China
- Prior art keywords
- flight
- user
- cabin
- next moment
- expected
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 230000010006 flight Effects 0.000 claims abstract description 47
- 239000013598 vector Substances 0.000 claims abstract description 18
- 238000006243 chemical reaction Methods 0.000 claims description 15
- 230000008901 benefit Effects 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 8
- 238000013528 artificial neural network Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000001105 regulatory effect Effects 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 10
- BASFCYQUMIYNBI-UHFFFAOYSA-N platinum Chemical compound [Pt] BASFCYQUMIYNBI-UHFFFAOYSA-N 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 230000008485 antagonism Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 229910052697 platinum Inorganic materials 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 1
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 1
- 239000010931 gold Substances 0.000 description 1
- 229910052737 gold Inorganic materials 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/23—Updating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Educational Administration (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of data analysis, and provides a flight inventory updating method, a flight inventory updating system, electronic equipment and a storage medium. The method comprises the following steps: acquiring reservation quantity M of the next moment according to the flight reservation data of the route; sequentially combining the M random vectors with the time characteristics of the next moment, and inputting the combined random vectors into a user generation model to obtain the user characteristics of M expected users at the next moment; the user characteristics of each expected user and the flight characteristics of each flight of the air route are sequentially combined and then input into a user selection model, the probability of booking each flight at the next moment of each expected user is obtained, the expected cabin inventory of each flight at the next moment is further obtained, and a reminder is sent to the browsing user of the target flight with the expected cabin inventory less than the target cabin inventory. According to the invention, the user distribution and the flight reservation probability at the future moment are predicted by combining the flight reservation data, and a prompt is sent to the intention users of the flights with the target cabin being about to be sold out, so that the intention users are ensured to successfully purchase the air ticket, and the user experience is improved.
Description
Technical Field
The invention relates to the technical field of big data analysis, in particular to a flight inventory updating method, a flight inventory updating system, electronic equipment and a storage medium.
Background
The current flight stock updating mode is that when a user successfully subscribes an air ticket, the stock of the corresponding flight is updated. This way of passively updating inventory is prone to situations where the preparing ticket purchaser cannot successfully purchase the ticket.
For example, some users are interrupted by telephone or other abrupt events when they browse the ticket, and when they return to the ticket page, they find that the ticket space in which they are browsing is sold out; for another example, when some users plan a business trip, after browsing the air ticket, they first go to the hotel, and after the hotel reservation is completed, they find that the air ticket is not stored, which affects their normal trip.
It should be noted that the information disclosed in the foregoing background section is only for enhancement of understanding of the background of the invention and thus may include information that does not form the prior art that is already known to those of ordinary skill in the art.
Disclosure of Invention
In view of the above, the invention provides a flight inventory updating method, a system, an electronic device and a storage medium, which can predict user distribution and flight reservation probability at future time by combining flight reservation data, send a reminder for an intended user of a flight with a target cabin being about to be sold out, ensure that the intended user successfully purchases an air ticket, and improve user experience.
One aspect of the present invention provides a flight inventory updating method, including: acquiring a preset quantity M of the next moment of an air route according to the flight reservation data of the air route; sequentially generating M random vectors, combining the M random vectors with the time characteristics of the next moment, and inputting a trained user generation model to obtain the user characteristics of M expected users of the next moment; sequentially combining the user characteristics of each expected user with the flight characteristics of each flight of the air route, and inputting a trained user selection model to obtain the probability of booking each flight at the next moment of each expected user; according to the probability of booking each flight at the next moment of each expected user, obtaining the expected cabin stock of each flight at the next moment, and updating the expected cabin stock into the cabin stock information of the corresponding flight; the reservation reminding information is pushed to a browsing user of a target flight with expected cabin inventory less than target cabin inventory, namely, about to be sold out.
In some embodiments, the user-generated model is trained based on generating an antagonism network, the training process comprising: extracting user characteristics of preset dimensions and time characteristics of ordering time from historical ordering data of all airlines, and constructing a training data set for generating an countermeasure network; inputting a random vector and a time feature to a generator to obtain a group of generated data containing user features and time features; extracting a group of real data containing user characteristics and time characteristics from the training data set, and inputting the generated data and the real data into a discriminator to obtain a judging result; according to the judging result, parameter tuning is carried out on the generator or the discriminator until the generator generates generated data which enables the discriminator to be incapable of judging the data source; and taking the trained generator as the user generation model.
In some embodiments, the training process, the loss function is:
Loss＝E x～pr (D(x,t))-E x～pG (D(x,t))；
wherein x is a user characteristic, t is a time characteristic, x-pr are data distribution of x meeting a training data set, and x-pG are data distribution of x meeting generated data; d (x, t) is the judgment result of the discriminator, E x～pr (D (x, t)) is the discriminationThe device judges the data source as the mathematical expectation of the training data set, E x～pG (D (x, t)) is a mathematical expectation that the arbiter determines the source of the data as the generated data.
In some embodiments, the preset dimensions include: age, business type, gender, and user rating.
In some embodiments, the user-selected model is an XGBoost model, which is an extensible machine learning model based on a lifting tree, and the results of a plurality of decision trees are summed to form a final predicted value; the input of the user selection model further comprises: the current bunk characteristics of each flight and the market characteristics of the airlines.
In some embodiments, the flight inventory updating method further includes: obtaining a cabin-regulating flight, and determining a plurality of groups of candidate cabin characteristics of the cabin-regulating flight; predicting the bilge benefit of the cabin-regulating flight based on each group of alternative bilge features in sequence, including: inputting the combination of the flight characteristics and a group of alternative cabin characteristics of the cabin-regulating flights, the flight characteristics and the current cabin characteristics of the rest flights of the air route and the market characteristics of the air route into the XGBoost model to obtain the selection probability of the cabin-regulating flights; according to the alternative bunk characteristics and the selection probability, obtaining bunk benefits of the cabin-regulating flight based on the alternative bunk characteristics; taking the alternative bunk characteristic corresponding to the maximum bunk income as the target bunk characteristic of the cabin-regulating flight at the next moment; and pushing reservation reminding information to a browsing user of the target bunk with the bunk inventory less than the threshold value at the next moment according to the target bunk characteristics.
In some embodiments, the flight inventory updating method further includes: according to the target cabin characteristics of the cabin-regulating flight, the current cabin characteristics of the associated flight are regulated through a trained associated reaction model; the association reaction model is constructed based on a neural network and is used for predicting whether the rest flights are accommodated at the next moment and updating the cabin characteristics of the rest flights at the next moment, and the loss function of the association reaction model is as follows:
FL(y)＝-α(1-y t ) γ logy t ；
Loss＝FL(y)+β*MSE(p next -p nextpred )；
wherein y is t For the other flights, the duty ratio of the cabin-regulating sample at the next moment, alpha and gamma are super parameters, t is the next moment, and p next For the actual bunk characteristics of the bunk adjustment sample at the next moment, p nextpred For the updated bunk characteristics of the next moment bunk sample, it is desirable to have the constraint that the closer the updated bunk characteristics of the next moment bunk sample are to the actual bunk characteristics.
In some embodiments, the obtaining the expected bunk inventory for each flight at the next time comprises: acquiring a flight deck corresponding to the maximum probability of each expected user as a reserved flight deck of each expected user; counting the number of each reserved flight space to obtain the total reserved quantity of each reserved flight space at the next moment; taking the difference between the current stock of each flight deck and the corresponding total booking quantity as the expected deck stock of each flight deck at the next moment.
In some embodiments, the flight reservation data is reservation data for flights in the airline having the same departure time; and before the same departure time, circularly executing the flight inventory updating method by taking a preset period as a time interval.
Another aspect of the present invention provides a flight inventory updating system, comprising: the reservation amount acquisition module is used for acquiring reservation amount M of the next moment of an air route according to the flight reservation data of the air route; the expected user generation module is used for sequentially generating M random vectors, inputting a trained user generation model after being combined with the time characteristics of the next moment, and obtaining the user characteristics of M expected users of the next moment; the booking probability calculation module is used for sequentially combining the user characteristics of each expected user and the flight characteristics of each flight of the air route and inputting the combined user characteristics into a trained user selection model to obtain the probability of booking each flight at the next moment of each expected user; the expected stock updating module is used for acquiring expected cabin stock of each flight at the next moment according to the probability of each expected user booking each flight at the next moment and updating the expected cabin stock into cabin stock information of the corresponding flight; and the flight reservation pushing module is used for pushing reservation reminding information to a browsing user of a target flight with the expected berth inventory less than the target berth inventory.
Yet another aspect of the present invention provides an electronic device, comprising: a processor; a memory having executable instructions stored therein; when the executable instructions are executed by the processor, the flight inventory updating method described in any embodiment is implemented.
A further aspect of the present invention provides a computer-readable storage medium storing a program which, when executed, implements the flight inventory updating method according to any of the above embodiments.
Compared with the prior art, the invention has the beneficial effects that at least:
the invention combines the flight reservation data to make predictions on the distribution of users at future time and the probability of flight reservation, and helps users to predict the ticket selling speed so as to decide to purchase; particularly for flights with the target berths about to be sold out, a prompt can be sent to the intended user of the target berths browsing the flights, so that the intended user can be ensured to successfully purchase the air ticket, and the user experience is improved; and for ticket platforms and air terminals, the air ticket selling can be promoted, and the operation cost is reduced.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention as claimed.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention. It is evident that the figures described below are only some embodiments of the invention, from which other figures can be obtained without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram showing steps of a flight inventory updating method according to an embodiment of the present invention;
FIG. 2 illustrates a training scenario diagram of generating a countermeasure network in an embodiment of the invention;
FIG. 3 is a schematic diagram showing steps for cabin-level adjustment of a cabin-adjusting flight in an embodiment of the present invention;
FIG. 4 is a schematic structural diagram of an associative reaction model in an embodiment of the present invention;
FIG. 5 shows a schematic diagram of a simulation scenario of a simulator in an embodiment of the invention;
FIG. 6 is a schematic block diagram of a flight inventory update system according to an embodiment of the invention;
fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present invention;
fig. 8 shows a schematic structure of a computer-readable storage medium in an embodiment of the present invention.
Detailed Description
Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the example embodiments to those skilled in the art.
The drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus a repetitive description thereof will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software or in one or more hardware modules or integrated circuits or in different networks and/or processor devices and/or microcontroller devices.
Furthermore, the flow shown in the drawings is merely illustrative and not necessarily all steps are included. For example, some steps may be decomposed, some steps may be combined or partially combined, and the order of actual execution may be changed according to actual situations. It should be noted that, without conflict, the embodiments of the present invention and features in different embodiments may be combined with each other.
Fig. 1 shows main steps of a flight inventory updating method in an embodiment, and referring to fig. 1, the flight inventory updating method in this embodiment includes: in step S110, a reservation amount M of a next time of an airline is obtained according to flight reservation data of the airline; in step S120, M random vectors are sequentially generated, and after being combined with the time characteristics of the next moment, the M random vectors are input into a trained user generation model to obtain user characteristics of M expected users of the next moment; in step S130, the user characteristics of each expected user and the flight characteristics of each flight of the route are sequentially combined and then input into a trained user selection model, so as to obtain the probability of booking each flight at the next moment of each expected user; in step S140, according to the probability of each expected user booking each flight at the next moment, obtaining the expected cabin inventory of each flight at the next moment, and updating the cabin inventory information of the corresponding flight; in step S150, reservation reminding information is pushed to a browsing user of a target flight with a desired bunk inventory less than the target bunk inventory.
The flight inventory updating method combines the flight reservation data to make predictions on the user distribution and the flight reservation probability at the future time, and helps the user to predict the ticket selling speed so as to decide to purchase; particularly for flights with the target berths about to be sold out, a prompt can be sent to the intended user of the target berths browsing the flights, so that the intended user can be ensured to successfully purchase the air ticket, and the user experience is improved; and for ticket platforms and air terminals, the air ticket selling can be promoted, and the operation cost is reduced.
The following describes the step flow of the flight inventory updating method in detail with reference to specific examples.
In step S110, the flight reservation data is specifically reservation data of flights with the same departure time in the route to the current time, and the reservation data shows the reservation condition of each flight for each bunk. The next time specifically refers to a period of time corresponding to the next execution cycle. That is, in this embodiment, the flight inventory updating method is circularly executed at intervals of a preset period before a preset departure time. For example, the preset period is 1 hour, and before the departure time, the inventory of the corresponding flights in the air line is actively updated once every 1 hour through a flight inventory updating method.
The reservation amount M of the next moment of the route can be obtained according to the change rule of the flight reservation data of the historical preset period of the route; existing methods, such as prediction by passenger source analysis, may also be employed, as the present invention is not limited in this regard; alternatively, the reservation amount M at the next time may be set in combination with the flight reservation data and the target sales data of the route.
In step S120, the user generated model is trained based on generating the countermeasure network, and the training process includes: extracting user characteristics of preset dimensions and time characteristics of ordering time from historical ordering data of all airlines, and constructing a training data set for generating an countermeasure network; inputting a random vector and a time feature to a generator to obtain a group of generated data containing user features and time features; extracting a group of real data containing user characteristics and time characteristics from the training data set, and inputting the generated data and the real data into a discriminator to obtain a judging result; according to the judging result, parameter tuning is carried out on the generator or the discriminator until the generator generates the generated data which enables the discriminator to be incapable of judging the data source; and taking the trained generator as a user generation model.
When the training data set is constructed, the user characteristics are marked as x, and attribute characteristics of dimensions such as age stage, business type, gender, user grade and the like of the ordering user are extracted from the historical ordering data. The linear correlation between age and ordering behavior is weak, so that the age and ordering behavior are divided into six stages of 0-2 years old, 2-12 years old, 12-18 years old, 18-23 years old, 33-55 years old and over 55 years old according to the purchasing power of different age stages; the business travel types are specifically classified into common, business travel and unknown, and are respectively encoded into 0, 1 and unow; sex characteristics are discretized into men and women, unknown conditions are eliminated, and codes are respectively 1 and 0; user ratings are classified into normal, gold, platinum, and masonry users. Because the user characteristics have strong correlation with the ordering time, the user characteristics can change along with the ordering time, so that the corresponding time characteristics of the ordering time are added at the same time when the user characteristics are constructed. Taking user i as an example, its user characteristic x based on time characteristic t ti Expressed as: x is x ti ＝(age ti ,business ti ,sex ti ,usertype ti T). Wherein, age ti Is the age group characteristic of user i based on time characteristic t, bussiness ti Sex for business trip type feature of user i based on time feature t ti User type for gender feature of user i based on time feature t ti Is a user type feature of user i based on the time feature t.
Taking the time feature t as an example of the ordering date, assume that the platinum user i1 successfully subscribes to an air ticket in the ordering date 2020-09-23 and is a 35 year old business trip client with gender as female, and the user feature x of the user i1 based on the time feature t ti1 Expressed as: x is x ti1 = (4,1,0,2,20200923). Since the machine cannot directly identify the discrete features and the date features, the discrete variables are encoded in a one-hot coded form and the date features are converted into usable information. For example, the date feature may be turned into a feature of week, spring festival, national celebration, holiday, aviation season, advance reservation date, etc. Thus, a machine-recognizable user data sample X can be constructed u I.e. a training data set. In this embodiment, a billion-level user data sample X is constructed by using massive historical downlink data of the ticketing platform u ＝concat(x t T). Through the constructed training data set, the antagonism network can be trained and generated, and the user characteristic distribution at the future moment can be learned, so that the expected user at the future moment can be simulated.
When training to generate an countermeasure network, referring to the training scenario shown in fig. 2, the generated countermeasure network is composed of two networks, namely, a generator G and a discriminator D. The input data 210 of the generator G is a random number and a time feature, and the generated data 220 is a user feature based on the time feature; the generated data 220 of the generator G and the real data 240 extracted from the training data set 230, including user characteristics and time characteristics, are input to the arbiter D, which resolves the data sources. When the judging device D judges that the generated data 220 is generated by the generator G, the parameters of the generator G are adjusted; when the arbiter D judges an error, that is, the generated data 220 is not resolved to be generated by the generator G, the parameters of the arbiter D are adjusted. After several rounds of iterative training, until the generator G generates the generated data 220 that makes the arbiter D unable to determine the source of the data, the generator G accurately learns the user feature distribution based on the time feature in the training data set 230.
In the training process, by means of the thought of WGAN, the loss function uses the bulldozing distance, and the specific formula is as follows:
Loss＝E x～pr (D(x,t))-E x～pG (D(x,t))；
wherein x is a user characteristic, t is a time characteristic, x-pr are data distribution of x meeting a training data set, and x-pG are data distribution of x meeting generated data; d (x, t) is the judgment result of the discriminator, E x～pr (D (x, t)) is a mathematical expectation that the arbiter determines that the data source is a training data set, E x～pG (D (x, t)) is a mathematical expectation that the arbiter determines the source of the data as the generated data. The optimization is continually countered by the generator and the arbiter until a Nash equilibrium state is reached.
In one embodiment, predictions of expected user distribution for a particular route are made for different lead times by a trained user generated model, starting at 40 days in advance and ending at the lead date, with daily generated samples of the user generated model being close to the real samples of the route.
Thus, using the trained user generation model, in step S120, each time according to the input random vector and the combination feature of the next moment, the user feature of an expected user based on the next moment can be output; through the combined input of the M random vectors and the time characteristics of the next moment, the user characteristics of M expected users based on the next moment can be obtained, and the real user distribution of the next moment can be simulated.
In step S130, the user selects the model to be the XGBoost model. The flight characteristics include the flight take-off and landing time, take-off and landing place, airline and other attribute characteristics. In one embodiment, the input of the user selection model also includes, considering that factors affecting the user selection of a flight also include the individual bilge prices of the flights, the condition of other flights on the same airline, and so on: the current bunk characteristics of each flight, in particular the bunk type and the price of each bunk type contained by each flight, and the market characteristics of the airlines. Thus, each time data of the user selection model is input, it includes: a user profile of the intended user, a flight table comprising attribute characteristics of each flight such as each bunk price and departure location/time, landing location/time, and airline profile. According to the input data, the user selection model can accurately predict the booking probability of the expected user on each flight deck in the flight table. Probability prediction based on XGBoost model is a mature technique and therefore will not be explained.
In the training process of the user selection model, for each sample user, flights selected by the sample user for ordering in the flight table (specifically, a certain flight position of a certain flight) are marked as 1, and other flight positions are marked as 0. The XGBoost model adds regularization term in the objective function, controls the complexity of the model, is beneficial to preventing over fitting and improves the generalization capability of the model; and the gradient is faster by using the second-order Taylor expansion, and the gradient has strong classifying capability, so that the training classification is selected to predict the booking probability of each sample user to each flight in the flight table. Finally, the flight with the highest booking probability is selected as the final choice of the sample user.
In step S140, the process of obtaining the expected inventory of each flight at the next moment specifically includes: acquiring a flight deck corresponding to the maximum probability of each expected user as a reserved flight deck of each expected user; counting the number of each reserved flight space to obtain the total reserved quantity of each reserved flight space at the next moment; taking the difference between the current stock of each flight deck and the corresponding total booking quantity as the expected deck stock of each flight deck at the next moment.
In step S150, the target flights with expected berth inventory less than the target berth inventory are screened out, and reservation reminding information is pushed to the browsing user, so that the intention user who browses the target flights in history can timely learn the information that the target berth of the target flights is about to be sold out, and the user can be assisted in making a purchase decision. In the booking reminding information, a purchase link of a bill page of a target bunk of the target flight can be additionally jumped to facilitate the user with the intention of purchasing demand to purchase in time.
In one embodiment, the flight inventory update method is further capable of dynamically adjusting the flight slots based on the predicted probability of flight reservation at the future time. The adjustment of the flight berth can be performed by adopting a data set of a user selection model, removing user characteristics from the data set, inputting the data set into the XGBoost model to obtain the selection probability of each flight berth, and in the step, the XGBoost model is used as a berth adjustment model. Alternatively, the output results of the user selection model may be weighted to obtain the overall selection probability of each flight deck.
Fig. 3 illustrates a procedure of cabin adjustment for a cabin-exchanging flight in an embodiment, and referring to fig. 3, the flight inventory updating method further includes: step S310, a cabin-regulating flight is obtained, and multiple groups of candidate cabin characteristics of the cabin-regulating flight are determined. Step S320, predicting a cabin profit of the cabin-regulating flight based on each group of candidate cabin characteristics in turn, including: the method comprises the steps of inputting a combination of flight characteristics of a cabin-regulating flight, a group of alternative cabin characteristics, flight characteristics of other flights of a route, current cabin characteristics and market characteristics (such as route heat) of the route into an XGBoost model to obtain selection probability of the cabin-regulating flight; and obtaining the cabin benefits of the cabin regulating flight based on the alternative cabin characteristics (the benefits are cabin price and cabin selection probability) according to the alternative cabin characteristics and the selection probability. Step S330, taking the alternative berth characteristic corresponding to the maximum berth gain as a target berth characteristic of the next-moment cabin-regulating flight; the method comprises the steps of adjusting the bunk characteristics of the cabin-regulating flight for a plurality of times, calculating expected benefits of the cabin-regulating flight based on the bunk characteristics after adjustment each time, and taking the bunk price with the maximum expected benefits as the updated bunk characteristics of the cabin-regulating flight at the next moment. Step S340, pushing reservation reminding information to a browsing user of a target bunk with the bunk inventory less than a threshold value at the next moment according to the target bunk characteristics; for example, after the price of the bunk is adjusted, if the inventory of the lowest price bunk is insufficient, reservation reminding information is pushed to a browsing user of the lowest price bunk.
When the flight deck is adjusted, the selling conditions of the cabin-adjusting flights and the airlines can be used as punishment items, and if the selling speed is too high, the punishment items are biased to be added; if the vending speed is too slow, the penalty term is biased towards a reduction in price.
Further, after the cabin characteristics of the cabin-regulating flights are updated, cabin regulation can be carried out on the associated flights under the same airlines, cabin response prices of the associated flights are predicted, and the whole market environment of the airlines is updated. This step may be accomplished by a correlation reaction model. Firstly, judging whether the state of the cabin-regulating flight changes, if the cabin characteristics and the inventory of the cabin-regulating flight are unchanged, other related flights keep the original cabin price, so that the running time of a model can be effectively reduced, and the system efficiency is improved. If the market environment, including the status of any flights under the airline, changes, then the correlation reaction model is run.
In the training process of the association reaction model, the whole of the flight samples needing cabin adjustment is small, and the whole of the flight samples generally only accounts for about 14% of all flight samples under the air route; if the regression algorithm is directly adopted to predict the cabin response price of the associated flight, the training result tends to deviate from the original price, so that the model is invalid; since neural networks can theoretically fit all complex functions, neural networks are employed to build models. In order to solve the problem of sample unbalance, the next time cabin-regulating sample is coded as 1, and the next time cabin-regulating sample is not coded as 0. The optimization target of the correlation reaction model is converted into two tasks, namely, the problem of classification of whether to tune the cabin is fitted, and the regression task of tuning the cabin at the next moment is optimized.
Referring to the principle of the correlation reaction model shown in fig. 4, the samples are classified into discrete variables and continuous variables, the continuous variables are standardized, the discrete variables are transmitted into the embedded layer 410 for dimension reduction after One-Hot encoding, and the dimension reduction can be set according to the needs. After the continuous variable and the discrete variable are processed by the preprocessing layer 420, the continuous variable and the discrete variable enter a prediction network 430 constructed by two neural network layers (or more complex structures) for training, and fit whether to tune the cabin label and the cabin tuning feature at the next moment. Since the positive samples, i.e. the pod samples, have too small a duty cycle, direct training can lead to information being submerged in many nonsensical samples, this step uses focallos to solve this problem:
FL(y)＝-α(1-y t ) γ logy t ；
wherein y is t Is positive toThe duty ratio of the sample, namely the duty ratio of the cabin-regulating sample in the rest flights at the next moment; alpha and gamma are hyper-parameters. The final optimized loss function of the network can be obtained by combining the two tasks:
Loss＝FL(y)+β*MSE(p next -p nextpred )；
wherein t is the next time, p next For the actual bunk characteristics of the bunk adjustment sample at the next moment, p nextpred And updating the cabin level characteristics of the cabin adjusting sample at the next moment.
Further, in one embodiment, a simulator is constructed by combining the user-generated model, the user-selected model, the cabin level adjustment model and the associated reaction model of each embodiment, and the model is a complete air ticket vending process. The method specifically comprises the following steps: initial configuration parameters, setting take-off dates, airlines, cabin-regulating flight numbers, reservation times and reservation amounts (which can be obtained through a passenger source prediction model or set according to requirements); s510, inputting a reservation amount and a reservation time, and generating a user sample by using a user generation model; s520, the user sample is associated with information such as flight market price and the like to form a new sample, the new sample is input into a user selection model, and flight reservation is generated through selection probability, so that the stock of each flight in the airline is updated; s530, calculating the optimal cabin position of the cabin-regulating flight through a cabin position adjustment model, and updating the cabin-regulating flight position; s540, adjusting reaction bilges of other associated flights in the airlines through the associated reaction model; completing a round of simulation; the reservation time and the reservation amount are then updated, and S510-S540 are repeatedly performed until the set departure date is reached. And finally, calculating the benefit of the overall simulation for the cabin adjustment reference of the airlines.
The cabin adjustment strategy can be automatically judged by simulating the ticket selling flow. Compared with the traditional manual cabin adjustment by means of personal experience, the influence of the cabin adjustment on the subsequent benefits cannot be predicted, the simulator can predict the future cabin adjustment benefits of the results according to the manual cabin adjustment results, and the optimal cabin adjustment operation is given for manual decision, so that the cabin adjustment effectiveness and certainty are effectively improved. Meanwhile, the simulator is different from the traditional aviation income management system, and can fully utilize the data advantage of the ticketing platform to learn the distribution hidden in the data. From the experimental effect, the user characteristic distribution generated by the user generated model is basically consistent with the actual distribution, and the KL divergence of each characteristic is lower than 0.003. The average absolute percentage error (MAPE) of the order quantity of flights after take-off is 45.11+/-0.17%, which is far smaller than the lowest berth taking method error (141.13%) and the uniform sampling error (60.88+/-0.41%).
The embodiment of the invention also provides a flight stock updating system which can be used for realizing the flight stock updating method described in any embodiment. The features and principles of the flight inventory update method described in any of the above embodiments are applicable to the following flight inventory update system embodiments. In the following embodiments of the flight inventory update system, the features and principles already explained with respect to the flight inventory update will not be repeated.
Fig. 6 shows main modules of the flight inventory updating system in the embodiment, and referring to fig. 6, the flight inventory updating system 600 in the embodiment includes: a reservation amount obtaining module 610, configured to obtain a reservation amount M of a next time of an airline according to flight reservation data of the airline; the expected user generation module 620 is configured to sequentially generate M random vectors, combine the random vectors with time features of a next moment, and input a trained user generation model to obtain user features of M expected users of the next moment; the booking probability calculation module 630 is configured to sequentially combine the user characteristics of each prospective user and the flight characteristics of each flight of the airline, and then input the combined user characteristics into a trained user selection model to obtain the probability of booking each flight at the next moment of each prospective user; the expected stock updating module 640 is configured to obtain an expected cabin stock of each flight at the next moment according to the probability that each expected user subscribes to each flight at the next moment, and update the expected cabin stock to cabin stock information of the corresponding flight; the flight reservation pushing module 650 is configured to push reservation reminding information to a browsing user of a target flight whose expected bunk inventory is less than the target bunk inventory.
Further, the flight inventory updating system 600 may further include modules that implement other flow steps of the embodiments of the above-described flight inventory updating method, and the specific principles of each module may refer to the descriptions of the embodiments of the above-described flight inventory updating method, which are not repeated herein.
As described above, the flight inventory updating system of the invention can combine the flight reservation data to make predictions on the user distribution and the flight reservation probability at the future time, and help the user to predict the ticket selling speed so as to make a decision for purchase; particularly for flights with the target berths about to be sold out, a prompt can be sent to the intended user of the target berths browsing the flights, so that the intended user can be ensured to successfully purchase the air ticket, and the user experience is improved; and for ticket platforms and air terminals, the air ticket selling can be promoted, and the operation cost is reduced.
The embodiment of the invention also provides electronic equipment, which comprises a processor and a memory, wherein executable instructions are stored in the memory, and when the executable instructions are executed by the processor, the flight inventory updating method described in any embodiment is realized.
As described above, the electronic device of the invention can combine the flight reservation data to make predictions on the user distribution and the flight reservation probability at the future time, and help the user to predict the ticket selling speed so as to make a decision for purchase; particularly for flights with the target berths about to be sold out, a prompt can be sent to the intended user of the target berths browsing the flights, so that the intended user can be ensured to successfully purchase the air ticket, and the user experience is improved; and for ticket platforms and air terminals, the air ticket selling can be promoted, and the operation cost is reduced.
Fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and it should be understood that fig. 7 is only a schematic diagram illustrating each module, and the modules may be virtual software modules or actual hardware modules, and the combination, splitting and addition of the remaining modules are all within the scope of the present invention.
As shown in fig. 7, the electronic device 700 is embodied in the form of a general purpose computing device. Components of electronic device 700 include, but are not limited to: at least one processing unit 710, at least one memory unit 720, a bus 730 connecting the different platform components (including memory unit 720 and processing unit 710), a display unit 740, and the like.
The storage unit stores therein a program code executable by the processing unit 710 to cause the processing unit 710 to perform the steps of the flight inventory updating method described in any of the above embodiments. For example, the processing unit 710 may perform the steps as shown in fig. 1.
The memory unit 720 may include readable media in the form of volatile memory units, such as Random Access Memory (RAM) 7201 and/or cache memory 7202, and may further include Read Only Memory (ROM) 7203.
The storage unit 720 may also include a program/utility 7204 having one or more program modules 7205, such program modules 7205 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment.
Bus 730 may be a bus representing one or more of several types of bus structures including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.
The electronic device 700 may also communicate with one or more external devices 800, which external devices 800 may be one or more of a keyboard, pointing device, bluetooth device, etc. These external devices 800 enable a user to interactively communicate with the electronic device 700. The electronic device 700 can also communicate with one or more other computing devices, including a router, modem, as shown. Such communication may occur through an input/output (I/O) interface 750. Also, electronic device 700 may communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN) and/or a public network, such as the Internet, through network adapter 760. Network adapter 760 may communicate with other modules of electronic device 700 via bus 730. It should be appreciated that although not shown, other hardware and/or software modules may be used in connection with electronic device 700, including, but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, data backup storage platforms, and the like.
The embodiment of the invention also provides a computer readable storage medium for storing a program, which when executed, implements the flight inventory updating method described in any of the above embodiments. In some possible implementations, aspects of the invention may also be implemented in the form of a program product comprising program code for causing a terminal device to carry out the flight inventory updating method described in any of the above-described embodiments, when the program product is run on the terminal device.
As described above, the computer-readable storage medium of the present invention is capable of making predictions of user distribution and flight reservation probability at future times in combination with flight reservation data, helping users to predict ticket vending speed to make decisions for purchase; particularly for flights with the target berths about to be sold out, a prompt can be sent to the intended user of the target berths browsing the flights, so that the intended user can be ensured to successfully purchase the air ticket, and the user experience is improved; and for ticket platforms and air terminals, the air ticket selling can be promoted, and the operation cost is reduced.
Fig. 8 is a schematic structural view of a computer-readable storage medium of the present invention. Referring to fig. 8, a program product 900 for implementing the above-described method according to an embodiment of the present invention is described, which may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited thereto, and in this document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the readable storage medium include, but are not limited to: an electrical connection having one or more wires, a portable disk, a hard disk, random Access Memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The computer readable storage medium may include a data signal propagated in baseband or as part of a carrier wave, with readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A readable storage medium may also be any readable medium that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Program code for carrying out operations of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device, partly on a remote computing device, or entirely on the remote computing device or server. In the case of remote computing devices, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device, such as through the Internet using an Internet service provider.
The foregoing is a further detailed description of the invention in connection with the preferred embodiments, and it is not intended that the invention be limited to the specific embodiments described. It will be apparent to those skilled in the art that several simple deductions or substitutions may be made without departing from the spirit of the invention, and these should be considered to be within the scope of the invention.
Claims (9)
1. A method for updating a flight inventory, comprising:
acquiring reservation quantity M of a next moment of an air route according to flight reservation data of the air route;
sequentially generating M random vectors, combining the M random vectors with the time characteristics of the next moment, and inputting a trained user generation model to obtain the user characteristics of M expected users of the next moment;
wherein the user generated model is trained based on generating an countermeasure network, the training process comprising: extracting user characteristics of preset dimensions and time characteristics of ordering time from historical ordering data of all airlines, and constructing a training data set for generating an countermeasure network; inputting a random vector and a time feature to a generator to obtain a group of generated data containing user features and time features; extracting a group of real data containing user characteristics and time characteristics from the training data set, and inputting the generated data and the real data into a discriminator to obtain a judging result; according to the judging result, parameter tuning is carried out on the generator or the discriminator until the generator generates generated data which enables the discriminator to be incapable of judging the data source; taking the trained generator as the user generation model;
Sequentially combining the user characteristics of each expected user with the flight characteristics of each flight of the air route, and inputting a trained user selection model to obtain the probability of booking each flight at the next moment of each expected user;
according to the probability of booking each flight at the next moment of each expected user, obtaining the expected cabin stock of each flight at the next moment, and updating the expected cabin stock into cabin stock placement information of the corresponding flight;
pushing reservation reminding information for browsing users of target flights with expected berth inventory less than target berth inventory;
the user selection model is an XGBoost model, the input of the user selection model further comprises the current cabin characteristics of each flight and the market characteristics of the airlines, and the flight inventory updating method further comprises the following steps:
obtaining a cabin-regulating flight, and determining a plurality of groups of candidate cabin characteristics of the cabin-regulating flight;
predicting the bilge benefit of the cabin-regulating flight based on each group of alternative bilge features in sequence, including: inputting the combination of the flight characteristics and a group of alternative cabin characteristics of the cabin-regulating flights, the flight characteristics and the current cabin characteristics of the rest flights of the air route and the market characteristics of the air route into the XGBoost model to obtain the selection probability of the cabin-regulating flights; according to the alternative bunk characteristics and the selection probability, obtaining bunk benefits of the cabin-regulating flight based on the alternative bunk characteristics;
Taking the alternative bunk characteristic corresponding to the maximum bunk income as the target bunk characteristic of the cabin-regulating flight at the next moment;
and pushing reservation reminding information to a browsing user of the target bunk with the bunk inventory less than the threshold value at the next moment according to the target bunk characteristics.
2. The flight inventory updating method according to claim 1, wherein in the training process, the loss function is:
Loss＝E x～pr (D(x,t))-E x～pG (D(x,t))；
wherein x is a user characteristic, t is a time characteristic, x-pr are data distribution of x meeting a training data set, and x-pG are data distribution of x meeting generated data; d (x, t) is the judgment result of the discriminator, E x～pr (D (x, t)) is a mathematical expectation that the arbiter determines that the data source is a training data set, E x～pG (D (x, t)) is a mathematical expectation that the arbiter determines the source of the data as the generated data.
3. The flight inventory updating method according to claim 1, wherein the preset dimensions include: age, business type, gender, and user rating.
4. The flight inventory updating method according to claim 1, further comprising:
according to the target cabin characteristics of the cabin-regulating flight, the current cabin characteristics of the associated flight are regulated through a trained associated reaction model;
The association reaction model is constructed based on a neural network and is used for predicting whether the rest flights are accommodated at the next moment and updating the cabin characteristics of the rest flights at the next moment, and the loss function of the association reaction model is as follows:
FL(y)＝-α(1-y t ) γ logy t ；
Loss＝FL(y)+β*MSE(p next -p nextpred )；
wherein y is t For the other flights, the duty ratio of the cabin-regulating sample at the next moment, alpha and gamma are super parameters, t is the next moment, and p next For the actual bunk characteristics of the bunk adjustment sample at the next moment, p nextpred And updating the cabin level characteristics of the cabin adjusting sample at the next moment.
5. The method for updating a flight inventory of claim 1, wherein obtaining an expected bunk inventory for each flight at a next time comprises:
acquiring a flight deck corresponding to the maximum probability of each expected user as a reserved flight deck of each expected user;
counting the number of each reserved flight space to obtain the total reserved quantity of each reserved flight space at the next moment;
taking the difference between the current stock of each flight deck and the corresponding total booking quantity as the expected deck stock of each flight deck at the next moment.
6. A flight inventory updating method according to any one of claims 1 to 5, wherein the flight reservation data is reservation data of flights having the same departure time in the route;
And before the same departure time, circularly executing the flight inventory updating method by taking a preset period as a time interval.
7. A flight inventory update system, comprising:
the reservation amount acquisition module is used for acquiring reservation amount M of the next moment of an air route according to the flight reservation data of the air route;
the expected user generation module is used for sequentially generating M random vectors, inputting a trained user generation model after being combined with the time characteristics of the next moment, and obtaining the user characteristics of M expected users of the next moment;
wherein the user generated model is trained based on generating an countermeasure network, the training process comprising: extracting user characteristics of preset dimensions and time characteristics of ordering time from historical ordering data of all airlines, and constructing a training data set for generating an countermeasure network; inputting a random vector and a time feature to a generator to obtain a group of generated data containing user features and time features; extracting a group of real data containing user characteristics and time characteristics from the training data set, and inputting the generated data and the real data into a discriminator to obtain a judging result; according to the judging result, parameter tuning is carried out on the generator or the discriminator until the generator generates generated data which enables the discriminator to be incapable of judging the data source; taking the trained generator as the user generation model;
The booking probability calculation module is used for sequentially combining the user characteristics of each expected user and the flight characteristics of each flight of the air route and inputting the combined user characteristics into a trained user selection model to obtain the probability of booking each flight at the next moment of each expected user;
the expected stock updating module is used for acquiring expected cabin stock of each flight at the next moment according to the probability of each expected user booking each flight at the next moment and updating the expected cabin stock into cabin stock information of the corresponding flight;
the flight reservation pushing module is used for pushing reservation reminding information to a browsing user of a target flight with expected berth inventory less than the target berth inventory;
the user selection model is an XGBoost model, the input of the user selection model further comprises the current cabin characteristics of each flight and the market characteristics of the airlines, and the flight inventory updating system further comprises a functional module for realizing the following steps:
obtaining a cabin-regulating flight, and determining a plurality of groups of candidate cabin characteristics of the cabin-regulating flight;
predicting the bilge benefit of the cabin-regulating flight based on each group of alternative bilge features in sequence, including: inputting the combination of the flight characteristics and a group of alternative cabin characteristics of the cabin-regulating flights, the flight characteristics and the current cabin characteristics of the rest flights of the air route and the market characteristics of the air route into the XGBoost model to obtain the selection probability of the cabin-regulating flights; according to the alternative bunk characteristics and the selection probability, obtaining bunk benefits of the cabin-regulating flight based on the alternative bunk characteristics;
Taking the alternative bunk characteristic corresponding to the maximum bunk income as the target bunk characteristic of the cabin-regulating flight at the next moment;
and pushing reservation reminding information to a browsing user of the target bunk with the bunk inventory less than the threshold value at the next moment according to the target bunk characteristics.
8. An electronic device, comprising:
a processor;
a memory having executable instructions stored therein;
wherein the executable instructions, when executed by the processor, implement the flight inventory updating method of any one of claims 1-6.
9. A computer-readable storage medium storing a program, wherein the program when executed implements the flight inventory updating method of any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110429295.4A CN112948412B (en) | 2021-04-21 | 2021-04-21 | Flight inventory updating method, system, electronic device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110429295.4A CN112948412B (en) | 2021-04-21 | 2021-04-21 | Flight inventory updating method, system, electronic device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112948412A CN112948412A (en) | 2021-06-11 |
| CN112948412B true CN112948412B (en) | 2024-03-12 |
Family
ID=76233165
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110429295.4A Active CN112948412B (en) | 2021-04-21 | 2021-04-21 | Flight inventory updating method, system, electronic device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112948412B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113486084B (en) * | 2021-06-30 | 2024-02-09 | 中国民航信息网络股份有限公司 | Lock cabin control method and device, electronic equipment and computer storage medium |
| CN113627999A (en) * | 2021-06-30 | 2021-11-09 | 中国铁道科学研究院集团有限公司电子计算技术研究所 | Combined transportation and ticketing method, combined transportation and ticketing system, ticketing system and electronic equipment |
| CN114330908A (en) * | 2021-12-31 | 2022-04-12 | 中国民航信息网络股份有限公司 | Seat booking demand prediction method and device and revenue management system |
| CN116596170B (en) * | 2023-07-18 | 2023-09-22 | 合肥城市云数据中心股份有限公司 | Intelligent prediction method for delivery time based on space-time attention mechanism |
| CN117436676B (en) * | 2023-12-21 | 2024-06-11 | 深圳市活力天汇科技股份有限公司 | Weight assignment method and system for flight data validity |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108846493A (en) * | 2018-06-22 | 2018-11-20 | 山东大学 | A kind of air ticket booking number prediction technique based on improved incremental model |
| CN109952572A (en) * | 2016-09-20 | 2019-06-28 | 谷歌有限责任公司 | Suggestion response based on message paster |
| CN111858690A (en) * | 2020-07-30 | 2020-10-30 | 中国民航信息网络股份有限公司 | Automatic adjustment method and system for limited sales number of flight space |
| CN112329816A (en) * | 2020-10-09 | 2021-02-05 | 北京嘀嘀无限科技发展有限公司 | Data classification method and device, electronic equipment and readable storage medium |
| CN112651671A (en) * | 2021-01-18 | 2021-04-13 | 中国民航信息网络股份有限公司 | Flight space adjusting method and related equipment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US12007980B2 (en) * | 2019-01-17 | 2024-06-11 | The Boston Consulting Group, Inc. | AI-driven transaction management system |
-
2021
- 2021-04-21 CN CN202110429295.4A patent/CN112948412B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109952572A (en) * | 2016-09-20 | 2019-06-28 | 谷歌有限责任公司 | Suggestion response based on message paster |
| CN108846493A (en) * | 2018-06-22 | 2018-11-20 | 山东大学 | A kind of air ticket booking number prediction technique based on improved incremental model |
| CN111858690A (en) * | 2020-07-30 | 2020-10-30 | 中国民航信息网络股份有限公司 | Automatic adjustment method and system for limited sales number of flight space |
| CN112329816A (en) * | 2020-10-09 | 2021-02-05 | 北京嘀嘀无限科技发展有限公司 | Data classification method and device, electronic equipment and readable storage medium |
| CN112651671A (en) * | 2021-01-18 | 2021-04-13 | 中国民航信息网络股份有限公司 | Flight space adjusting method and related equipment |
Non-Patent Citations (3)
| Title |
|---|
| A deep reinforcement learning approach to seat inventory control for airline revenue management;Syed A;IEEE(第01期);全文 * |
| 基于旅客预订行为的航班收益辅助系统建设;王忠韬;李彬;;空运商务(第05期);全文 * |
| 基于轮辐式航线网络的航班舱位控制动态优化;高金敏;乐美龙;;工业工程(05);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112948412A (en) | 2021-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112948412B (en) | Flight inventory updating method, system, electronic device and storage medium | |
| Yu et al. | Prediction of highway tunnel pavement performance based on digital twin and multiple time series stacking | |
| CN104169950B (en) | Utilize the Database Systems of the calculating towards batch processing | |
| Kazak et al. | Artificial intelligence in the tourism sphere | |
| CN110516910A (en) | Declaration form core based on big data protects model training method and core protects methods of risk assessment | |
| CN110288121A (en) | Flight based on multiple time granularity attention mechanism visits rate prediction technique | |
| CN111899059A (en) | Navigation driver revenue management dynamic pricing method based on block chain | |
| CN111340536A (en) | Model training method, passenger seat rate progress prediction method, system, device and medium | |
| CN113656691A (en) | Data prediction method, device and storage medium | |
| CN113724039A (en) | Hotel recommendation method, system and device for OTA website and readable storage medium | |
| Liu et al. | Stochastic airline fleet assignment with risk aversion | |
| CN114782114A (en) | Air ticket price prediction method based on CNN-GRU and SSA-GWO-LightGBM | |
| CN113672797A (en) | Content recommendation method and device | |
| CN113269402B (en) | Flight space control method and device and computer equipment | |
| CN114780600A (en) | Flight searching method, system, equipment and storage medium | |
| CN116662815B (en) | Training method of time prediction model and related equipment | |
| CN111861801A (en) | Hotel full room prediction method, system, equipment and storage medium | |
| Ayaydın et al. | Deep learning based forecasting of delay on flights | |
| Alamdari et al. | Deep reinforcement learning in seat inventory control problem: an action generation approach | |
| CN115049458A (en) | Commodity pushing method and device based on user crowd modeling, medium and equipment | |
| Parbat et al. | Understanding the customer perception using machine learning while booking flight tickets | |
| Gupta et al. | Dynamic Flight Price Prediction Using Machine Learning Algorithms | |
| Wang et al. | Multi-period portfolio with a dynamic reference point considering disappointment feelings | |
| Mojoodi et al. | Designing an algorithm for predicting plane ticket prices using feedforward neural network modeling | |
| Ukai et al. | An Aircraft Deployment Prediction Model Using Machine Learning Techniques |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |