A method, device, medium, and electronic device for refreshing ticket data

Abstract

This application relates to the technical field of online transactions, and provides a method, device, medium, and electronic device for refreshing ticket data. This application obtains factor eigenvalue of each type based on various types of key factor data; obtains multiple prediction values of each type and the weight values corresponding to the prediction values based on the factor eigenvalues of various types; obtains the target factor value of the corresponding type based on the multiple prediction values of each of the various types and the weight values corresponding to the prediction values; obtains the trigger factor value of the corresponding type based on the trigger factor data of various types; when the trigger factor value of any type meets the preset refresh condition of the corresponding type, obtains the refresh frequency of ticket data based on the target factor values of various types and the trigger factor values of various types. By dynamically adjusting the frequency of refreshing tickets and adapting to the frequency of data update, it ensures that the system can obtain key data in a timely manner.

Description

Technical Field

[0001] This application relates to the technical field of online transactions. Specifically, it relates to a method, device, medium, and electronic device for refreshing ticket data. Background Art

[0002] An online trading platform (full English name: Over-The-Air, abbreviated as OTA) upgrades the firmware and software of devices through a wireless network. It enables airlines to upgrade and manage the avionics equipment, sensors, and software of aircraft without having to dock the aircraft on the ground. This can reduce the workload of ground crew and downtime, and improve the efficiency and safety of the fleet.

[0003] With the popularity of global air travel and the intensification of competition, there are challenges in dynamically adjusting ticket prices and cabin classes. Currently, the method of updating ticket data relies on price verification, booking failure, or a fixed expiration time. OTA tickets are either refreshed passively or at too high a frequency, resulting in a poor user experience.

[0004] Therefore, this application provides a method for refreshing ticket data to solve the above technical problems. Summary of the Invention

[0005] The purpose of this application is to provide a method, device, medium, and electronic device for refreshing ticket data, which can solve at least one of the above-mentioned technical problems. The specific solutions are as follows:

[0006] According to the specific embodiments of this application, in the first aspect, this application provides a method for refreshing ticket data, including:

[0007] Obtain various types of key factor data that affect ticket data refreshing, where each of the various types of key factor data includes at least trigger factor data of the corresponding type;

[0008] Obtain factor characteristic values of each type based on the various types of key factor data;

[0009] Obtain multiple prediction values of each type and the weight values of the corresponding prediction values based on the various types of factor characteristic values;

[0010] Obtain the target factor value of the corresponding type based on the multiple prediction values of each type and the weight values of the corresponding prediction values;

[0011] Obtain the trigger factor value of the corresponding type based on the various types of trigger factor data;

[0012] When the trigger factor value of any type meets the preset refresh condition of the corresponding type, obtain the refresh frequency of the ticket data based on the various types of target factor values and the various types of trigger factor values.

[0013] Optionally, obtaining multiple prediction values of each type and the weight values of the corresponding prediction values based on the factor eigenvalue of multiple types includes:

[0014] Applying the factor eigenvalue of multiple types to multiple prediction models respectively to obtain multiple prediction values of the corresponding type;

[0015] Obtaining the weight values of the multiple prediction values of the corresponding type respectively based on the factor eigenvalue of multiple types and the multiple prediction values of the corresponding type.

[0016] Optionally, the multiple prediction models include: time series model, deep learning model, and Bayesian-integration model.

[0017] Optionally, obtaining the weight values of the multiple prediction values of the corresponding type respectively based on the factor eigenvalue of multiple types and the multiple prediction values of the corresponding type includes:

[0018] Calculating the difference between the factor eigenvalue of each type and each prediction value of the corresponding type to obtain the residual value of the corresponding prediction value;

[0019] Calculating the standard deviation of the residual values of each prediction value to obtain the standard deviation of the corresponding prediction value;

[0020] Obtaining the weight values of the multiple prediction values of the corresponding type respectively based on the standard deviation of the multiple prediction values of each type.

[0021] Optionally, obtaining the weight values of the multiple prediction values of the corresponding type respectively based on the standard deviation of the multiple prediction values of each type includes the following formula:

[0022] ;

[0023] Wherein, ω ij represents the weight value of the i th j prediction value of the σ ij represents the standard deviation of the i th j prediction value of the σ ik represents the standard deviation of the i th k prediction value of the N represents the number of the multiple prediction values of the i th j is less than or equal to N .

[0024] Optionally, obtaining the target factor value of the corresponding type based on the multiple prediction values of each of the multiple types and the weight values of the corresponding prediction values includes the following formula:

[0025]

[0026] Wherein, α i represents the target factor value of the i th type, ω ik represents the weight value of the i th prediction value of the k th type, α ik represents the i th prediction value of the k th type, N represents the number of the multiple prediction values of the i th type.

[0027] Optionally, obtaining the refresh frequency of ticket data based on the target factor values of the multiple types and the trigger factor values of the multiple types includes the following formula:

[0028]

[0029] Wherein, F update represents the refresh frequency of ticket data, α i represents the target factor value of the i th type, P i represents the trigger factor value of the i th type, and M represents the number of the multiple types.

[0030] Optionally, the multiple types include: ticket type, cabin type, and external type;

[0031] The key factor data of the ticket type includes a time series value and a ticket price, and the ticket price of the ticket type represents the trigger factor data of the ticket type;

[0032] The key factor data of the cabin type includes: time series value, total number of cabins, and number of sold seats, and the total number of cabins and the number of sold seats of the cabin type represent the trigger factor data of the cabin type;

[0033] The key factor data of the external type includes: time series value, keyword popularity value, key temperature value, key rainfall value, key wind speed value, and key news sentiment value. The keyword popularity value, key temperature value, key rainfall value, key wind speed value, and key news sentiment value of the external type characterize the trigger factor data of the external type.

[0034] Optionally, the factor characteristic values of the ticket type include: time series characteristic value, average ticket price, and increase / decrease rate;

[0035] The factor characteristic values of the cabin type include: time series characteristic value, remaining cabin number, and ratio of sold seats;

[0036] The factor characteristic values of the external type include: time series characteristic value, topic popularity characteristic value, weather impact characteristic value, and news sentiment characteristic value.

[0037] Optionally, the trigger factor value of the ticket type includes price volatility;

[0038] The trigger factor value of the cabin type includes cabin change rate;

[0039] The trigger factor value of the external type includes external factor change rate.

[0040] According to the specific implementation manner of the present application, in the second aspect, the present application provides a ticket data refreshing device, including:

[0041] An acquisition unit, configured to acquire various types of key factor data that affect ticket data refreshing, where at least the trigger factor data of the corresponding type is included in the various types of key factor data;

[0042] A first obtaining unit, configured to obtain the factor characteristic values of each type based on the various types of key factor data;

[0043] A second obtaining unit, configured to obtain multiple prediction values of each type and the weight values of the corresponding prediction values based on the various types of factor characteristic values;

[0044] A third obtaining unit, configured to obtain the target factor value of the corresponding type based on the multiple prediction values of each type and the weight values of the corresponding prediction values;

[0045] A fourth obtaining unit, configured to obtain the trigger factor value of the corresponding type based on the various types of trigger factor data;

[0046] A fifth obtaining unit, configured to obtain the refreshing frequency of ticket data based on the various types of target factor values and the various types of trigger factor values when the trigger factor value of any type meets the preset refreshing condition of the corresponding type.

[0047] Optionally, obtaining multiple prediction values of each type and weight values of corresponding prediction values based on the factor eigenvalue of multiple types includes:

[0048] Applying the factor eigenvalue of multiple types to multiple prediction models respectively to obtain multiple prediction values of corresponding types;

[0049] Obtaining weight values of the multiple prediction values of corresponding types respectively based on the factor eigenvalue of multiple types and the multiple prediction values of corresponding types.

[0050] Optionally, the multiple prediction models include: time series model, deep learning model and Bayesian-integration model.

[0051] Optionally, obtaining weight values of the multiple prediction values of corresponding types respectively based on the factor eigenvalue of multiple types and the multiple prediction values of corresponding types includes:

[0052] Calculating the difference between the factor eigenvalue of each type and each prediction value of corresponding type to obtain the residual value of corresponding prediction value;

[0053] Calculating the standard deviation value of the residual value of each prediction value to obtain the standard deviation value of corresponding prediction value;

[0054] Obtaining weight values of the multiple prediction values of corresponding types respectively based on the standard deviation value of the multiple prediction values of each type.

[0055] Optionally, obtaining weight values of the multiple prediction values of corresponding types respectively based on the standard deviation value of the multiple prediction values of each type includes the following formula:

[0056] ;

[0057] Wherein, ω ij represents the weight value of the i th j th prediction value of the σ ij represents the standard deviation value of the i th j th prediction value of the σ ik represents the standard deviation value of the i th k th prediction value of the N represents the number of the multiple prediction values of the i th type, j less than or equal to N .

[0058] Optionally, obtaining the target factor value of the corresponding type based on the multiple prediction values of each of the multiple types and the weight values of the corresponding prediction values includes the following formula:

[0059]

[0060] Wherein, α i represents the target factor value of the i th type, ω ik represents the weight value of the i th prediction value of the k th type, α ik represents the i th prediction value of the k th type, N represents the number of the multiple prediction values of the i th type.

[0061] Optionally, obtaining the refresh frequency of ticket data based on the target factor values of the multiple types and the trigger factor values of the multiple types includes the following formula:

[0062]

[0063] Wherein, F update represents the refresh frequency of ticket data, α i represents the target factor value of the i th type, P i represents the trigger factor value of the i th type, and M represents the number of the multiple types.

[0064] Optionally, the multiple types include: ticket type, cabin type, and external type;

[0065] The key factor data of the ticket type includes a time series value and a ticket price, and the ticket price of the ticket type represents the trigger factor data of the ticket type;

[0066] The key factor data of the cabin type includes: a time series value, the total number of cabins, and the number of sold seats, and the total number of cabins and the number of sold seats of the cabin type represent the trigger factor data of the cabin type;

[0067] The key factor data of the external type includes: time series value, keyword heat value, key temperature value, key rainfall value, key wind speed value, and key news sentiment value. The keyword heat value, key temperature value, key rainfall value, key wind speed value, and key news sentiment value of the external type characterize the trigger factor data of the external type.

[0068] Optionally, the factor characteristic values of the ticket type include: time series characteristic value, average ticket price, and rise and fall value;

[0069] The factor characteristic values of the cabin type include: time series characteristic value, remaining cabin number, and ratio of sold seats;

[0070] The factor characteristic values of the external type include: time series characteristic value, topic heat characteristic value, weather impact characteristic value, and news sentiment characteristic value.

[0071] Optionally, the trigger factor value of the ticket type includes price volatility;

[0072] The trigger factor value of the cabin type includes cabin change rate;

[0073] The trigger factor value of the external type includes external factor change rate.

[0074] According to the specific implementation manners of the present application, in the third aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the ticket data refreshing method described in any one of the above.

[0075] According to the specific implementation manners of the present application, in the fourth aspect, the present application provides an electronic device, including: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement the ticket data refreshing method described in any one of the above.

[0076] The above solutions of the embodiments of the present application have at least the following beneficial effects compared with the prior art:

[0077] The present application provides a method, apparatus, medium, and electronic device for refreshing ticket data. The present application obtains factor eigenvalue of each type based on multiple types of key factor data; obtains multiple prediction values of each type and the weight values of the corresponding prediction values based on the factor eigenvalues of multiple types; obtains the target factor value of the corresponding type based on the multiple prediction values of each of the multiple types and the weight values of the corresponding prediction values; obtains the trigger factor value of the corresponding type based on the trigger factor data of multiple types; when the trigger factor value of any type meets the preset refresh condition of the corresponding type, obtains the refresh frequency of the ticket data based on the target factor values of multiple types and the trigger factor values of multiple types. By dynamically adjusting the frequency of refreshing tickets and adapting to the frequency of data update, it ensures that the system can obtain key data in a timely manner. It not only maintains the accuracy and adaptability of prediction, but also improves the refresh efficiency, interpretability, and flexibility, making the refresh more adaptable to the changes in ticket data. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 FIG. shows a flowchart of a method for refreshing ticket data according to an embodiment of the present application;

[0079] Figure 2 FIG. shows a block diagram of units of a device for refreshing ticket data according to an embodiment of the present application. DETAILED DESCRIPTION

[0080] In order to make the objectives, technical solutions, and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0081] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. The singular forms "a", "the", and "said" used in the embodiments of the present application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. "Multiple" generally includes at least two.

[0082] It should be understood that the term " / and / " used herein is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after.

[0083] It should be understood that although terms such as first, second, and third may be used in the embodiments of the present application for description, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, without departing from the scope of the embodiments of the present application, the first may also be referred to as the second, and similarly, the second may also be referred to as the first.

[0084] Depending on the context, the words "if" or "when" as used herein may be interpreted as "when" or "while" or "in response to determining" or "in response to detecting". Similarly, depending on the context, the phrase "if determined" or "if detected (stated condition or event)" may be interpreted as "when determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)".

[0085] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a commodity or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such commodity or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the commodity or device comprising said element.

[0086] It should be particularly noted that symbols and / or numbers present in the specification that are not marked in the figure description are not figure reference numerals.

[0087] The optional embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0088] An embodiment of the present application is provided, which is an embodiment of a method for refreshing ticket data.

[0089] The following will be combined with Figure 1 The embodiments of the present application will be described in detail.

[0090] Step S101, obtain various types of key factor data that affect ticket data refreshing.

[0091] Among them, each of the various types of key factor data includes at least trigger factor data corresponding to the type.

[0092] The trigger factor data is used to obtain a trigger factor value, and it is determined whether the ticket data is refreshed based on the trigger factor value.

[0093] In some specific embodiments, the various types include: ticket type, cabin type, and external type.

[0094] Embodiments of this application comprehensively judge the refresh frequency of ticket data through multiple types, improve the real-time response efficiency of ticket data, and reduce unnecessary resource waste.

[0095] Key factor data refers to historical data that affects ticket data.

[0096] The time series value is the time point when data is collected.

[0097] The key factor data of the ticket type includes the time series value and the ticket price, and the ticket price of the ticket type represents the trigger factor data of the ticket type.

[0098] The key factor data of the cabin type includes: the time series value, the total number of cabins, and the number of sold seats. The total number of cabins and the number of sold seats of the cabin type represent the trigger factor data of the cabin type.

[0099] The key factor data of the external type includes: the time series value, the keyword heat value, the key temperature value, the key rainfall value, the key wind speed value, and the key news sentiment value. The keyword heat value, the key temperature value, the key rainfall value, the key wind speed value, and the key news sentiment value of the external type represent the trigger factor data of the external type.

[0100] The key factor data of the external type is derived from social media data, news hot data, and weather data.

[0101] Social media data: Analyze trend topics, popular destinations, and popular events. Obtained through keyword monitoring and sentiment analysis models. For example, the keyword heat value of social media.

[0102] News hot data: Extract information such as events, festivals, and news sentiment related to the destination or time point. For example, the keyword heat value and the key news sentiment value of news hotspots.

[0103] Weather data: Collect the weather changes between the departure place and the destination to evaluate the impact on travel demand. For example, the key temperature value, the key rainfall value, and the key wind speed value.

[0104] This specific embodiment flexibly accesses external media data (such as social media, news hotspots, etc.) and real-time change information, making the prediction more comprehensive and agile, and capable of coping with the impact of emergencies (such as weather, news).

[0105] This specific embodiment classifies the key factor data of the external type and formulates a priority strategy.

[0106] Classification by region: Classify tourist attractions and popular locations according to historical popularity and climate conditions into priority attention areas and general attention areas.

[0107] Priority Focus Areas: Updated monthly based on historical data, areas with higher popularity and better climate in the current month will be listed as priority focus areas (such as northern summer resorts in summer, southern warm areas in winter, etc.).

[0108] Exceptional Events: Entertainment events such as concerts and press conferences are not restricted by geographical classification. For such events, key attention will be paid starting 1 month in advance from the specific date of the official announcement of the event.

[0109] Regarding the collection frequency of social media data, if it is a priority focus area, the interval is 1 day / time; if it is a general focus area, the interval is 3 days / time; if there are changes in the event density, new entertainment events or sudden hotspots occur in the priority focus area or general focus area, the collection frequency of the corresponding area will be automatically increased to ensure the timeliness of the data.

[0110] Regarding the collection frequency during the event, if the event occurs during the priority time period, the 1 week before the start of the event to 1 week after the end of the event will be the key attention period. 1 week before the event, maintain the regular frequency (interval of 1 day / time in the priority focus area, interval of 3 days / time in the general focus area); 3 days before the event, the frequency is accelerated, interval of 12 hours / time; 1 day before the event, the collection frequency increases to interval of 1 hour / time; during the event, maintain interval of 1 hour / time to ensure real-time collection; 1 week after the event, interval of 12 hours / time, and return to the original frequency after the end. During the key attention period, the flight ticket data in the relevant event areas is updated synchronously; the update frequency of the general focus area is restored 1 week after the event ends.

[0111] Regarding the collection frequency of weather data. For the update of regular weather data, the default interval for collecting weather data of the destination and scenic spots is 1 day / time. During the period approaching the event, the update frequency of weather data is adjusted accordingly to analyze the impact of weather on the event in a timely manner; for 3 days before the event and 2 days after the event, the collection frequency of weather data is interval of 1 hour / time; for other time periods, the update frequency of weather data returns to interval of 1 day / time.

[0112] Regarding the attention mode for special events. For events with continuous popularity (such as "Village Super League in Guizhou"), pay attention in the week before and after, follow the above non-special event attention, and collect frequently 1 week before and 1 week after the start of the event; for multiple closely-spaced events, if the interval between sessions is no more than 20 days, continuously pay attention during this period and shorten the prediction frequency to ensure that the data meets the needs of the hotspot; for multiple sessions with a long interval, if the interval between sessions exceeds 20 days, return to the attention frequency of the above non-special events, and the collection frequency also returns to the frequency of the above non-special events.

[0113] Step S102, obtain the factor eigenvalue of each type based on the key factor data of the multiple types.

[0114] In some specific embodiments, the factor feature values of the ticket type include: time series feature values, average ticket price, and increase / decrease rate values.

[0115] The time series feature value is a time feature extracted from the time series value. For example, the time series value indicates which day of the week it is, whether it is a holiday, or whether it is a peak tourist season.

[0116] For example,

[0117]

[0118] Among them, Price mean represents the average ticket price of the ticket type, N represents the total number of time series feature values, Price t represents the t ticket price of the

[0119] For example,

[0120]

[0121] Among them, Price diff represents the increase / decrease rate value of the ticket type, Price t represents the t ticket price of the Price t-1 represents the t- 1st time series feature value of the ticket price.

[0122] The factor feature values of the cabin type include: time series feature values, remaining cabin seats, and ratio of sold seats.

[0123] For example,

[0124]

[0125] Among them, Seats availabel represents the remaining cabin seats of the cabin type, Total Seats represents the total number of cabin seats of the cabin type, Seats sold represents the number of sold seats of the cabin type.

[0126] For example,

[0127]

[0128] Among them, sold_ratio represents the ratio of sold seats of the cabin type, Total SeatsThe total number of cabins representing cabin types, Seats sold The number of sold seats representing cabin types.

[0129] The factor eigenvalue of the external type includes: time series eigenvalue, topic popularity eigenvalue, weather impact eigenvalue, and news sentiment eigenvalue.

[0130] For example,

[0131]

[0132] Among them, Social trend Represents the topic popularity eigenvalue of the external type, weight k Represents the k th preset weight value of the keyword of the external type, trend k Represents the k th keyword popularity value of the keyword of the external type, K Represents the number of keywords.

[0133] For example,

[0134]

[0135] Among them, Weather impact Represents the weather impact eigenvalue of the external type, F Represents the mapping function, Temperature Represents the key temperature value of the external type, Rainfall Represents the key rainfall value of the external type, WindSpeed Represents the key wind speed value of the external type.

[0136] For example,

[0137]

[0138] Among them, News sentiment Represents the news sentiment eigenvalue of the external type, sentiment m Represents the m th key news sentiment value of the keyword of the external type, weight m Represents the m th preset weight value of the keyword of the external type, M represents the number of keywords of the external type.

[0139] Before prediction, duplicate data removal, missing value filling, regional data difference processing, data augmentation processing, and standardization processing are also performed on factor eigenvalues of multiple types.

[0140] The differential processing of regional data can conduct hierarchical analysis on the demand differences in different regions and improve the prediction accuracy. The demand characteristics of different destinations and origins vary greatly, and the regional refinement model can reduce the prediction error.

[0141] The data enhancement processing can generate diverse time series, increase the richness of training data, and enhance the generalization ability of the model.

[0142] Step S103: Obtain multiple prediction values of each type and the weight values corresponding to the prediction values based on the multiple types of factor feature values.

[0143] In some specific embodiments, the obtaining of multiple prediction values of each type and the weight values corresponding to the prediction values based on the multiple types of factor feature values includes:

[0144] Step S103-1: Apply the multiple types of factor feature values to multiple prediction models respectively to obtain multiple prediction values of the corresponding types.

[0145] In some specific embodiments, the multiple prediction models include: time series model, deep learning model, and Bayesian-integration model.

[0146] The time series model is used to capture the linear trend and time series feature values in price and cabin changes, and specifically includes the following formula:

[0147]

[0148] Wherein, α SARIMAX_i represents the i th type of prediction value output by the time series model, f SARIMAX represents the time series model, X it represents at the t th time series feature value the i th type of factor feature value.

[0149] The deep learning model is used to capture the non-linear relationship in the input data, and specifically includes the following formula:

[0150]

[0151] Wherein, α SLTM_i represents the i th type of prediction value output by the deep learning model, SLTM represents the deep learning model, X it represents at the t th time series feature value thei The eigenvalue of a certain type of factor h it-1 Indicates that the deep learning model is at the t-1 On the time series eigenvalue, the i Hidden state of a certain type. The hidden state carries the t-1 Context information on the time series eigenvalue, and its main function is to assist in the prediction of the current time series eigenvalue.

[0152] The Bayesian-integration model includes a Bayesian model and an integration model. The Bayesian model uses Bayesian inference for uncertainty analysis, risk control, and credibility analysis. The Bayesian model specifically includes the following formula:

[0153]

[0154] Where, α i Indicates the i Value of the prediction target variable of a certain type, X it Indicates at the t On the time series eigenvalue, the i Eigenvalue of a certain type of factor, P（α i |X it ） Indicates the credibility value, P(X it |α i ） The likelihood function, which represents the probability of observing α i Under the condition of X it The probability of, P（α i ） Indicates the i Value of the prior probability of a certain type, P(X it ） Indicates the preset marginal probability value.

[0155] The integration model specifically includes the following formula:

[0156]

[0157] Where, α Bayes_i Indicates the i Value of the prediction of a certain type output by the integration model. The integration model is used to calculate the expected value of price prediction, that is, to perform a weighted average of all credibility values.

[0158] This specific embodiment integrates a time series model and a deep learning model, combining the linear trends and periodic features that the time series model is good at with the non-linear modeling ability of the deep learning model, which can not only cope with the long-term trend changes of price and cabin class, but also handle short-term non-linear fluctuations.

[0159] Using the confidence evaluation of the Bayesian-integral model to assign dynamic weights to different models to ensure that the fusion result is more reliable. When significant changes occur in ticket data, the Bayesian-integral model can adaptively adjust the weights, effectively reducing the prediction errors caused by model biases.

[0160] Bayesian model inference helps the model better quantify the uncertainty of the prediction results, thereby avoiding more frequent real-time ticket data updates due to huge prediction errors. In this way, in the face of high volatility and complex market changes, the prediction model is more robust and applicable to scenarios of dynamic pricing and real-time adjustment.

[0161] Step S103-2: Obtain the weight values of the multiple predicted values of each corresponding type based on the multiple types of factor feature values and the multiple predicted values of the corresponding types.

[0162] In some specific embodiments, the obtaining the weight values of the multiple predicted values of each corresponding type based on the multiple types of factor feature values and the multiple predicted values of the corresponding types includes:

[0163] Step S103-21: Calculate the difference between the factor feature value of each type and each predicted value of the corresponding type to obtain the residual value of the corresponding predicted value.

[0164] Step S103-22: Calculate the standard deviation of the residual values of each predicted value to obtain the standard deviation of the corresponding predicted value.

[0165] Step S103-23: Obtain the weight values of the multiple predicted values of each corresponding type based on the standard deviations of the multiple predicted values of each type.

[0166] In some specific embodiments, the obtaining the weight values of the multiple predicted values of each corresponding type based on the standard deviations of the multiple predicted values of each type includes the following formula:

[0167] ;

[0168] Wherein, ω ij represents the weight value of the i th predicted value of the j th type, σ ij represents the i th type of thej The standard deviation of a predicted value σ ik indicating the i th type of the k standard deviation of the N th predicted value, i indicating the number of the multiple predicted values of the j th type, which is less than or equal to N .

[0169] Step S104: Obtain the target factor value of each type based on the multiple predicted values of each of the multiple types and the weight values of the corresponding predicted values.

[0170] In some specific embodiments, obtaining the target factor value of each type based on the multiple predicted values of each of the multiple types and the weight values of the corresponding predicted values includes the following formula:

[0171]

[0172] where, α i indicates the target factor value of the i th type, ω ik indicates the weight value of the i th predicted value of the k th type, α ik indicates the i th predicted value of the k th type, N indicates the number of the multiple predicted values of the i th type.

[0173] Step S105: Obtain the trigger factor value of each type based on the trigger factor data of the multiple types.

[0174] In some specific embodiments, the trigger factor value of the ticket type includes the price volatility; the trigger factor value of the cabin type includes the cabin change rate; the trigger factor value of the external type includes the external factor change rate.

[0175] The price volatility specifically includes the following formula:

[0176]

[0177] where, indicates the average ticket price of the ticket type, Price t indicates the time series value t of the ticket price of the ticket type, Price t-1Indicates the timing value t-1 of the ticket price for the ticket type.

[0178] The cabin class change rate, specifically including the following formula:

[0179]

[0180] Among them, V seat represents the cabin class change rate of the cabin class type, sold_ratio represents the ratio of sold seats of the cabin class type, Seats sold represents the number of sold seats of the cabin class type, Total Seats represents the total number of cabin classes of the cabin class type.

[0181] The external factor change rate, specifically including the following formula:

[0182]

[0183] Among them, Event impact represents the external factor change rate of the external type, Social trend represents the topic heat characteristic value of the external type, Weather impact represents the weather impact characteristic value of the external type, News sentiment represents the news sentiment characteristic value of the external type, φ、 and ψ are the weight values of the impacts of each part, and their sum value is equal to 1.

[0184] Step S106, when the trigger factor value of any type meets the preset refresh condition of the corresponding type, obtain the refresh frequency of the ticket data based on the target factor values of the multiple types and the trigger factor values of the multiple types.

[0185] For example, if the price volatility rate of the ticket type is greater than the price volatility threshold, obtain the refresh frequency of the ticket data based on the target factor values of the multiple types and the trigger factor values of the multiple types; calculate the price standard deviation and the average historical price change based on the historical ticket prices, and then obtain the price volatility threshold based on the product value of the preset multiple value (such as 1.5 times or 2 times) and the price standard deviation, and then the sum value with the average historical price change.

[0186] In some specific embodiments, the obtaining the refresh frequency of the ticket data based on the target factor values of the multiple types and the trigger factor values of the multiple types includes the following formula:

[0187]

[0188] Among them, F update represents the refresh frequency of ticket data, α i represents the i th type of target factor value, P i represents the i th type of trigger factor value, and M represents the number of the multiple types.

[0189] For example, continuing with the above example, specifically, the formula is as follows:

[0190]

[0191] Among them, F update represents the refresh frequency of ticket data, represents the average fare of the ticket type, α represents the target factor value of the ticket type, V seat represents the cabin change rate of the cabin type, β represents the target factor value of the cabin type, Event impact represents the external factor change rate of the external type, γ represents the target factor value of the external type.

[0192] In the embodiments of the present application, factor eigenvalue of each type is obtained based on key factor data of multiple types; multiple predicted values of each type and weight values corresponding to the predicted values are obtained based on the factor eigenvalues of the multiple types; target factor values of corresponding types are obtained based on the multiple predicted values of each of the multiple types and the weight values corresponding to the predicted values; trigger factor values of corresponding types are obtained based on the trigger factor data of the multiple types; when the trigger factor value of any type meets the preset refresh condition of the corresponding type, the refresh frequency of ticket data is obtained based on the target factor values of the multiple types and the trigger factor values of the multiple types. By dynamically adjusting the frequency of refreshing tickets and adapting to the frequency of data update, it ensures that the system can obtain key data in a timely manner. It not only maintains the accuracy and adaptability of prediction, but also improves the refresh efficiency, interpretability and flexibility, making the refresh more adaptable to the change of ticket data.

[0193] The present application also provides an apparatus embodiment for implementing the method steps as described in the above embodiments. The explanation based on the same name meaning is the same as that in the above embodiments, and it has the same technical effects as the above embodiments, which will not be elaborated here.

[0194] Such as Figure 2As shown in the figure, the present application provides a refresh device 200 for ticket data, including:

[0195] An acquisition unit 201, configured to acquire various types of key factor data that affect the refresh of ticket data, where at least the trigger factor data of the corresponding type is included in each of the various types of key factor data;

[0196] A first obtaining unit 202, configured to obtain factor feature values of each type based on the various types of key factor data;

[0197] A second obtaining unit 203, configured to obtain multiple prediction values of each type and weight values corresponding to the prediction values based on the various types of factor feature values;

[0198] A third obtaining unit 204, configured to obtain target factor values of the corresponding type based on the multiple prediction values of each type and the weight values corresponding to the prediction values;

[0199] A fourth obtaining unit 205, configured to obtain trigger factor values of the corresponding type based on the various types of trigger factor data;

[0200] A fifth obtaining unit 206, configured to obtain the refresh frequency of ticket data based on the various types of target factor values and the various types of trigger factor values when the trigger factor value of any type meets the preset refresh condition corresponding to that type.

[0201] Optionally, the obtaining of multiple prediction values of each type and weight values corresponding to the prediction values based on the various types of factor feature values includes:

[0202] Applying the various types of factor feature values to multiple prediction models respectively to obtain multiple prediction values of the corresponding type;

[0203] Obtaining the weight values of the multiple prediction values of the corresponding type based on the various types of factor feature values and the multiple prediction values of the corresponding type.

[0204] Optionally, the multiple prediction models include: time series model, deep learning model, and Bayesian-integration model.

[0205] Optionally, the obtaining of the weight values of the multiple prediction values of the corresponding type based on the various types of factor feature values and the multiple prediction values of the corresponding type includes:

[0206] Calculating the difference between the factor feature value of each type and each prediction value of the corresponding type to obtain the residual value of the corresponding prediction value;

[0207] Calculating the standard deviation of the residual values of each prediction value to obtain the standard deviation of the corresponding prediction value;

[0208] The weight value of each of the multiple prediction values of each type is obtained based on the standard deviation value of each of the multiple prediction values of the corresponding type.

[0209] Optionally, the obtaining of the weight value of each of the multiple prediction values of each type based on the standard deviation value of each of the multiple prediction values of the corresponding type includes the following formula:

[0210] ;

[0211] where, ω ij represents the weight value of the i th j prediction value of the σ ij th i type, j represents the standard deviation value of the σ ik th i prediction value of the k th N type, i represents the number of the multiple prediction values of the j th N type, and is less than or equal to

[0212] Optionally, the obtaining of the target factor value of each type based on the multiple prediction values of each of the multiple types and the weight values of the corresponding prediction values includes the following formula:

[0213]

[0214] where, α i represents the target factor value of the i th ω ik represents the weight value of the i th k prediction value of the α ik th i type, k represents the N th i prediction value of the

[0215] Optionally, the obtaining of the refresh frequency of ticket data based on the target factor values of the multiple types and the trigger factor values of the multiple types includes the following formula:

[0216]

[0217] Among them, F update represents the refresh frequency of ticket data, α i represents the i th type of target factor value, P i represents the i th type of trigger factor value, and M represents the number of the multiple types.

[0218] Optionally, the multiple types include: ticket type, cabin type, and external type;

[0219] The key factor data of the ticket type includes a time series value and a ticket price, and the ticket price of the ticket type represents the trigger factor data of the ticket type;

[0220] The key factor data of the cabin type includes: a time series value, the total number of cabins, and the number of sold seats. The total number of cabins and the number of sold seats of the cabin type represent the trigger factor data of the cabin type;

[0221] The key factor data of the external type includes: a time series value, a keyword heat value, a key temperature value, a key rainfall value, a key wind speed value, and a key news sentiment value. The keyword heat value, key temperature value, key rainfall value, key wind speed value, and key news sentiment value of the external type represent the trigger factor data of the external type.

[0222] Optionally, the factor characteristic values of the ticket type include: a time series characteristic value, an average fare, and a rise and fall value;

[0223] The factor characteristic values of the cabin type include: a time series characteristic value, the remaining number of cabins, and the ratio of sold seats;

[0224] The factor characteristic values of the external type include: a time series characteristic value, a topic heat characteristic value, a weather impact characteristic value, and a news sentiment characteristic value.

[0225] Optionally, the trigger factor value of the ticket type includes a price volatility;

[0226] The trigger factor value of the cabin type includes a cabin change rate;

[0227] The trigger factor value of the external type includes an external factor change rate.

[0228] Embodiments of the present application obtain eigenvalue of each type of factor based on multiple types of key factor data; obtain multiple prediction values of each type and the weight values corresponding to the prediction values based on the eigenvalue of each type of factor; obtain the target factor value of the corresponding type based on the multiple prediction values of each type and the weight values corresponding to the prediction values; obtain the trigger factor value of the corresponding type based on the multiple types of trigger factor data; when the trigger factor value of any type meets the preset refresh condition of the corresponding type, obtain the refresh frequency of the ticket data based on the target factor values of the multiple types and the trigger factor values of the multiple types. By dynamically adjusting the frequency of refreshing tickets and adapting to the frequency of data update, it ensures that the system can obtain key data in a timely manner. It not only maintains the accuracy and adaptability of prediction, but also improves the refresh efficiency, interpretability and flexibility, making the refresh more adaptable to the changes in ticket data.

[0229] Embodiment 3

[0230] This embodiment provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method steps as described in the above embodiment.

[0231] Embodiment 4

[0232] Embodiments of the present application provide a non-volatile computer storage medium, which stores computer-executable instructions that can execute the method steps as described in the above embodiment.

[0233] Finally, it should be noted that the embodiments in this specification are described in a progressive manner, and the key point of each embodiment is to describe the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other. For the systems or devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0234] The above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for refreshing air ticket data, characterized in that: include: Acquire multiple types of key factor data that affect the refreshing of air ticket data, wherein the multiple types include: air ticket type, cabin type and external type, and the multiple types of key factor data at least include trigger factor data of corresponding types; Obtaining factor characteristic values ​​of respective types based on the multiple types of key factor data; Obtaining a plurality of prediction values ​​of respective types based on the plurality of types of factor feature values; Calculate the difference between each type of factor characteristic value and each predicted value of the corresponding type to obtain the residual value of the corresponding predicted value; Calculate the standard deviation of the residual value of each predicted value to obtain the standard deviation of the corresponding predicted value; Obtaining weight values ​​of the plurality of prediction values ​​of the corresponding type based on the standard deviation values ​​of the plurality of prediction values ​​of each type; Obtaining a target factor value of a corresponding type based on the plurality of prediction values ​​of the plurality of types and the weight values ​​of the plurality of prediction values ​​of the corresponding type; Obtaining trigger factor values ​​of corresponding types based on the multiple types of trigger factor data; When any type of trigger factor value satisfies a corresponding type of preset refresh condition, the refresh frequency of the air ticket data is obtained based on the multiple types of target factor values ​​and the multiple types of trigger factor values.

2. The method according to claim 1, characterized in that The obtaining of a plurality of prediction values ​​of respective types based on the plurality of types of factor characteristic values ​​comprises: The multiple types of factor feature values ​​are applied to multiple prediction models respectively to obtain the multiple prediction values ​​of corresponding types.

3. The method according to claim 2, characterized in that The multiple prediction models include: time series model, deep learning model and Bayesian-integral model.

4. The method according to claim 1, characterized in that The method of obtaining the weight values ​​of the corresponding types of the plurality of predicted values ​​based on the respective standard deviation values ​​of the plurality of predicted values ​​of each type comprises the following formula: in, ω ij Indicates i Type j The weight of the predicted value, σ ij Indicates i Type j The standard deviation of the predicted values, σ ik Indicates i Type k The standard deviation of the predicted values, N Indicates i the number of the plurality of predicted values ​​of the type, j Less than or equal to N .

5. The method according to claim 1, characterized in that The target factor value of the corresponding type is obtained based on the respective multiple predicted values ​​of the multiple types and the respective weight values ​​of the multiple predicted values ​​of the corresponding type, Includes the following formulas: in, α i Indicates i Types of target factor values, ω ik Indicates i Type k The weight of the predicted value, α ik Indicates i Type k Prediction values, N Indicates i The number of the multiple predicted values ​​of the type.

6. The method according to claim 1, characterized in that The method of obtaining the refresh frequency of the air ticket data based on the multiple types of target factor values ​​and the multiple types of trigger factor values ​​includes the following formula: in, F update Indicates the refresh frequency of ticket data. α i Indicates i Types of target factor values, P i Indicates i types of trigger factor values, and M represents the number of the multiple types.

7. The method according to claim 1, characterized in that The key factor data of the ticket type includes a time series value and a ticket price, and the ticket price of the ticket type represents the trigger factor data of the ticket type; The key factor data of the class type includes: time series value, total number of classes and number of seats sold, and the total number of classes and number of seats sold of the class type represent the trigger factor data of the class type; The key factor data of the external type include: timing value, keyword heat value, key temperature value, key rainfall value, key wind speed value and key news sentiment value. The keyword heat value, key temperature value, key rainfall value, key wind speed value and key news sentiment value of the external type represent the trigger factor data of the external type.

8. The method according to claim 7, characterized in that The factor characteristic values ​​of the ticket type include: time series characteristic value, ticket price mean and increase / decrease value; The factor characteristic values ​​of the class type include: time series characteristic value, number of remaining class spaces and ratio of sold seats; The external factor characteristic values ​​include: time series characteristic values, topic popularity characteristic values, weather impact characteristic values ​​and news sentiment characteristic values.

9. The method according to claim 7, characterized in that: The trigger factor value of the ticket type includes price volatility; The trigger factor value of the space type includes the space change rate; The external type trigger factor value includes the external factor change rate.

10. A device for refreshing air ticket data, characterized in that: include: An acquisition unit, used to acquire multiple types of key factor data affecting the refreshing of air ticket data, wherein the multiple types include: air ticket type, cabin type and external type, and the multiple types of key factor data at least include trigger factor data of corresponding types; A first obtaining unit, configured to obtain factor characteristic values ​​of respective types based on the multiple types of key factor data; A second obtaining unit, configured to obtain a plurality of prediction values ​​of respective types and weight values ​​of corresponding prediction values ​​based on the plurality of types of factor characteristic values; The step of obtaining a plurality of prediction values ​​of respective types and weight values ​​of corresponding prediction values ​​based on the plurality of types of factor characteristic values ​​includes: Obtaining a plurality of prediction values ​​of respective types based on the plurality of types of factor feature values; Calculate the difference between each type of factor characteristic value and each predicted value of the corresponding type to obtain the residual value of the corresponding predicted value; Calculate the standard deviation of the residual value of each predicted value to obtain the standard deviation of the corresponding predicted value; Obtaining weight values ​​of the plurality of prediction values ​​of the corresponding type based on the standard deviation values ​​of the plurality of prediction values ​​of each type; A third obtaining unit, configured to obtain a target factor value of a corresponding type based on the plurality of prediction values ​​of the plurality of types and the weight values ​​of the plurality of prediction values ​​of the corresponding type; a fourth obtaining unit, configured to obtain trigger factor values ​​of corresponding types based on the multiple types of trigger factor data; The fifth obtaining unit is used to obtain the refresh frequency of the air ticket data based on the multiple types of target factor values ​​and the multiple types of trigger factor values ​​when any type of trigger factor value meets the corresponding type of preset refresh condition.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

12. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as claimed in any one of claims 1 to 9.

Images