Order prediction method, device, electronic device and readable storage medium

Abstract

An order prediction method, device, electronic device and computer-readable storage medium, wherein historical order data of at least three hotel reservation platforms are obtained; data to be analyzed are extracted from the historical order data; features of multiple dimensions of the data to be analyzed are extracted and feature splicing is performed to obtain a feature data set; a pre-built order prediction model is trained using the feature data set to obtain a pre-trained order prediction model; when an order prediction instruction is received, order prediction input data of a target hotel reservation platform is obtained, the order prediction input data is input into the pre-trained order prediction model, and a real-time order prediction result of the target hotel reservation platform is obtained. The above-mentioned order prediction method, device, electronic device and computer-readable storage medium can achieve the purpose of improving the accuracy of hotel order prediction.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer technology, and in particular, to an order prediction method, device, electronic device, and computer-readable storage medium. Background Art

[0002] With the continuous improvement of people's living standards, traveling has become a common leisure activity in people's daily lives. At the same time, business activities are becoming increasingly frequent, which has led to a sustained growth in hotel booking business. The order volume of many hotels has reached a considerable scale.

[0003] As an important place for tourism and business activities, the order volume of hotels is affected by the interaction of many complex factors. Different seasons have obvious differences in tourism peak and off-peak seasons. For example, in summer, there is a strong demand for hotels in coastal cities, while in winter, it may turn to hotels near hot springs or ski resorts; holidays are the peak of travel. During holidays such as the Spring Festival and National Day, people go out or return home, and hotel bookings will rise sharply; during special events, such as large exhibitions and sports events, hotels will also usher in orders. Moreover, the fluctuation of demand at different time points every day (such as morning, noon, and evening) should not be underestimated. The morning may be the peak time for business travelers to check out, and the evening may be the concentrated time for tourists to check in. These factors combined make the hotel order volume highly dynamic and uncertain. Traditional forecasting methods are difficult to accurately grasp its changing trends, which brings great difficulties to the prediction of hotel order volume. Therefore, it is urgent to design an accurate forecasting method for hotel order volume. Summary of the invention

[0004] The embodiments of the present application provide an order prediction method, device, electronic device and computer-readable storage medium, which can achieve the purpose of improving the accuracy of hotel order prediction.

[0005] The present application provides an order prediction method, including:

[0006] Obtain historical order data from at least three hotel booking platforms;

[0007] Extracting data to be analyzed from the historical order data, the data to be analyzed including order time, order time feature identifier, order quantity, order price, inventory quantity, order source, and scheduled lead time;

[0008] Extracting features of multiple dimensions of the data to be analyzed and performing feature splicing to obtain a feature data set, wherein the features of the multiple dimensions include time dimension features, price dimension features, and inventory dimension features;

[0009] Using the feature data set to train a pre-built order prediction model to obtain a pre-trained order prediction model;

[0010] When an order prediction instruction is received, order prediction input data of the target hotel reservation platform is obtained according to the order prediction instruction, and the order prediction input data is input into the pre-trained order prediction model to obtain a real-time prediction result of the order of the target hotel reservation platform.

[0011] The present application also provides an order prediction device, including:

[0012] An acquisition module is used to obtain historical order data of at least three hotel booking platforms;

[0013] A first extraction module is used to extract the data to be analyzed from the historical order data, wherein the data to be analyzed includes order time, order time feature identifier, order quantity, order price, inventory quantity, order source, and scheduled lead time;

[0014] A second extraction module is used to extract features of multiple dimensions of the data to be analyzed and perform feature splicing to obtain a feature data set, wherein the features of the multiple dimensions include time dimension features, price dimension features, and inventory dimension features;

[0015] A training module, used to train a pre-built order prediction model using the feature data set to obtain a pre-trained order prediction model;

[0016] The prediction module is used to obtain the order prediction input data of the target hotel reservation platform according to the order prediction instruction when receiving the order prediction instruction, input the order prediction input data into the pre-trained order prediction model, and obtain the real-time prediction result of the order of the target hotel reservation platform.

[0017] An embodiment of the present application further provides an electronic device, including:

[0018] A memory and a processor; the memory stores executable program code;

[0019] The processor coupled to the memory calls the executable program code stored in the memory to execute the order prediction method provided in the above embodiment.

[0020] On the one hand, an embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the order prediction method provided in the above embodiment is executed.

[0021] From the above embodiments of the present application, it can be known that historical order data of at least three hotel reservation platforms are obtained; data to be analyzed are extracted from the historical order data, and the data to be analyzed include order time, order time feature identifier, order quantity, order price, inventory, order source, and booking lead time; features of multiple dimensions of the data to be analyzed are extracted and feature splicing is performed to obtain a feature data set, wherein the features of the multiple dimensions include time dimension features, price dimension features, and inventory dimension features; a pre-constructed order prediction model is trained using the feature data set to obtain a pre-trained order prediction model; when an order prediction instruction is received, order prediction input data of the target hotel reservation platform is obtained according to the order prediction instruction, and the order prediction input data is input into the pre-trained order prediction model to obtain a real-time prediction result of the order of the target hotel reservation platform. This embodiment uses a large amount of data for model training, thereby obtaining a pre-trained order prediction model that efficiently and accurately predicts the order quantity. When an order prediction instruction is received, the order quantity of the target hotel reservation platform is directly predicted by the model, which can improve the accuracy of hotel order quantity prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0023] Figure 1 A flowchart of an implementation method of an order prediction method provided in an embodiment of the present application;

[0024] Figure 2 A schematic diagram of the structure of an order prediction device provided in one embodiment of the present application;

[0025] Figure 3 A schematic diagram of the hardware structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] See also Figure 1 , Figure 1This is a flowchart of an implementation of an order prediction method provided in an embodiment of the present application. The method can be applied to an electronic device, which can be a personal computer device or a server, such as Figure 1 As shown, the method specifically includes:

[0028] Step S11, obtaining historical order data of at least three hotel reservation platforms.

[0029] In this embodiment, at least three hotel booking platforms include OTA platforms and may also include non-OTA platforms. OTA (Online Travel Agency) refers to the network platform of online travel service providers, such as Ctrip, Fliggy and other platforms. Non-OTA platform refers to the hotel's online sales platform, such as the hotel's official website and the hotel's online store.

[0030] The historical order data refers to the order data of the hotel in the past period of time (i.e., the number of hotel room reservations), for example, the order data in the past three years, or the order data in the past five years. Specifically, the order data includes order data in different years, different seasons, holidays, and non-holidays.

[0031] By collecting order data from different years, seasons, holidays, and non-holidays over the past period of time, we can enrich the diversity of data and improve the accuracy of model predictions.

[0032] Step S12, extracting the data to be analyzed from the historical order data, wherein the data to be analyzed includes order time, order time feature identifier, order quantity, order price, inventory quantity, order source, and scheduled lead time.

[0033] In this embodiment, the order time can be specific to the day, hour and minute, and the order time represents the scheduled time of each order. The order time feature identifier is used to identify whether the time is a holiday. The order quantity is the number of orders at that time point (specifically, it may include the number of orders for each room type), and the order price may be the unit price of each order at the order time, or the total price of all orders at the order time, or the price range of the orders at the order time. The inventory is the number of rooms available for reservation in the hotel. The order origin refers to the customer's place of origin corresponding to each order. The booking lead time refers to the time interval between the order time of each order and the check-in time.

[0034] Furthermore, in an optional embodiment of the present invention, after extracting the data to be analyzed from the historical order data, the method further includes:

[0035] Identifying missing data in the data to be analyzed, and filling the identified missing data; and

[0036] Identify abnormal data in the data to be analyzed, and correct the identified abnormal data.

[0037] In this embodiment, there may be missing data in the data to be analyzed due to system failure, platform interface, etc. Therefore, filling the missing data can restore the real data and improve the accuracy of the prediction. For example, if NA is identified, it is determined to be missing data, and the identified real data can be filled by using methods such as mean filling.

[0038] In this embodiment, if data that is significantly deviated from a preset normal range is identified, the data is determined to be abnormal data, and the abnormal value is corrected. The preset normal range value can be determined based on the statistical distribution of historical data; the correction method can be to replace it with the value of adjacent data or to estimate it based on the trend of the overall data.

[0039] Step S13, extracting features of multiple dimensions of the data to be analyzed and performing feature concatenation to obtain a feature data set, wherein the features of the multiple dimensions include time dimension features, price dimension features, inventory dimension features, and order origin dimension features.

[0040] In this embodiment, after extracting the features of multiple dimensions of the data to be analyzed, the features of multiple dimensions at each time may be concatenated to form time series feature data.

[0041] In this embodiment, the time dimension feature may include multiple sub-features.

[0042] Specifically, the time dimension features include predetermined time lag sub-features. For example, if the hourly order volume is predicted, the order volume in the previous hour, previous two hours, previous 24 hours, etc. can be extracted as features to capture the short-term dynamic changes in the order volume.

[0043] Furthermore, the time dimension feature can also include seasonal sub-features, such as daily, weekly, and monthly periodic features. Specifically, seasonality can be represented by calculating statistics such as the average daily, weekly, and monthly order volume. These features help capture the impact of periodic changes such as tourist peak seasons and off-seasons on order volume.

[0044] Furthermore, the time dimension feature may also include a booking lead time sub-feature, that is, the interval time from the order booking time to the user's check-in time (such as the number of days or hours). The booking lead time feature can reflect the relationship between the user's booking habits and the order volume.

[0045] In this embodiment, the price dimension feature includes a price change rate sub-feature, that is, the change ratio of the current price to the price in the previous time period (such as the previous day or the previous hour), so as to capture the impact of price fluctuations on order volume. Specifically, the order volume can be grouped and counted according to different price ranges to capture the distribution of order volumes at different price levels.

[0046] Furthermore, the price dimension feature also includes a price mean sub-feature, that is, the average value or standard deviation of the order price at the current booking time.

[0047] In this embodiment, the inventory dimension feature includes an inventory remaining rate sub-feature, that is, the ratio of the current inventory remaining quantity to the initial inventory quantity, thereby reflecting the potential impact of the inventory level on the order quantity.

[0048] Furthermore, the inventory dimension feature also includes an inventory mean sub-feature, that is, the average value or standard deviation of the inventory at the current scheduled time.

[0049] In this embodiment, the order origin dimension feature specifically counts the order volume proportions of users from different origins to analyze the impact of differences in booking behavior of users from different regions on the order volume.

[0050] Step S14, using the feature data set to train a pre-built order prediction model to obtain a pre-trained order prediction model.

[0051] In this embodiment, the pre-built order prediction model can be any one of a time series analysis model, a random forest model, a neural network model, or a combination of any two models.

[0052] Further, in an optional embodiment of the present invention, the using the feature data set to train the pre-built order prediction model to obtain the pre-trained order prediction model includes:

[0053] Classifying the feature data set according to the category of the hotel reservation platform to obtain classified feature data sets of different categories;

[0054] The classification feature data set of each category is used to train the pre-built order prediction model corresponding to each category to obtain pre-trained order prediction models of different categories.

[0055] In this embodiment, the categories of hotel reservation platforms can be classified according to the characteristics of the user groups of the hotel reservation platforms (including characteristics such as region, age, price sensitivity, etc.). For example, the categories of hotel reservation platforms include first category platforms, second category platforms, and third category platforms. Among them, the first category platform is a platform mainly for domestic users, the second category is a platform mainly for foreign users, and the third category is a comprehensive platform with equal domestic and foreign users.

[0056] For example, if the hotel reservation platform is a platform that mainly serves the domestic market, mainly has a young user group and a price-sensitive user group, then the pre-built order prediction model corresponding to this platform type may be an XGBoost or LightGBM model; if the hotel reservation platform is a comprehensive platform with equal domestic and foreign users, then the pre-built order prediction model corresponding to this platform type may be a deep neural network, such as CNN or RNN; if the hotel reservation platform is a platform with local users greater than a preset number of users (such as seventy percent), then the pre-built order prediction model corresponding to this platform type may be a time series model, such as ARIMA or SARIMA.

[0057] In this embodiment, models with different features are selected for training for the classification feature data sets corresponding to each type of hotel reservation platform, so that models suitable for different hotel reservation platforms can be obtained, and then analysis and prediction can be performed based on the differences in the data to improve the accuracy of the prediction.

[0058] Step S15, when an order prediction instruction is received, the order prediction input data of the target hotel reservation platform is obtained according to the order prediction instruction, and the order prediction input data is input into the pre-trained order prediction model to obtain the real-time prediction result of the order of the target hotel reservation platform.

[0059] In this embodiment, the order prediction instruction may be an order prediction instruction triggered after the hotel staff clicks the order prediction button.

[0060] Specifically, the order prediction instruction may be parsed to determine which hotel reservation platforms need to predict the order volume at which time, or determine which hotel reservation platforms need to predict the order volume at which time and for which type of rooms.

[0061] Specifically, order forecast input data of a target hotel reservation platform is obtained according to the order forecast instruction, including: obtaining a forecast time and a target hotel reservation platform from the order forecast instruction; obtaining order data of the target hotel reservation platform before the forecast time (such as order data three months or six months before the forecast time, and order data with a forecast time in the same month of the previous year or the year before last), performing feature extraction on the order data (such as extracting forecast time features, historical time dimension features, historical price dimension features, and historical inventory dimension features), and determining that the data obtained by the feature extraction is feature data of the target hotel reservation platform.

[0062] After the feature data of the target hotel reservation platform is input into the pre-trained order prediction model, the model makes predictions based on the input feature data.

[0063] In this embodiment, the target hotel reservation platform may include all or part of the hotel reservation platforms that can be predicted by the pre-trained order prediction model, or may be a hotel reservation platform that can be predicted by the pre-trained order prediction model. The real-time prediction results of the orders of the target hotel reservation platform may also include the prediction results of different room types of the target hotel platform.

[0064] In specific implementation, if pre-trained order prediction models of different categories are obtained through training, then when predicting, if prediction is only made for a certain hotel reservation platform, prediction can be made directly through the pre-trained order prediction model corresponding to the category of the hotel reservation platform to be predicted, thereby improving the prediction efficiency.

[0065] Further, in an optional embodiment of the invention, after obtaining the real-time prediction result of the order of the target hotel reservation platform, the method further includes:

[0066] Obtain the number of visits, clicks, and conversion rate of the target hotel reservation platform within a preset time period before the current prediction time;

[0067] Determine a correction factor based on the visits, clicks and conversion rates;

[0068] The correction factor is used to correct the real-time prediction result of the order of the target hotel reservation platform.

[0069] In this embodiment, the number of visits, clicks and conversion rate of the hotel reservation platform and the order data can be regressed in advance to determine the linear or nonlinear relationship between the number of visits, clicks and conversion rate of the hotel reservation platform and the order data. The correction factor is determined according to the number of visits, clicks and conversion rate, including: determining the correction factor according to the linear or nonlinear relationship between the number of visits, clicks and conversion rate of the hotel reservation platform and the order data.

[0070] Furthermore, the real-time order prediction result is corrected by using the correction factor, including: performing linear or nonlinear calculation on the real-time order prediction result by using the correction factor.

[0071] For example, the real-time order prediction result predicts that the order volume of the target hotel reservation platform is 1,000 rooms, and the recent visits to the target hotel reservation platform have increased by 20%. According to the regression relationship of historical data, this may lead to a 10% increase in the order volume of the target hotel reservation platform. In this case, the 1,000 rooms predicted by the real-time order prediction result can be revised to 1,100 rooms.

[0072] Further, in an optional embodiment of the present invention, after the real-time prediction result of the order of the target hotel reservation platform is obtained, the method further includes:

[0073] The room price or inventory is adjusted according to the real-time prediction result of the order of the target hotel reservation platform and the real-time inventory of the target hotel reservation platform.

[0074] In this embodiment, the real-time order prediction result of the target hotel reservation platform is the order volume of the target hotel reservation platform at the prediction time, and the real-time inventory of the target hotel reservation platform is the remaining bookable inventory displayed on the target hotel reservation platform.

[0075] Specifically, if the difference between the order volume of the target hotel reservation platform at the predicted time and the real-time inventory of the target hotel reservation platform is greater than the inventory safety threshold, an instruction to increase the inventory of the target hotel platform is sent to increase the number of rooms that can be booked on the target hotel platform. For example, if the order volume of platform A is predicted to be 300 rooms in a certain period of time, and the current inventory is 250 rooms, it means that there is a gap of 50 rooms, and the inventory needs to be adjusted in advance for users on platform A to place orders and increase customer acquisition rate and user experience.

[0076] And, if the order volume of the target hotel reservation platform at the forecast time is the first preset multiple of the real-time inventory of the target hotel reservation platform, and the user group of the target hotel reservation platform is not price sensitive, then send an instruction to increase the room price of the target hotel platform, so that the scheduled price displayed on the target hotel platform increases. If the real-time inventory of the target hotel reservation platform is the second preset multiple of the order volume of the target hotel reservation platform at the forecast time, and the user group of the target hotel reservation platform is price sensitive, then send an instruction to reduce the room price of the target hotel platform, so that the scheduled price displayed on the target hotel platform is reduced. Wherein, the first preset multiple and the second preset multiple can be the same or different values, and whether the user group of the target hotel reservation platform is price sensitive can be determined by obtaining the attribute identification of the target hotel reservation platform, and the attribute identification can be digitally identified by whether the user group is price sensitive.

[0077] Furthermore, in an optional embodiment of the present invention, after adjusting the room price or inventory, the method further includes:

[0078] An adjustment report is generated, and the adjustment report is sent to a target person.

[0079] In this embodiment, after each price and inventory adjustment, an adjustment report is generated and sent to a target person (eg, an operation team member).

[0080] Specifically, when generating an adjustment report, you can also collect all relevant adjustment data, including price changes, inventory updates, order volume changes, etc.; then organize the collected data for easy analysis and presentation; and then extract key information, such as adjustment time, adjustment reasons, comparative data before and after adjustment, etc.; finally, use templates or automated tools to generate reports to ensure that the report content is clear and logical.

[0081] Specifically, adjustment reports can use charts and graphs to visualize data, making the report more intuitive and easy to understand. In addition, adjustment reports can be sent via email, internal website or other communication tools.

[0082] Furthermore, before sending the adjustment report to the target personnel, it can be reviewed and verified to ensure the accuracy and completeness of the adjustment report. It can also receive feedback from the operations team on the report to continuously improve the quality and usefulness of the report.

[0083] Further, in an optional embodiment of the present invention, after obtaining the real-time prediction result of the order of the target hotel reservation platform, the method further includes:

[0084] Determine whether the real-time forecast result of the order is less than a preset standard threshold;

[0085] If the real-time prediction result of the order is less than the preset standard threshold, determine whether there is a system failure in the hotel system;

[0086] If it is determined that there is a fault in the hotel system, an early warning message is sent.

[0087] In this embodiment, the preset standard threshold value may be a preset value, for example, a value preset by a system administrator. Alternatively, the preset standard threshold value may be the lowest value of the historical order volume in the same period.

[0088] In this embodiment, the hotel system refers to the hotel's operating system, which communicates with various hotel reservation platforms through interface calls.

[0089] In this embodiment, if the order real-time prediction result is less than the preset standard threshold, it is determined whether the system is faulty to identify whether the reduction in order volume is caused by system connection. If it is determined that the system is faulty, an early warning message can be sent to the system maintenance personnel, or a repair script can be run and a repair configuration can be performed to repair the system.

[0090] In this embodiment, the real-time order prediction results can be used to timely check and warn the system operation status, thereby improving the stability of system operation.

[0091] Further, in an optional embodiment of the invention, the step of determining whether there is a system failure in the hotel system includes:

[0092] Obtain system logs and performance metrics for hotel systems;

[0093] comparing the performance indicator with a performance threshold, and searching for target information in the system log;

[0094] Determine whether the hotel system has a system failure based on the comparison result and the search result.

[0095] In this embodiment, the performance indicators of the hotel system include but are not limited to the CPU usage, memory usage, network delay, request response time, process status, and disk I / O of the hotel system.

[0096] In this embodiment, the performance threshold can be preset, or can be the normal range of the performance indicator predicted by the machine learning model. For example, if the CPU utilization exceeds a certain threshold (such as 90%), there may be performance problems; if the memory usage rate exceeds 80%, there may be memory leaks or insufficient resources; when the disk usage reaches 80%-90%, it may cause the disk to be full and data writing to fail; if certain key services (such as databases, Web servers) are not started or stopped, the system operation is affected.

[0097] In this embodiment, searching for target information in the system log includes searching for information identifying abnormalities such as error codes, abnormal request modes, and server errors in the system log.

[0098] In this embodiment, determining whether the hotel system has a system failure according to the comparison result and the search result includes:

[0099] If the target performance index is abnormal in the comparison result, it is determined that the system has a fault; or

[0100] If the target information exists in the system log, it is determined that there is a system failure; or

[0101] If there is an abnormality in the target performance indicator in the comparison result and there is target information in the system log, it is determined that there is a fault in the system.

[0102] In this embodiment, the number of target performance indicators may be one or at least two. For example, when the CPU usage rate exceeds 90%, it is determined that a system fault exists.

[0103] In this embodiment, when an error, warning or abnormal information appears in the system log, it can be determined that the system has a fault. It can also be determined whether there is a system fault by combining the system log and the performance index for comprehensive judgment.

[0104] In an embodiment of the present application, historical order data of at least three hotel reservation platforms are obtained; data to be analyzed are extracted from the historical order data, and the data to be analyzed include order time, order time feature identifier, order quantity, order price, inventory, order source, and booking lead time; features of multiple dimensions of the data to be analyzed are extracted and feature splicing is performed to obtain a feature data set, wherein the features of the multiple dimensions include time dimension features, price dimension features, and inventory dimension features; a pre-constructed order prediction model is trained using the feature data set to obtain a pre-trained order prediction model; when an order prediction instruction is received, order prediction input data of the target hotel reservation platform is obtained according to the order prediction instruction, and the order prediction input data is input into the pre-trained order prediction model to obtain a real-time prediction result of the order of the target hotel reservation platform. This embodiment uses a large amount of data to perform model training, thereby obtaining a pre-trained order prediction model that efficiently and accurately predicts the order quantity. When an order prediction instruction is received, the order quantity of the target hotel reservation platform is directly predicted by the model, which can improve the accuracy of hotel order quantity prediction.

[0105] See also Figure 2 , a schematic diagram of the structure of an order prediction device provided by an embodiment of the present application. For the convenience of explanation, only the parts related to the embodiment of the present application are shown. The device can be set in an electronic device, which can be a personal computer device or a server. The device includes:

[0106] The acquisition module 201 is used to acquire historical order data of at least three hotel reservation platforms.

[0107] In this embodiment, at least three hotel booking platforms include OTA platforms and may also include non-OTA platforms. OTA (Online Travel Agency) refers to the network platform of online travel service providers, such as Ctrip, Fliggy and other platforms. Non-OTA platform refers to the hotel's online sales platform, such as the hotel's official website and the hotel's online store.

[0108] The historical order data refers to the order data of the hotel in the past period of time (i.e., the number of hotel room reservations), for example, the order data in the past three years, or the order data in the past five years. Specifically, the order data includes order data in different years, different seasons, holidays, and non-holidays.

[0109] By collecting order data from different years, seasons, holidays, and non-holidays over the past period of time, we can enrich the diversity of data and improve the accuracy of model predictions.

[0110] The first extraction module 202 is used to extract the data to be analyzed from the historical order data, wherein the data to be analyzed includes order time, order time feature identifier, order quantity, order price, inventory, order source, and scheduled lead time.

[0111] In this embodiment, the order time can be specific to the day, hour and minute, and the order time represents the scheduled time of each order. The order time feature identifier is used to identify whether the time is a holiday. The order quantity is the number of orders at that time point (specifically, it may include the number of orders for each room type), and the order price may be the unit price of each order at the order time, or the total price of all orders at the order time, or the price range of the orders at the order time. The inventory is the number of rooms available for reservation in the hotel. The order origin refers to the customer's place of origin corresponding to each order. The booking lead time refers to the time interval between the order time of each order and the check-in time.

[0112] Furthermore, in an optional embodiment of the present invention, the first extraction module 202 is also used to identify missing data in the data to be analyzed after extracting the data to be analyzed from the historical order data, and fill in the identified missing data; and identify abnormal data in the data to be analyzed, and correct the identified abnormal data.

[0113] In this embodiment, there may be missing data in the data to be analyzed due to system failure, platform interface, etc. Therefore, filling the missing data can restore the real data and improve the accuracy of the prediction. For example, if NA is identified, it is determined to be missing data, and the identified real data can be filled by using methods such as mean filling.

[0114] In this embodiment, if data that is significantly deviated from a preset normal range is identified, the data is determined to be abnormal data, and the abnormal value is corrected. The preset normal range value can be determined based on the statistical distribution of historical data; the correction method can be to replace it with the value of adjacent data or to estimate it based on the trend of the overall data.

[0115] The second extraction module 203 is used to extract features of multiple dimensions of the data to be analyzed and perform feature concatenation to obtain a feature data set, wherein the features of the multiple dimensions include time dimension features, price dimension features, and inventory dimension features.

[0116] In this embodiment, after extracting the features of multiple dimensions of the data to be analyzed, the features of multiple dimensions at each time may be concatenated to form time series feature data.

[0117] In this embodiment, the time dimension feature may include multiple sub-features.

[0118] Specifically, the time dimension features include predetermined time lag sub-features. For example, if the hourly order volume is predicted, the order volume in the previous hour, previous two hours, previous 24 hours, etc. can be extracted as features to capture the short-term dynamic changes in the order volume.

[0119] Furthermore, the time dimension feature can also include seasonal sub-features, such as daily, weekly, and monthly periodic features. Specifically, seasonality can be represented by calculating statistics such as the average daily, weekly, and monthly order volume. These features help capture the impact of periodic changes such as tourist peak seasons and off-seasons on order volume.

[0120] Furthermore, the time dimension feature may also include a booking lead time sub-feature, that is, the interval time from the order booking time to the user's check-in time (such as the number of days or hours). The booking lead time feature can reflect the relationship between the user's booking habits and the order volume.

[0121] In this embodiment, the price dimension feature includes a price change rate sub-feature, that is, the change ratio of the current price to the price in the previous time period (such as the previous day or the previous hour), so as to capture the impact of price fluctuations on order volume. Specifically, the order volume can be grouped and counted according to different price ranges to capture the distribution of order volumes at different price levels.

[0122] Furthermore, the price dimension feature also includes a price mean sub-feature, that is, the average value or standard deviation of the order price at the current booking time.

[0123] In this embodiment, the inventory dimension feature includes an inventory remaining rate sub-feature, that is, the ratio of the current inventory remaining quantity to the initial inventory quantity, thereby reflecting the potential impact of the inventory level on the order quantity.

[0124] Furthermore, the inventory dimension feature also includes an inventory mean sub-feature, that is, the average value or standard deviation of the inventory at the current scheduled time.

[0125] In this embodiment, the order origin dimension feature specifically counts the order volume proportions of users from different origins to analyze the impact of differences in booking behavior of users from different regions on the order volume.

[0126] The training module 204 is used to train the pre-built order prediction model using the feature data set to obtain a pre-trained order prediction model.

[0127] In this embodiment, the pre-built order prediction model can be any one of a time series analysis model, a random forest model, a neural network model, or a combination of any two models.

[0128] Furthermore, in an optional embodiment of the present invention, the training module 204 is specifically used for:

[0129] Classifying the feature data set according to the category of the hotel reservation platform to obtain classified feature data sets of different categories;

[0130] The classification feature data set of each category is used to train the pre-built order prediction model corresponding to each category to obtain pre-trained order prediction models of different categories.

[0131] In this embodiment, the categories of hotel reservation platforms can be classified according to the characteristics of the user groups of the hotel reservation platforms (including characteristics such as region, age, price sensitivity, etc.). For example, the categories of hotel reservation platforms include first category platforms, second category platforms, and third category platforms. Among them, the first category platform is a platform mainly for domestic users, the second category is a platform mainly for foreign users, and the third category is a comprehensive platform with equal domestic and foreign users.

[0132] For example, if the hotel reservation platform is a platform that mainly serves the domestic market, mainly has a young user group and a price-sensitive user group, then the pre-built order prediction model corresponding to this platform type may be an XGBoost or LightGBM model; if the hotel reservation platform is a comprehensive platform with equal domestic and foreign users, then the pre-built order prediction model corresponding to this platform type may be a deep neural network, such as CNN or RNN; if the hotel reservation platform is a platform with local users greater than a preset number of users (such as seventy percent), then the pre-built order prediction model corresponding to this platform type may be a time series model, such as ARIMA or SARIMA.

[0133] In this embodiment, models with different features are selected for training for the classification feature data sets corresponding to each type of hotel reservation platform, so that models suitable for different hotel reservation platforms can be obtained, and then analysis and prediction can be performed based on the differences in the data to improve the accuracy of the prediction.

[0134] The prediction module 205 is used to obtain the order prediction input data of the target hotel reservation platform according to the order prediction instruction when receiving the order prediction instruction, input the order prediction input data into the pre-trained order prediction model, and obtain the real-time prediction result of the order of the target hotel reservation platform.

[0135] In this embodiment, the order prediction instruction may be an order prediction instruction triggered after the hotel staff clicks the order prediction button.

[0136] Specifically, the order prediction instruction may be parsed to determine which hotel reservation platforms need to predict the order volume at which time, or determine which hotel reservation platforms need to predict the order volume at which time and for which type of rooms.

[0137] Specifically, order forecast input data of a target hotel reservation platform is obtained according to the order forecast instruction, including: obtaining a forecast time and a target hotel reservation platform from the order forecast instruction; obtaining order data of the target hotel reservation platform before the forecast time (such as order data three months or six months before the forecast time, and order data with a forecast time in the same month of the previous year or the year before last), performing feature extraction on the order data (such as extracting forecast time features, historical time dimension features, historical price dimension features, and historical inventory dimension features), and determining that the data obtained by the feature extraction is feature data of the target hotel reservation platform.

[0138] After the feature data of the target hotel reservation platform is input into the pre-trained order prediction model, the model makes predictions based on the input feature data.

[0139] In this embodiment, the target hotel reservation platform may include all or part of the hotel reservation platforms that can be predicted by the pre-trained order prediction model, or may be a hotel reservation platform that can be predicted by the pre-trained order prediction model. The real-time prediction results of the orders of the target hotel reservation platform may also include the prediction results of different room types of the target hotel platform.

[0140] In specific implementation, if pre-trained order prediction models of different categories are obtained through training, then when predicting, if prediction is only made for a certain hotel reservation platform, prediction can be made directly through the pre-trained order prediction model corresponding to the category of the hotel reservation platform to be predicted, thereby improving the prediction efficiency.

[0141] Furthermore, in an optional embodiment of the invention, the order prediction device further includes:

[0142] The correction module is used to obtain the real-time prediction results of orders of the target hotel reservation platform, and then obtain the visits, clicks and conversion rate of the target hotel reservation platform in a preset time period before the current prediction time; determine the correction factor according to the visits, clicks and conversion rate; and use the correction factor to correct the real-time prediction results of orders of the target hotel reservation platform.

[0143] In this embodiment, the number of visits, clicks and conversion rate of the hotel reservation platform and the order data can be regressed in advance to determine the linear or nonlinear relationship between the number of visits, clicks and conversion rate of the hotel reservation platform and the order data. The correction factor is determined according to the number of visits, clicks and conversion rate, including: determining the correction factor according to the linear or nonlinear relationship between the number of visits, clicks and conversion rate of the hotel reservation platform and the order data.

[0144] Furthermore, the correction module uses the correction factor to correct the real-time order prediction result, including: the correction module uses the correction factor to perform linear or nonlinear calculation on the real-time order prediction result.

[0145] For example, the real-time order prediction result predicts that the order volume of the target hotel reservation platform is 1,000 rooms, and the recent visits to the target hotel reservation platform have increased by 20%. According to the regression relationship of historical data, this may lead to a 10% increase in the order volume of the target hotel reservation platform. In this case, the 1,000 rooms predicted by the real-time order prediction result can be revised to 1,100 rooms.

[0146] Furthermore, in an optional embodiment of the present invention, the order prediction device further includes:

[0147] The adjustment module is used to adjust the room price or inventory according to the real-time prediction result of the order of the target hotel reservation platform and the real-time inventory of the target hotel reservation platform after the real-time prediction result of the order of the target hotel reservation platform is obtained.

[0148] In this embodiment, the real-time order prediction result of the target hotel reservation platform is the order volume of the target hotel reservation platform at the prediction time, and the real-time inventory of the target hotel reservation platform is the remaining bookable inventory displayed on the target hotel reservation platform.

[0149] Specifically, if the difference between the order volume of the target hotel reservation platform at the predicted time and the real-time inventory of the target hotel reservation platform is greater than the inventory safety threshold, an instruction to increase the inventory of the target hotel platform is sent to increase the number of rooms that can be booked on the target hotel platform. For example, if the order volume of platform A is predicted to be 300 rooms in a certain period of time, and the current inventory is 250 rooms, it means that there is a gap of 50 rooms, and the inventory needs to be adjusted in advance for users on platform A to place orders and increase customer acquisition rate and user experience.

[0150] And, if the order volume of the target hotel reservation platform at the forecast time is the first preset multiple of the real-time inventory of the target hotel reservation platform, and the user group of the target hotel reservation platform is not price sensitive, then send an instruction to increase the room price of the target hotel platform, so that the scheduled price displayed on the target hotel platform increases. If the real-time inventory of the target hotel reservation platform is the second preset multiple of the order volume of the target hotel reservation platform at the forecast time, and the user group of the target hotel reservation platform is price sensitive, then send an instruction to reduce the room price of the target hotel platform, so that the scheduled price displayed on the target hotel platform is reduced. Wherein, the first preset multiple and the second preset multiple can be the same or different values, and whether the user group of the target hotel reservation platform is price sensitive can be determined by obtaining the attribute identification of the target hotel reservation platform, and the attribute identification can be digitally identified by whether the user group is price sensitive.

[0151] Furthermore, in an optional embodiment of the present invention, the order prediction device further includes:

[0152] The reporting module is used to generate an adjustment report after adjusting the room price or inventory, and send the adjustment report to the target person.

[0153] In this embodiment, after each price and inventory adjustment, an adjustment report is generated and sent to a target person (eg, an operation team member).

[0154] Specifically, when generating an adjustment report, you can also collect all relevant adjustment data, including price changes, inventory updates, order volume changes, etc.; then organize the collected data for easy analysis and presentation; and then extract key information, such as adjustment time, adjustment reasons, comparative data before and after adjustment, etc.; finally, use templates or automated tools to generate reports to ensure that the report content is clear and logical.

[0155] Specifically, adjustment reports can use charts and graphs to visualize data, making the report more intuitive and easy to understand. In addition, adjustment reports can be sent via email, internal website or other communication tools.

[0156] Furthermore, before sending the adjustment report to the target personnel, it can be reviewed and verified to ensure the accuracy and completeness of the adjustment report. It can also receive feedback from the operations team on the report to continuously improve the quality and usefulness of the report.

[0157] Further, in an optional embodiment of the present invention, after obtaining the real-time prediction result of the order of the target hotel reservation platform, the method further includes:

[0158] Determine whether the real-time forecast result of the order is less than a preset standard threshold;

[0159] If the real-time prediction result of the order is less than the preset standard threshold, determine whether there is a system failure in the hotel system;

[0160] If it is determined that there is a fault in the hotel system, an early warning message is sent.

[0161] In this embodiment, the preset standard threshold value may be a preset value, for example, a value preset by a system administrator. Alternatively, the preset standard threshold value may be the lowest value of the historical order volume in the same period.

[0162] In this embodiment, the hotel system refers to the hotel's operating system, which communicates with various hotel reservation platforms through interface calls.

[0163] In this embodiment, if the order real-time prediction result is less than the preset standard threshold, it is determined whether the system is faulty to identify whether the reduction in order volume is caused by system connection. If it is determined that the system is faulty, an early warning message can be sent to the system maintenance personnel, or a repair script can be run and a repair configuration can be performed to repair the system.

[0164] In this embodiment, the real-time order prediction results can be used to timely check and warn the system operation status, thereby improving the stability of system operation.

[0165] Further, in an optional embodiment of the invention, the step of determining whether there is a system failure in the hotel system includes:

[0166] Obtain system logs and performance metrics for hotel systems;

[0167] comparing the performance indicator with a performance threshold, and searching for target information in the system log;

[0168] Determine whether the hotel system has a system failure based on the comparison result and the search result.

[0169] In this embodiment, the performance indicators of the hotel system include but are not limited to the CPU usage, memory usage, network delay, request response time, process status, and disk I / O of the hotel system.

[0170] In this embodiment, the performance threshold can be preset, or can be the normal range of the performance indicator predicted by the machine learning model. For example, if the CPU utilization exceeds a certain threshold (such as 90%), there may be performance problems; if the memory usage rate exceeds 80%, there may be memory leaks or insufficient resources; when the disk usage reaches 80%-90%, it may cause the disk to be full and data writing to fail; if certain key services (such as databases, Web servers) are not started or stopped, the system operation is affected.

[0171] In this embodiment, searching for target information in the system log includes searching for information identifying abnormalities such as error codes, abnormal request modes, and server errors in the system log.

[0172] In this embodiment, determining whether the hotel system has a system failure according to the comparison result and the search result includes:

[0173] If the target performance index is abnormal in the comparison result, it is determined that the system has a fault; or

[0174] If the target information exists in the system log, it is determined that there is a system failure; or

[0175] If there is an abnormality in the target performance indicator in the comparison result and there is target information in the system log, it is determined that there is a fault in the system.

[0176] In this embodiment, the number of target performance indicators may be one or at least two. For example, when the CPU usage rate exceeds 90%, it is determined that a system fault exists.

[0177] In this embodiment, when an error, warning or abnormal information appears in the system log, it can be determined that the system has a fault. It can also be determined whether there is a system fault by combining the system log and the performance index for comprehensive judgment.

[0178] In an embodiment of the present application, historical order data of at least three hotel reservation platforms are obtained; data to be analyzed are extracted from the historical order data, and the data to be analyzed include order time, order time feature identifier, order quantity, order price, inventory, order source, and booking lead time; features of multiple dimensions of the data to be analyzed are extracted and feature splicing is performed to obtain a feature data set, wherein the features of the multiple dimensions include time dimension features, price dimension features, and inventory dimension features; a pre-constructed order prediction model is trained using the feature data set to obtain a pre-trained order prediction model; when an order prediction instruction is received, order prediction input data of the target hotel reservation platform is obtained according to the order prediction instruction, and the order prediction input data is input into the pre-trained order prediction model to obtain a real-time prediction result of the order of the target hotel reservation platform. This embodiment uses a large amount of data to perform model training, thereby obtaining a pre-trained order prediction model that efficiently and accurately predicts the order quantity. When an order prediction instruction is received, the order quantity of the target hotel reservation platform is directly predicted by the model, which can improve the accuracy of hotel order quantity prediction.

[0179] See also Figure 3 , a schematic diagram of the hardware structure of an electronic device provided in one embodiment of the present application.

[0180] Exemplarily, the electronic device can be any one of various types of computer system devices that are mobile or portable and perform wireless communication. Specifically, the electronic device can be a mobile phone or a smart phone (e.g., a phone based on iPhoneTM or AndroidTM).

[0181] The electronic device may also be any of a plurality of electronic devices including, but not limited to, cellular phones, smart phones, other wireless communication devices, personal digital assistants, other media players, calculators, programmable remote controls, laptop computers, desktop computers, and combinations thereof.

[0182] In some cases, the electronic device can perform multiple functions (e.g., play music, display video, store pictures, and receive and send phone calls). If desired, the electronic device can be a portable device such as a cellular phone, a media player, other handheld device, a wristwatch device, or other compact portable device.

[0183] like Figure 3 As shown, the electronic device 10 may include a control circuit, which may include a storage and processing circuit 30. The storage and processing circuit 30 may include a memory, such as a hard disk drive memory, a non-volatile memory (such as a flash memory or other electronically programmable memory with limited deletion for forming a solid-state drive, etc.), a volatile memory (such as a static or dynamic random access memory, etc.), etc., which is not limited in the embodiment of the present application. The processing circuit in the storage and processing circuit 30 can be used to control the operation of the electronic device 10. The processing circuit can be implemented based on one or more microprocessors, microcontrollers, digital signal processors, baseband processors, power management units, audio codec chips, application-specific integrated circuits, display driver integrated circuits, etc.

[0184] The storage and processing circuit 30 can be used to run software in the electronic device 10, such as an Internet browsing application, a Voice over Internet Protocol (VOIP) phone call application, an email application, a media playback application, an operating system function, etc. These software can be used to perform some control operations, such as image acquisition based on a camera, ambient light measurement based on an ambient light sensor, proximity sensor measurement based on a proximity sensor, information display functions based on status indicators such as status indicators of light-emitting diodes, touch event detection based on a touch sensor, functions associated with displaying information on multiple (e.g., layered) displays, operations associated with performing wireless communication functions, operations associated with collecting and generating audio signals, control operations associated with collecting and processing button press event data, and other functions in the electronic device 10, etc., which are not limited by the embodiments of the present application.

[0185] Furthermore, the memory stores an executable program code, and the processor coupled to the memory calls the executable program code stored in the memory to execute the above Figure 1 The order prediction method described in the illustrated embodiment.

[0186] The executable program code includes the above Figure 2 The various modules in the order prediction device described in the illustrated embodiment, for example, are: an acquisition module, a first extraction module, a second extraction module, a training module, and a prediction module.

[0187] The electronic device 10 may further include an input-output circuit 42. The input-output circuit 42 may be used to enable the electronic device 10 to input and output data, that is, to allow the electronic device 10 to receive data from an external device and also to allow the electronic device 10 to output data from the electronic device 10 to an external device. The input-output circuit 42 may further include a sensor 32. The sensor 32 may include an ambient light sensor, a proximity sensor based on light and capacitance, a touch sensor (e.g., based on a light touch sensor and / or a capacitive touch sensor, wherein the touch sensor may be part of a touch display screen or may be used independently as a touch sensor structure), an acceleration sensor, and other sensors.

[0188] The input-output circuit 42 may also include one or more displays, such as the display 14. The display 14 may include a liquid crystal display, an organic light-emitting diode display, an electronic ink display, a plasma display, a display using other display technologies, or a combination of several of them. The display 14 may include a touch sensor array (i.e., the display 14 may be a touch-controlled display screen). The touch sensor may be a capacitive touch sensor formed by an array of transparent touch sensor electrodes (e.g., indium tin oxide (ITO) electrodes), or may be a touch sensor formed using other touch technologies, such as sonic touch, pressure-sensitive touch, resistive touch, optical touch, etc., which is not limited in the embodiments of the present application.

[0189] The electronic device 10 may also include an audio component 36. The audio component 36 may be used to provide audio input and output functions for the electronic device 10. The audio component 36 in the electronic device 10 may include a speaker, a microphone, a buzzer, a tone generator, and other components for generating and detecting sound.

[0190] The communication circuit 38 can be used to provide the electronic device 10 with the ability to communicate with external devices. The communication circuit 38 may include analog and digital input / output interface circuits, and wireless communication circuits based on radio frequency signals and / or optical signals. The wireless communication circuit in the communication circuit 38 may include a radio frequency transceiver circuit, a power amplifier circuit, a low noise amplifier, a switch, a filter, and an antenna. For example, the wireless communication circuit in the communication circuit 38 may include a circuit for supporting near field communication (NFC) by transmitting and receiving near field coupled electromagnetic signals. For example, the communication circuit 38 may include a near field communication antenna and a near field communication transceiver. The communication circuit 38 may also include a cellular phone transceiver and antenna, a wireless local area network transceiver circuit and antenna, etc.

[0191] The electronic device 10 may further include a battery, power management circuits and other input / output units 40. The input / output units 40 may include buttons, joysticks, click wheels, scroll wheels, touch pads, keypads, keyboards, cameras, light emitting diodes and other status indicators.

[0192] A user may input commands through the input-output circuit 42 to control the operation of the electronic device 10 , and may use the output data of the input-output circuit 42 to receive status information and other outputs from the electronic device 10 .

[0193] Furthermore, an embodiment of the present invention further provides a computer-readable storage medium, which may be disposed in the electronic device in the above embodiments. The computer-readable storage medium may be the above Figure 3 The memory in the storage and processing circuit 30 in the illustrated embodiment. The computer readable storage medium stores a computer program, which, when executed by the processor, implements the aforementioned Figure 1 The order prediction method described in the illustrated embodiment. Further, the computer storable medium may also be a U disk, a mobile hard disk, a read-only memory (ROM), a RAM, a magnetic disk or an optical disk, or other medium that can store program codes.

[0194] It should be noted that, for the convenience of description, the aforementioned method embodiments are all described as a series of action combinations, but those skilled in the art should be aware that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0195] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0196] The above is a description of the order prediction method, device and computer-readable storage medium provided by the present invention. For a person skilled in the art, according to the idea of ​​the embodiments of the present invention, there may be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. An order prediction method, characterized in that: The method comprises: Obtain historical order data from at least three hotel booking platforms; Extracting data to be analyzed from the historical order data, the data to be analyzed including order time, order time feature identifier, order quantity, order price, inventory quantity, order source, and scheduled lead time; Extracting features of multiple dimensions of the data to be analyzed and performing feature splicing to obtain a feature data set, wherein the features of the multiple dimensions include time dimension features, price dimension features, and inventory dimension features; Using the feature data set to train a pre-built order prediction model to obtain a pre-trained order prediction model; When an order prediction instruction is received, order prediction input data of the target hotel reservation platform is obtained according to the order prediction instruction, and the order prediction input data is input into the pre-trained order prediction model to obtain a real-time prediction result of the order of the target hotel reservation platform.

2. The method according to claim 1, characterized in that The method of using the feature data set to train a pre-built order prediction model to obtain a pre-trained order prediction model includes: Classifying the feature data set according to the category of the hotel reservation platform to obtain classified feature data sets of different categories; The classification feature data set of each category is used to train the pre-built order prediction model corresponding to each category to obtain pre-trained order prediction models of different categories.

3. The method according to claim 1, characterized in that After obtaining the real-time prediction result of the order of the target hotel reservation platform, the method further includes: Obtain the number of visits, clicks, and conversion rates of different hotel booking platforms in the preset time period before the current forecast time; Determine a correction factor based on the visits, clicks and conversion rates; The correction factor is used to correct the real-time order prediction result.

4. The method according to any one of claims 1 to 3, characterized in that After obtaining the real-time prediction result of the order of the target hotel reservation platform, the method further includes: The room price or inventory is adjusted according to the real-time prediction result of the order of the target hotel reservation platform and the real-time inventory of the target hotel reservation platform.

5. The method according to any one of claims 1 to 3, characterized in that After obtaining the real-time prediction result of the order of the target hotel reservation platform, the method further includes: Determine whether the real-time forecast result of the order is less than a preset standard threshold; If the real-time prediction result of the order is less than the preset standard threshold, determine whether there is a system failure in the hotel system; If it is determined that there is a fault in the hotel system, an early warning message is sent.

6. The method according to claim 5, characterized in that The determining whether there is a system failure in the hotel system includes: Obtain system logs and performance metrics for hotel systems; comparing the performance indicator with a performance threshold, and searching for target information in the system log; Determine whether the hotel system has a system failure based on the comparison result and the search result.

7. The method according to claim 5, characterized in that After determining whether the hotel system has a system failure according to the comparison result and the search result, the method further includes: If it is determined that the hotel system has a system failure, conduct an attribution analysis on the system failure; Generate attribution analysis reports of system failures.

8. An order prediction device, characterized in that: The device comprises: An acquisition module is used to obtain historical order data of at least three hotel booking platforms; A first extraction module is used to extract the data to be analyzed from the historical order data, wherein the data to be analyzed includes order time, order time feature identifier, order quantity, order price, inventory quantity, order source, and scheduled lead time; A second extraction module is used to extract features of multiple dimensions of the data to be analyzed and perform feature splicing to obtain a feature data set, wherein the features of the multiple dimensions include time dimension features, price dimension features, and inventory dimension features; A training module, used to train a pre-built order prediction model using the feature data set to obtain a pre-trained order prediction model; The prediction module is used to obtain the order prediction input data of the target hotel reservation platform according to the order prediction instruction when receiving the order prediction instruction, input the order prediction input data into the pre-trained order prediction model, and obtain the real-time prediction result of the order of the target hotel reservation platform.

9. An electronic device, characterized in that: The electronic device comprises: Memory and processor; The memory stores executable program code; The processor coupled to the memory calls the executable program code stored in the memory to execute the order forecasting method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the order prediction method according to any one of claims 1 to 7 is implemented.

Images