Hotel resource recommendation method, device, program product and medium

Abstract

This application provides a hotel resource recommendation method, device, program product, and medium, belonging to the field of travel resource management technology. This method can solve the problem of employees being unable to check in due to failures in hotel reservations using a fixed differential, thereby improving the success rate and reliability of hotel reservations. The method comprises: obtaining a first hotel dataset in the destination area, inputting this first hotel dataset into a trained isolation forest model to eliminate outlier hotels and filter out a clustered hotel dataset; then, adaptively determining a first differential based on the hotel prices in the clustered hotel dataset; and recommending hotels for the trip based on the first differential.

Description

Technical Field

[0001] The present application relates to the technical field of travel resource management, and in particular to a hotel resource recommendation method, device, program product, and medium. Background Art

[0002] At present, business travel often uses a larger geographical area (such as YY City) where the destination (such as XX Building) is located as the difference control area. This difference control area often corresponds to a fixed travel standard (hereinafter referred to as the difference standard). The price of the booked hotel must not exceed this difference standard to control travel costs.

[0003] However, hotel prices in a smaller area of ​​the destination (e.g., the block where Building XX is located) may generally be higher than the standard. For example, hotel prices in the destination may generally rise due to factors such as tourism and conferences, or current hotel prices in the destination may generally be higher than the standard previously determined. This may result in a lack of hotels near the destination that meet the standard, leading to failed hotel reservations and employees being unable to check in. This indicates that existing hotel reservation methods are unreliable. Summary of the Invention

[0004] The embodiments of the present application provide a hotel resource recommendation method, device, program product, and medium, which can solve the problem of employees being unable to check in due to failure to book a hotel using a fixed difference, thereby improving the success rate and reliability of hotel reservations.

[0005] To achieve the above objectives, the present invention provides the following technical solutions:

[0006] In a first aspect, a hotel resource recommendation method is provided, which is applicable to a server and includes:

[0007] Receive a hotel reservation request from a terminal, the hotel reservation request carrying the destination of this business trip;

[0008] Obtain a first hotel dataset, where the first hotel dataset includes one or more of the following features of hotels in the destination area: location, price, travel distance, travel duration, or type;

[0009] The first hotel dataset is input into the isolation forest model for data screening to obtain the sociable hotel dataset. The isolation forest model is trained based on the hotel sample dataset, which is the same type of dataset as the first hotel dataset.

[0010] If the number of hotels in the group hotel dataset is greater than or equal to the quantity threshold, then determine the first difference standard based on the price of each hotel in the group hotel dataset;

[0011] The hotels in the Hequn Hotel dataset whose prices are less than or equal to the first difference standard are determined as recommended hotels, and the recommended hotels are used to book hotels for this business trip;

[0012] A hotel resource recommendation message is sent to the terminal, where the hotel resource recommendation message carries a list of recommended hotels.

[0013] In one possible design, the first hotel dataset is input into the isolation forest model for data screening to obtain a group hotel dataset, including:

[0014] The first hotel dataset is input into the isolation forest model for data screening to obtain the second hotel dataset;

[0015] If the number of hotels in the second hotel dataset is greater than or equal to the number threshold, the second hotel dataset is determined to be a group hotel dataset.

[0016] Optionally, the first hotel dataset is input into the isolation forest model for data screening to obtain a second hotel dataset, including:

[0017] Input the first hotel data set into the isolation forest model, traverse each binary tree in the isolation forest model, and determine the average height of each hotel data in the first hotel data set in the isolation forest model;

[0018] Hotel data with an average height greater than or equal to a first height threshold are screened out from the first hotel data set to construct a second hotel data set.

[0019] Furthermore, the first hotel dataset is input into the isolation forest model for data screening to obtain a second hotel dataset, which also includes:

[0020] If the number of hotel data with an average height greater than or equal to the first height threshold in the first hotel data set is less than the number threshold, hotel data with an average height greater than or equal to the second height threshold are filtered out from the first hotel data set to construct a second hotel data set, and the second height threshold is less than the first height threshold.

[0021] Optionally, the first hotel dataset is input into the isolation forest model for data screening to obtain a second hotel dataset, including:

[0022] Input the first hotel data set into the isolation forest model, traverse each binary tree in the isolation forest model, and determine the average height of each hotel data in the first hotel data set in the isolation forest model;

[0023] In descending order of average height, hotel data with a number greater than or equal to a threshold value are filtered out from the first hotel data set to construct the second hotel data.

[0024] In one possible design, the first difference standard is determined based on the prices of each hotel in the group hotel dataset, including:

[0025] Determine the second difference standard based on the prices of each hotel in the Hequn Hotel Dataset. The second difference standard is the average or median of the lowest prices of each hotel in the Hequn Hotel Dataset.

[0026] The first difference mark is determined according to the second difference mark and a difference mark threshold, where the difference mark threshold corresponds to the destination.

[0027] Optionally, determining the first difference standard according to the second difference standard and a difference standard threshold includes:

[0028] If the second difference mark is less than or equal to the difference mark threshold, the second difference mark is determined as the first difference mark.

[0029] Optionally, determining the first difference standard according to the second difference standard and a difference standard threshold includes:

[0030] If the second difference standard is greater than the difference standard threshold, and there is a hotel in the group hotel set whose price is less than or equal to the difference standard threshold, the difference standard threshold is determined as the first difference standard.

[0031] Optionally, determining the first difference standard according to the second difference standard and a difference standard threshold includes:

[0032] If the price of each hotel in the group hotel set is greater than the difference threshold, the second difference is determined as the first difference.

[0033] Further, determining the second difference standard as the first difference standard includes:

[0034] If the deviation between the second difference standard and the difference standard threshold is less than or equal to the difference standard deviation threshold, the second difference standard is determined as the first difference standard.

[0035] In one possible design, obtaining the first hotel dataset includes:

[0036] Hotel data in the area of ​​the destination is filtered according to a first filtering rule to generate a first hotel data set. The first filtering rule includes one or more of the following: information about a first geographical area of ​​the destination, a first hotel price range, a first distance or a first travel time between the hotel and the destination, and a first hotel type.

[0037] Optionally, the method further includes:

[0038] If the number of hotels in the group hotel dataset is less than the quantity threshold, and / or the prices of the hotels in the group hotel dataset are all greater than the difference threshold, then filtering the hotel data in the area where the destination is located according to a second filtering rule to generate a first hotel dataset, where the second filtering rule includes one or more of the following: information about a second geographical area where the destination is located, a second hotel price range, a second distance or a second travel time between the hotel and the destination, and a second hotel type;

[0039] wherein the second geographical area includes the first geographical area and is larger than the first geographical area;

[0040] The second hotel price range includes the first hotel price range and is greater than the first hotel price range;

[0041] The second distance is greater than the first distance;

[0042] The second passage time is longer than the first passage time;

[0043] The number of types of the second hotel type is greater than the number of types of the first hotel type.

[0044] Optionally, the hotel resource recommendation message further carries difference calculation information, and the difference calculation information is used to record the calculation method and result of the first difference.

[0045] Furthermore, the method further comprises:

[0046] Receive a hotel reservation confirmation message from the terminal, the hotel reservation confirmation message carrying reservation information for a target hotel from the recommended hotels, the target hotel being used to provide services for this business trip;

[0047] Send a hotel reservation completion message to the terminal, which carries the reservation success information of the target hotel.

[0048] Optionally, the difference calculation information and booking success information are used for travel reimbursement.

[0049] In a second aspect, a hotel resource recommendation device is provided. The device may be a server or other device that can execute the hotel resource recommendation method described in the first aspect. The device includes: a processing module and a transceiver module; wherein:

[0050] The transceiver module is used to receive a hotel reservation request from a terminal, where the hotel reservation request carries the destination of the current business trip;

[0051] a processing module configured to obtain a first hotel dataset, the first hotel dataset including one or more of the following characteristics of hotels in an area of ​​a destination: location, price, travel distance, travel duration, or type;

[0052] The processing module is further configured to input the first hotel dataset into the isolation forest model for data screening to obtain a sociable hotel dataset, wherein the isolation forest model is trained based on the hotel sample dataset, and the hotel sample dataset and the first hotel dataset are of the same type;

[0053] The processing module is further configured to determine a first difference standard based on the price of each hotel in the group hotel data set if the number of hotels in the group hotel data set is greater than or equal to the number threshold;

[0054] The processing module is further configured to determine hotels in the group hotel data set whose prices are less than or equal to the first difference standard as recommended hotels, and the recommended hotels are used to book hotels for this business trip;

[0055] The transceiver module is further configured to send a hotel resource recommendation message to the terminal, wherein the hotel resource recommendation message carries a list of recommended hotels.

[0056] In one possible design solution, the processing module is further configured to:

[0057] The first hotel dataset is input into the isolation forest model for data screening to obtain the second hotel dataset;

[0058] If the number of hotels in the second hotel dataset is greater than or equal to the number threshold, the second hotel dataset is determined to be a group hotel dataset.

[0059] Optionally, the processing module is further configured to:

[0060] Input the first hotel data set into the isolation forest model, traverse each binary tree in the isolation forest model, and determine the average height of each hotel data in the first hotel data set in the isolation forest model;

[0061] Hotel data with an average height greater than or equal to a first height threshold are screened out from the first hotel data set to construct a second hotel data set.

[0062] Furthermore, the processing module is also used to filter out hotel data with an average height greater than or equal to a second height threshold from the first hotel data set to construct a second hotel data set if the number of hotel data with an average height greater than or equal to a first height threshold in the first hotel data set is less than a quantity threshold, and the second height threshold is less than the first height threshold.

[0063] Optionally, the processing module is further configured to:

[0064] Input the first hotel data set into the isolation forest model, traverse each binary tree in the isolation forest model, and determine the average height of each hotel data in the first hotel data set in the isolation forest model;

[0065] In descending order of average height, hotel data with a number greater than or equal to a threshold value are filtered out from the first hotel data set to construct the second hotel data.

[0066] In one possible design solution, the processing module is further configured to:

[0067] Determine the second difference standard based on the prices of each hotel in the Hequn Hotel Dataset. The second difference standard is the average or median of the lowest prices of each hotel in the Hequn Hotel Dataset.

[0068] The first difference mark is determined according to the second difference mark and a difference mark threshold, where the difference mark threshold corresponds to the destination.

[0069] Optionally, the processing module is further configured to determine the second difference mark as the first difference mark if the second difference mark is less than or equal to the difference mark threshold.

[0070] Optionally, the processing module is further configured to determine the difference threshold as the first difference if the second difference is greater than the difference threshold and there is a hotel in the group hotel set whose price is less than or equal to the difference threshold.

[0071] Optionally, the processing module is further configured to determine the second difference standard as the first difference standard if the price of each hotel in the group hotel set is greater than the difference standard threshold.

[0072] Furthermore, the processing module is further configured to determine the second difference mark as the first difference mark if a deviation between the first difference mark and the difference mark threshold is less than or equal to the difference mark deviation threshold.

[0073] In one possible design scheme, the processing module is also used to filter hotel data in the area where the destination is located according to a first filtering rule to generate a first hotel data set. The first filtering rule includes one or more of the following: information about the first geographical area where the destination is located, a first hotel price range, a first distance or a first travel time between the hotel and the destination, and a first hotel type.

[0074] Optionally, the processing module is further configured to, if the number of hotels in the group hotel data set is less than a quantity threshold, and / or the prices of the hotels in the group hotel data set are all greater than a difference threshold, filter the hotel data in the area where the destination is located according to a second filtering rule to generate a first hotel data set, where the second filtering rule includes one or more of the following: information about a second geographical area where the destination is located, a second hotel price range, a second distance or a second travel time between the hotel and the destination, and a second hotel type;

[0075] wherein the second geographical area includes the first geographical area and is larger than the first geographical area;

[0076] The second hotel price range includes the first hotel price range and is greater than the first hotel price range;

[0077] The second distance is greater than the first distance;

[0078] The second passage time is longer than the first passage time;

[0079] The number of types of the second hotel type is greater than the number of types of the first hotel type.

[0080] Optionally, the hotel resource recommendation message further carries difference calculation information, and the difference calculation information is used to record the calculation method and result of the first difference.

[0081] Furthermore, the transceiver module is also used to:

[0082] Receive a hotel reservation confirmation message from the terminal, the hotel reservation confirmation message carrying reservation information for a target hotel from the recommended hotels, the target hotel being used to provide services for this business trip;

[0083] Send a hotel reservation completion message to the terminal, which carries the reservation success information of the target hotel.

[0084] Optionally, the difference calculation information and booking success information are used for travel reimbursement.

[0085] In a third aspect, an electronic device is provided, comprising: a processor coupled to a memory;

[0086] A processor is configured to execute a computer program stored in a memory so that the device executes the method as described in any implementation manner of the first aspect.

[0087] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a program or instruction, and when the program or instruction is run on a computer, the computer executes the method described in any implementation manner of the first aspect.

[0088] In a fifth aspect, a computer program product is provided, the computer program product comprising: a computer program code, which, when executed on a computer, enables the computer to execute the method as described in any implementation manner in the first aspect.

[0089] Based on the technical solution provided in the embodiment of the present application, the server can obtain a first hotel dataset in the area where the destination is located, and input the first hotel dataset into a trained isolation forest model to eliminate outlier hotels and screen out a clustered hotel dataset. Then, the server adaptively determines a first difference standard based on the hotel prices in the clustered hotel dataset, and recommends hotels for this business trip based on the first difference standard. This can solve the problem of hotel reservation failures and employees being unable to check in because the prices of destination hotels are generally higher than the fixed difference standard, thereby improving the success rate and reliability of hotel reservations.

[0090] Specifically, the first hotel data set can be input into the isolation forest model to traverse all binary trees to obtain the average height of each hotel data in the first hotel data set, and then the hotel data with an average height greater than or equal to the first height threshold or the second height threshold are combined into a second hotel data set. Then, the second hotel data set containing a number of hotel data greater than or equal to the number threshold is used as a clustered hotel data set to eliminate outlier hotel data and ensure that the clustered hotel data set has hotel data that meets the requirements and is sufficient in number for selection, further improving the success rate and reliability of hotel reservations.

[0091] Optionally, instead of using a height threshold, a sufficient number of hotel data may be directly selected from the first hotel dataset in descending order of average height to construct the second hotel data and the group hotel data, which can further improve the success rate and reliability of hotel reservations.

[0092] Specifically, the average or median price of each hotel in the group hotel dataset can be used as a dynamic differential standard (second differential standard), and then based on the comparison result of the dynamic differential standard and the fixed differential standard (difference standard threshold), the first differential standard applicable to this trip can be adaptively determined, and hotels can be recommended or booked based on the first differential standard to improve the success rate and reliability of hotel reservations, while taking into account travel costs as much as possible.

[0093] Optionally, when the second difference (dynamic difference) is less than or equal to the difference threshold (fixed difference), it indicates that the consumption level of the destination is lower than or equal to the fixed difference, and it can be ensured that a hotel can be booked. In this scenario, a dynamic difference (second difference) that is not higher than the fixed difference can also be used to book a hotel to save travel costs as much as possible.

[0094] Optionally, when the second difference is greater than the difference threshold, and there are hotels in the group hotel set whose prices are less than or equal to the difference threshold, it indicates that although the destination consumption level is higher than or equal to the fixed difference, there are still hotels with prices lower than the fixed difference, and it is guaranteed that the hotel can be booked. In this scenario, the fixed difference (difference threshold) can be used to book the hotel to save travel costs as much as possible.

[0095] Optionally, when the price of each hotel in the group hotel set is greater than the difference threshold, it indicates that the consumption level of the destination is generally higher than the fixed difference, and it is impossible to book a hotel using the fixed difference. In this scenario, you can consider using a dynamic difference (second difference) to book a hotel, that is, give priority to the destination consumption level to book a hotel, so as to ensure the success rate and reliability of hotel reservations. BRIEF DESCRIPTION OF THE DRAWINGS

[0096] Figure 1 A schematic diagram of the architecture of the travel management system provided in an embodiment of the present application;

[0097] Figure 2 A flowchart of a hotel resource recommendation method provided in an embodiment of the present application;

[0098] Figure 3 A schematic diagram of a process for filtering a first hotel dataset based on an isolation forest model provided in an embodiment of the present application;

[0099] Figure 4 A schematic diagram of another process for screening a first hotel dataset based on an isolation forest model provided in an embodiment of the present application;

[0100] Figure 5 A schematic diagram of another process for screening a first hotel dataset based on an isolation forest model provided in an embodiment of the present application;

[0101] Figure 6 A schematic diagram of another process for screening a first hotel dataset based on an isolation forest model provided in an embodiment of the present application;

[0102] Figure 7 A schematic diagram of a process for determining a first difference based on the prices of each hotel in a group hotel dataset provided in an embodiment of the present application;

[0103] Figure 8 A schematic diagram of another process for determining a first difference based on the prices of each hotel in a group hotel dataset provided in an embodiment of the present application;

[0104] Figure 9 A schematic diagram of another process for determining a first difference based on the prices of each hotel in the group hotel data set provided in an embodiment of the present application;

[0105] Figure 10 A schematic diagram of another process for determining a first difference based on the prices of each hotel in the group hotel data set provided in an embodiment of the present application;

[0106] Figure 11 A schematic diagram of another process for determining a first difference based on the prices of each hotel in the group hotel data set provided in an embodiment of the present application;

[0107] Figure 12 A schematic diagram of the structure of a hotel resource recommendation device provided in an embodiment of the present application;

[0108] Figure 13 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0109] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the present application will be described below in conjunction with the accompanying drawings.

[0110] For example, Figure 1 This is a schematic diagram of the architecture of the travel management system provided in this application embodiment. Figure 1 As shown, the system includes a terminal and a server.

[0111] The server is located on the network side of the travel management system, and the terminal device communicates with the server via wireless or wired communication. The server may also be referred to as a cloud device or network device, and may be a mainframe computer, computing device, or other device capable of processing travel services, but this embodiment of the application is not limited thereto.

[0112] The terminal can also be called a terminal device, user equipment (UE), mobile station, mobile terminal, client, etc. The terminal device can be a mobile phone, tablet computer, Pad, personal computer, etc. The embodiment of the present application does not limit the specific technology and specific device form adopted by the terminal device. It should be noted that the hotel resource recommendation method provided in the embodiment of the present application can be used for Figure 1 Between the terminal and the server shown.

[0113] After receiving a hotel reservation request containing a destination from a terminal, the server can obtain a first hotel dataset in the area where the destination is located, and input the first hotel dataset into the trained isolation forest model to eliminate outlier hotels and screen out a clustered hotel dataset. Then, the server adaptively determines a first difference standard based on the hotel prices in the clustered hotel dataset, and recommends hotels for this business trip based on the first difference standard. This can solve the problem of hotel reservation failures and employees being unable to check in because the prices of destination hotels are generally higher than the fixed difference standard, thereby improving the success rate and reliability of hotel reservations.

[0114] Those skilled in the art will understand that Figure 1 The schematic diagram is merely a simplified example for ease of understanding. The architecture shown does not constitute any limitation on the travel management system. An actual travel management system may include more or fewer devices than shown in the diagram. For example, it may also include relay devices such as base stations that establish communication connections between terminals and servers, and / or other terminals. Figure 1 Not drawn in.

[0115] The following combination Figure 2-Figure 11 The hotel resource recommendation method provided in the embodiment of the present application is specifically described.

[0116] For example, Figure 2 A flowchart of a hotel resource recommendation method provided in an embodiment of the present application, which is applicable to Figure 1 The server shown in . Figure 2 As shown, the method includes:

[0117] S1, the server receives a hotel reservation request from a terminal.

[0118] Among them, the hotel reservation request carries the destination of this business trip, which can be the address of the customer where the employee is on a business trip, the meeting address, etc. For example, it can be the customer's office location, the xx building hosting the meeting, etc. That is, the destination referred to in the embodiment of this application is usually a specific location, and its geographical scope is usually much smaller than the destination referred to in the travel standard, such as yy city.

[0119] Optionally, the hotel reservation request may also carry other information, such as the name, gender, start and end dates of the trip, the location of the hotel desired to stay in, price, room type (single room, double room, suite, etc.), the acceptable round-trip distance between the hotel and the destination, the travel time, etc., as well as whether the hotel needs to provide personalized services such as Internet access and breakfast, etc., which are not limited in this embodiment of the present application.

[0120] S2: The server obtains the first hotel dataset.

[0121] Among them, the area where the destination is located is usually a smaller geographical area that includes the location of the destination, such as a geographical area within 2 kilometers around the destination. The hotels in the area where the destination is located are various types of business entities that can provide accommodation services in the geographical area, such as star hotels, express hotels, guesthouses, farmhouses, couples hotels, etc., which are not limited in this embodiment of the present application.

[0122] The first hotel dataset includes one or more of the following characteristics for hotels in the destination area: location, price, travel distance, travel duration, or type. Location can be latitude, longitude, or geodetic coordinates. Price can be the unit price of a hotel's accommodation service, such as 500 yuan / day or 300 yuan / day, and can have multiple values ​​depending on room type. Travel distance can be the straight-line distance or travel distance between the hotel and the destination. Travel duration can be the commuting time between the hotel and the destination. Both travel distance and travel duration can be associated with travel mode (e.g., walking, bicycle, driving, bus, subway, etc.). This hotel data can be obtained through online searches and stored on a server for backup.

[0123] In one possible design, S2, the server obtains the first hotel dataset, including:

[0124] The server filters the hotel data in the area where the destination is located according to the first filtering rule to generate a first hotel data set.

[0125] The first filtering rule includes one or more of the following: information about the first geographical area where the destination is located, a first hotel price range, a first distance or a first travel time between the hotel and the destination, and a first hotel type.

[0126] Among them, the first geographical area can be a smaller geographical area around the destination, such as an area within 2 kilometers around the destination. The first hotel price range can be the price range of hotels in the first geographical area that are acceptable for this business trip, such as 400 yuan / day to 1,000 yuan / day for a double room and 300 yuan / day to 600 yuan / day for a single room. The first distance can be the straight-line distance between the hotel and the destination, or the travel distance corresponding to a certain mode of transportation. The first duration can be the commuting time between the hotel and the destination, such as the commuting time corresponding to a certain mode of transportation. The first hotel type can be the hotel type acceptable for this business trip, such as single rooms and double rooms are acceptable, but suites are not acceptable.

[0127] Specifically, the server can filter hotel data within a specified range (first geographic area) around the destination based on the first filtering rule through network search or other technical means to generate a first hotel data set. One hotel data set can be generated for each hotel, or multiple hotel data sets can be generated, which is not limited in this embodiment.

[0128] In addition, in the process of obtaining the first hotel data set, data preprocessing can be performed on the obtained hotel data, such as missing value processing, outlier processing, deduplication processing or noise data processing, format conversion, etc., to improve the accuracy and reliability of the data, facilitate subsequent processing, and improve efficiency.

[0129] It should be noted that data preprocessing typically only removes obvious or easily tractable anomalies. To further improve the quality of the first hotel dataset and ensure the accuracy of subsequent model training, the first hotel dataset can be iteratively cleaned using the isolation forest model to thoroughly remove any anomalies contained therein, i.e., perform S3 below. The isolation forest model is a fast anomaly detection method with linear time complexity and high accuracy.

[0130] S3: The server inputs the first hotel dataset into the isolation forest model for data screening to obtain a group hotel dataset.

[0131] The isolation forest model is trained based on a hotel sample dataset. The hotel sample dataset and the first hotel dataset are of the same type and are typically constructed based on historical hotel data. This historical hotel data may include, but is not limited to, hotel data for a larger geographic area (e.g., City YY) within the destination. For example, if both the hotel sample data and the first hotel data include the attribute (feature) of hotel price, and the destination is City A, and the historical hotel data is hotel data for City B, then the historical hotel data for City B can be normalized, such as price of a hotel in City B / average hotel price in City B. The isolation forest model can then be trained using the historical hotel data for City B containing normalized hotel prices to improve the generalization capability of the isolation forest model. Accordingly, the price of each hotel in the destination area in the first hotel dataset can be converted to a normalized price (e.g., price of the destination hotel / average hotel price in City A). This normalized first hotel dataset can then be input into the trained isolation forest model for data screening. Alternatively, normalization can be performed upon obtaining the first hotel dataset, and then converted back to actual prices to facilitate booking after the recommended hotels are determined in S5 below.

[0132] In some embodiments, using the isolation forest algorithm to screen the first hotel dataset mainly includes two processes: training an isolation forest model based on the hotel sample dataset and screening the first hotel dataset.

[0133] Specifically, training the hotel sample dataset usually includes the following steps:

[0134] Construct an initial isolation forest model;

[0135] The acquired historical hotel dataset is input into the initial isolation forest model for training to generate an isolation forest model (hereinafter referred to as isolation forest model 1) to complete the training process of the standard dataset.

[0136] For example, Ψ sample points can be randomly selected from the hotel sample dataset as a subsample set and placed at the root node of the tree. A random feature (e.g., selected from hotel price, distance, travel time, or hotel type number) is then randomly assigned. A cut point p is randomly generated in the current node data (the cut point is between the maximum and minimum values ​​of the specified feature at the current node). This cut point generates a hyperplane, which partitions the current node data space into two subspaces: hotel sample data with a value of the specified feature less than the cut point p is placed in the left child of the current node, and hotel sample data with a value greater than or equal to the cut point p is placed in the right child of the current node. Then, the above method is recursively performed on the left and right child nodes, continuously constructing new left and right child nodes until all child nodes contain only one hotel sample (no further cuts are possible) or the child node reaches its maximum height. This completes the construction of a binary tree. This process is repeated until the number of constructed binary trees reaches the specified number, thus completing the training process of the isolation forest model.

[0137] Furthermore, in the embodiment of the present application, before generating the isolation forest model, the parameters of the isolation forest model need to be set. For example, the maximum height of the binary tree can be 50% of the number of samples in the subsample set, and the number of binary trees in the isolation forest model can be twice the number of features.

[0138] Then, the isolation forest model 1 is used to perform anomaly detection on the first hotel dataset. That is, the first hotel dataset is input into the isolation forest model 1 for data screening, and a part of the abnormal data (outlier hotel data) is eliminated to obtain the group hotel dataset.

[0139] The following combination Figure 3-Figure 11 Describe the data screening process in detail.

[0140] For example, Figure 3 A flowchart of a method for filtering a first hotel dataset based on an isolation forest model is provided in an embodiment of the present application. Figure 3 As shown in S3, the server inputs the first hotel dataset into the isolation forest model for data screening, and obtains the group hotel dataset, including S31-S32:

[0141] S31, the server inputs the first hotel dataset into the isolation forest model for data screening to obtain a second hotel dataset.

[0142] Specifically, each hotel data point in the first hotel dataset is input into the isolation forest model, and each binary tree in the isolation forest model is traversed. Next, the layer on which each hotel data point in the first hotel dataset ultimately falls within each binary tree is calculated, yielding the average height of each hotel data point in the isolation forest model. Finally, based on the average height of each hotel data point in the first hotel dataset, a portion of hotel data points (outlier hotel data points) are removed, and the remaining hotel data points form the second hotel dataset.

[0143] One possible design solution is to combine Figure 3 ,like Figure 4 As shown, in S31, the server inputs the first hotel dataset into the isolation forest model for data screening to obtain the second hotel dataset, including S311-S312:

[0144] S311, the server inputs the first hotel data set into the isolation forest model, traverses each binary tree in the isolation forest model, and determines the average height of each hotel data in the first hotel data set in the isolation forest model.

[0145] S312: The server selects hotel data with an average height greater than or equal to a first height threshold from the first hotel data set to construct a second hotel data set.

[0146] Specifically, hotel data with an average height less than a first height threshold (outlier hotel data) may be removed from the first hotel data set, and the remaining hotel data may be used to generate the second hotel data set.

[0147] Among them, the first height threshold can be 30%, 20%, etc. of the maximum value of the average height of all hotel data in the first hotel data set, or it can be the average or median value of the average height of all hotel data in the first hotel data set, or it can also be a preset value, which is not limited in the embodiment of the present application.

[0148] Further, combined with Figure 4 ,like Figure 5 As shown, in S312, the server filters out hotel data with an average height greater than or equal to a first height threshold from the first hotel data set to construct a second hotel data set, including S3121:

[0149] S3121: If the number of hotel data with an average height greater than or equal to the first height threshold in the first hotel data set is less than the number threshold, the server filters out hotel data with an average height greater than or equal to the second height threshold from the first hotel data set to construct a second hotel data set.

[0150] The second height threshold is smaller than the first height threshold so as to relax the requirement and screen out more hotel data for selection and booking, so as to further improve the reliability of hotel booking.

[0151] Another possible design solution is to combine Figure 3 ,like Figure 6 As shown, in S31, the server inputs the first hotel dataset into the isolation forest model for data screening to obtain the second hotel dataset, including S311 and S314:

[0152] S311, the server inputs the first hotel data set into the isolation forest model, traverses each binary tree in the isolation forest model, and determines the average height of each hotel data in the first hotel data set in the isolation forest model.

[0153] S314 , the server filters out hotel data whose number is greater than or equal to a threshold value from the first hotel data set in descending order of average height to construct a second hotel data set.

[0154] Compared with the screening scheme in S312-S313, in S314, as long as the number of hotel data in the first hotel data set is greater than the quantity threshold, it can be ensured that a sufficient number of hotel data are screened out to construct the second hotel data set, which can further improve the reliability of hotel reservations.

[0155] It should be noted that the screening scheme in S312-S313 and the screening scheme in S314 can also be implemented in combination. For example, the number of hotel data must meet the quantity threshold requirement, and the average height of the screened hotel data must also meet the first height threshold or the second height threshold requirement. This embodiment of the present application is not limited.

[0156] S32: If the number of hotels in the second hotel dataset is greater than or equal to the number threshold, the server determines the second hotel dataset as a group hotel dataset.

[0157] Furthermore, to ensure the cleanliness of the obtained clustered hotel dataset, an iterative training-filtering process can be performed on the first hotel dataset. For example, the clustered hotel dataset, which is obtained after filtering the first hotel dataset using Isolation Forest Model 1, can be re-input into Isolation Forest Model 1 for iterative training to obtain another Isolation Forest Model (Isolation Forest Model 2). Isolation Forest Model 2 can then be used to filter the first hotel dataset again to obtain another clustered hotel dataset. This training-filtering process can be repeated one or more times until the number of clustered hotel datasets meets a set requirement. For example, the set requirement may be that the number of hotels in the clustered hotel dataset must be greater than or equal to a hotel threshold. In practical applications, the number of training-filtering iterations depends on the trade-off between data cleanliness and information loss. A greater number of iterations results in relatively cleaner hotel data in the clustered hotel dataset, but this may also result in the loss of more useful information.

[0158] It should be noted that to reduce the number of splits, facilitate rapid convergence, and improve training and screening efficiency, parameters such as the number of binary trees and maximum height in the isolation forest model can be adjusted. For example, the maximum height of the binary tree in the isolation forest model can be a fraction of the number of hotel data in the historical hotel sample dataset or the first hotel dataset, such as 50%, 25%, etc.

[0159] S4. If the number of hotels in the group hotel data set is greater than or equal to the number threshold, the server determines a first difference standard according to the price of each hotel in the group hotel data set.

[0160] One possible design solution is to combine Figure 2 ,like Figure 7 As shown, in S4, the server determines the first difference standard based on the prices of each hotel in the group hotel dataset, including:

[0161] S41, the server determines a second difference mark based on the prices of the hotels in the group hotel dataset, where the second difference mark is the average or median of the lowest prices of each hotel in the group hotel dataset;

[0162] S42: The server determines a first difference mark according to the second difference mark and a difference mark threshold, where the difference mark threshold corresponds to the destination.

[0163] Among them, the second differential is a differential that is dynamically determined based on the hotel prices at the destination, which can truly reflect the consumption level of the destination. The first differential determined by comprehensively considering the second differential and the fixed differential (difference threshold) can not only truly reflect the consumption level of the destination, but also take into account the travel cost control needs, thus having greater flexibility in the hotel booking process.

[0164] Specifically, for the following three scenarios, different methods can be used to determine the first difference standard:

[0165] Scenario 1, combined Figure 7 ,like Figure 8 As shown, in S42, the server determines the first difference mark according to the second difference mark and the difference mark threshold, including:

[0166] S421: If the second difference mark is less than or equal to the difference mark threshold, the server determines the second difference mark as the first difference mark.

[0167] Among them, the second difference standard (dynamic difference standard) is less than or equal to the difference standard threshold (fixed difference standard), indicating that the consumption level of the destination is lower than or equal to the fixed difference standard, which can ensure that the hotel can be booked. In this scenario, you can also use the dynamic difference standard (second difference standard) that is not higher than the fixed difference standard to book the hotel to save travel costs as much as possible.

[0168] Scenario 2, combined Figure 7 ,like Figure 9 As shown, in S42, the server determines the first difference mark according to the second difference mark and the difference mark threshold, including:

[0169] S422: If the second difference is greater than the difference threshold, and there is a hotel in the group hotel set whose price is less than or equal to the difference threshold, the server determines the difference threshold as the first difference.

[0170] Among them, although the second difference standard (dynamic difference standard) is greater than the difference standard threshold (fixed difference standard), there are hotels in the group hotel set with prices less than or equal to the difference standard threshold. This shows that although the destination consumption level is higher than or equal to the fixed difference standard, there are still hotels with prices lower than the fixed difference standard. It is guaranteed that the hotel can be booked. In this scenario, the fixed difference standard (difference standard threshold) can be used to book the hotel to save travel costs as much as possible.

[0171] Optionally, combined Figure 7 ,like Figure 10 As shown, in S42, the server determines the first difference mark according to the second difference mark and the difference mark threshold, including:

[0172] S423: If the price of each hotel in the group hotel set is greater than the difference threshold, the server determines the second difference as the first difference.

[0173] Among them, the price of each hotel in the group hotel set is greater than the difference threshold (it is easy to understand that the second difference will also be greater than the difference threshold at this time), indicating that the consumption level of the destination is generally higher than the fixed difference. It is impossible to book a hotel using the fixed difference. In this scenario, you can consider using a dynamic difference (the second difference) to book a hotel, that is, give priority to the destination consumption level to book a hotel, so as to ensure the success rate and reliability of hotel reservations.

[0174] Further, combined with Figure 10 ,like Figure 11 As shown, S423, if the price of each hotel in the group hotel set is greater than the difference threshold, the server determines the second difference as the first difference, including:

[0175] S4231: If the price of each hotel in the group hotel set is greater than the difference threshold, and the deviation between the second difference and the difference threshold is less than or equal to the difference deviation threshold, the server determines the second difference as the first difference.

[0176] The standard deviation threshold can be a specific value, such as the hotel price increase does not exceed 100 yuan / day, or a percentage, such as the hotel price increase does not exceed 20%*fixed standard deviation.

[0177] To prevent excessive costs from being incurred due to the second differential being significantly greater than the fixed differential, the increase in travel costs can also be controlled. On the one hand, if the increase in the second differential compared to the differential threshold is within an acceptable range (e.g., less than or equal to the differential deviation threshold), the remaining hotel reservation process can be automatically executed, thereby improving hotel booking efficiency. On the other hand, if the increase in the second differential compared to the differential threshold exceeds the acceptable range (e.g., greater than the differential deviation threshold), the remaining hotel reservation process is halted and an alarm is output. This alarm serves to alert travel standard decision-makers, such as the company's general manager and other management personnel, to determine a new fixed differential.

[0178] S5. The server determines the hotels in the group hotel data set whose prices are less than or equal to the first difference standard as recommended hotels. The recommended hotels are used to book hotels for this business trip.

[0179] S6. The server sends a hotel resource recommendation message to the terminal. The hotel resource recommendation message carries a list of recommended hotels.

[0180] Specifically, the list of recommended hotels can be sent to the terminal and displayed on a display screen of the terminal. The employee can determine a hotel to book based on the list of recommended hotels on a hotel reservation interface provided by the terminal.

[0181] Optionally, the method further includes:

[0182] If the number of hotels in the group hotel dataset is less than the quantity threshold, and / or the prices of the hotels in the group hotel dataset are all greater than the difference threshold, then filtering the hotel data in the area where the destination is located according to a second filtering rule to generate a first hotel dataset, where the second filtering rule includes one or more of the following: information about a second geographical area where the destination is located, a second hotel price range, a second distance or a second travel time between the hotel and the destination, and a second hotel type;

[0183] wherein the second geographical area includes the first geographical area and is larger than the first geographical area;

[0184] The second hotel price range includes the first hotel price range and is greater than the first hotel price range;

[0185] The second distance is greater than the first distance;

[0186] The second passage time is longer than the first passage time;

[0187] The number of types of the second hotel type is greater than the number of types of the first hotel type.

[0188] It can be seen that the second filtering rule is looser than the first filtering rule, and more hotel data can be filtered out from the hotel data in the destination area to construct the first hotel data, so that the server can filter out more hotel data from the first hotel data to construct the group hotel data, thereby improving the success rate and reliability of hotel reservations.

[0189] Optionally, the hotel resource recommendation message further carries difference calculation information, and the difference calculation information is used to record the calculation method and result of the first difference.

[0190] For further information, please refer to Figure 2 , the method further comprises:

[0191] S7, the server receives a hotel reservation confirmation message from the terminal, the hotel reservation confirmation message carries reservation information of a target hotel from the recommended hotels, and the target hotel is used to provide services for this business trip.

[0192] S8, the server sends a hotel reservation completion message to the terminal, and the hotel reservation completion message carries the reservation success information of the target hotel.

[0193] It should be noted that the interaction process of S1-S8 can provide users (such as employees on business trips) with a complete process from submitting a hotel reservation request to completing a hotel reservation, thereby providing a full-process closed-loop service for the hotel reservation process and improving the user experience.

[0194] Optionally, the difference calculation information and booking success information are used for travel reimbursement to improve the transparency of travel information and facilitate travel cost control and financial accounting.

[0195] It should be noted that the hotel resource recommendation method provided in the embodiment of the present application is not only applicable to scenarios where hotel reservation services are provided to employees in business travel scenarios, but can also be applied to other scenarios such as tourism where hotel reservations are required, and the embodiment of the present application is not limited thereto. For example, in a tourism scenario, the destination can be a tourist attraction, and the fixed difference standard (difference standard threshold) can be understood as the hotel accommodation budget in the area where the tourist attraction is located. The first filtering rule and the second filtering rule can be determined based on the personal preferences and economic conditions of the tourist. The second difference standard can be the current hotel price calculated in real time based on the prices of various hotels in the area where the tourist attraction is located (such as within a radius of 2 kilometers), and then the current hotel price is compared with the travel budget to determine the hotel price (first difference standard) acceptable to the tourist, and hotels are recommended to the tourist based on the acceptable price. The tourist can book a hotel from the list of recommended hotels.

[0196] Based on the hotel resource recommendation method provided in the embodiment of the present application, the server can obtain a first hotel dataset in the area where the destination is located, and input the first hotel dataset into a trained isolation forest model to eliminate outlier hotels and screen out a clustered hotel dataset. Then, the server adaptively determines a first difference standard based on the hotel prices in the clustered hotel dataset, and recommends hotels for this business trip based on the first difference standard. This can solve the problem of hotel reservation failures and employees being unable to check in due to the fact that the prices of destination hotels are generally higher than the fixed difference standard, thereby improving the success rate and reliability of hotel reservations.

[0197] Combination of the above Figure 2-Figure 11 The hotel resource recommendation method provided by the embodiment of this application is described in detail. Figure 12 and Figure 13 The hotel resource recommendation device and electronic device provided in the embodiments of the present application are described in detail.

[0198] For example, Figure 12 This is a schematic diagram of the structure of a hotel resource recommendation device provided in an embodiment of the present application. The device can be a server or other device that can perform the server function in the hotel resource recommendation method described in the above method embodiment, or a component or assembly that can be set in a server or other device.

[0199] like Figure 12 As shown, the hotel resource recommendation device 1200 includes: a processing module 1201 and a transceiver module 1202; wherein,

[0200] The transceiver module 1202 is configured to receive a hotel reservation request from a terminal, the hotel reservation request carrying the destination of the current business trip;

[0201] Processing module 1201 is configured to obtain a first hotel dataset, where the first hotel dataset includes one or more of the following features of hotels in the destination area: location, price, travel distance, travel duration, or type;

[0202] The processing module 1201 is further configured to input the first hotel dataset into an isolation forest model for data screening to obtain a sociable hotel dataset. The isolation forest model is trained based on a hotel sample dataset, and the hotel sample dataset and the first hotel dataset are of the same type.

[0203] The processing module 1201 is further configured to determine a first difference standard based on the price of each hotel in the group hotel data set if the number of hotels in the group hotel data set is greater than or equal to the number threshold;

[0204] The processing module 1201 is further configured to determine the hotels in the group hotel data set whose prices are less than or equal to the first difference standard as recommended hotels, and the recommended hotels are used to book hotels for this business trip;

[0205] The transceiver module 1202 is further configured to send a hotel resource recommendation message to the terminal, where the hotel resource recommendation message carries a list of recommended hotels.

[0206] In one possible design solution, the processing module 1201 is further configured to:

[0207] The first hotel dataset is input into the isolation forest model for data screening to obtain the second hotel dataset;

[0208] If the number of hotels in the second hotel dataset is greater than or equal to the number threshold, the second hotel dataset is determined to be a group hotel dataset.

[0209] Optionally, the processing module 1201 is further configured to:

[0210] Input the first hotel data set into the isolation forest model, traverse each binary tree in the isolation forest model, and determine the average height of each hotel data in the first hotel data set in the isolation forest model;

[0211] Hotel data with an average height greater than or equal to a first height threshold are screened out from the first hotel data set to construct a second hotel data set.

[0212] Furthermore, the processing module 1201 is also used to filter out hotel data with an average height greater than or equal to a second height threshold from the first hotel data set to construct a second hotel data set if the number of hotel data with an average height greater than or equal to a first height threshold in the first hotel data set is less than a quantity threshold, and the second height threshold is less than the first height threshold.

[0213] Optionally, the processing module 1201 is further configured to:

[0214] Input the first hotel data set into the isolation forest model, traverse each binary tree in the isolation forest model, and determine the average height of each hotel data in the first hotel data set in the isolation forest model;

[0215] In descending order of average height, hotel data with a number greater than or equal to a threshold value are filtered out from the first hotel data set to construct the second hotel data.

[0216] In one possible design solution, the processing module 1201 is further configured to:

[0217] Determine the second difference standard based on the prices of each hotel in the Hequn Hotel Dataset. The second difference standard is the average or median of the lowest prices of each hotel in the Hequn Hotel Dataset.

[0218] The first difference mark is determined according to the second difference mark and a difference mark threshold, where the difference mark threshold corresponds to the destination.

[0219] Optionally, the processing module 1201 is further configured to determine the second difference mark as the first difference mark if the second difference mark is less than or equal to the difference mark threshold.

[0220] Optionally, the processing module 1201 is further configured to determine the difference threshold as the first difference if the second difference is greater than the difference threshold and there is a hotel in the group hotel set whose price is less than or equal to the difference threshold.

[0221] Optionally, the processing module 1201 is further configured to determine the second difference standard as the first difference standard if the price of each hotel in the group hotel set is greater than the difference standard threshold.

[0222] Furthermore, the processing module 1201 is further configured to determine the second difference standard as the first difference standard if the deviation between the first difference standard and the difference standard threshold is less than or equal to the difference standard deviation threshold.

[0223] In one possible design scheme, the processing module 1201 is also used to filter hotel data in the area where the destination is located according to the first filtering rule to generate a first hotel data set. The first filtering rule includes one or more of the following: information about the first geographical area where the destination is located, the first hotel price range, the first distance or first travel time between the hotel and the destination, and the first hotel type.

[0224] Optionally, the processing module 1201 is further configured to, if the number of hotels in the group hotel dataset is less than a quantity threshold, and / or the prices of all hotels in the group hotel dataset are greater than a difference threshold, filter hotel data in the area where the destination is located according to a second filtering rule to generate a first hotel dataset, where the second filtering rule includes one or more of the following: information about a second geographical area where the destination is located, a second hotel price range, a second distance or a second travel time between the hotel and the destination, and a second hotel type;

[0225] wherein the second geographical area includes the first geographical area and is larger than the first geographical area;

[0226] The second hotel price range includes the first hotel price range and is greater than the first hotel price range;

[0227] The second distance is greater than the first distance;

[0228] The second passage time is longer than the first passage time;

[0229] The number of types of the second hotel type is greater than the number of types of the first hotel type.

[0230] Optionally, the hotel resource recommendation message further carries difference calculation information, and the difference calculation information is used to record the calculation method and result of the first difference.

[0231] Furthermore, the transceiver module 1202 is further configured to:

[0232] Receive a hotel reservation confirmation message from the terminal, the hotel reservation confirmation message carrying reservation information for a target hotel from the recommended hotels, the target hotel being used to provide services for this business trip;

[0233] Send a hotel reservation completion message to the terminal, which carries the reservation success information of the target hotel.

[0234] Optionally, the difference calculation information and booking success information are used for travel reimbursement.

[0235] For example, Figure 13 A schematic diagram of the structure of an electronic device applicable to the hotel resource recommendation method provided in an embodiment of the present application. The electronic device may be a server, or a chip or other component or assembly used in the server, or other device capable of performing the server functions in the above method embodiment.

[0236] like Figure 13 As shown, the electronic device 1300 may include at least one processor 1301, a memory 1302, and a transceiver 1303. The at least one processor 1301, the memory 1302, and the transceiver 1303 are signal-connected to each other, for example, via a bus.

[0237] The following combination Figure 13 The components of the electronic device 1300 are described in detail.

[0238] Processor 1301 is the control center of electronic device 1300 and can be a single processor or a collective term for multiple processing elements. For example, processor 1301 can be one or more central processing units (CPUs), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present application, such as one or more digital signal processors (DSPs) or one or more field programmable gate arrays (FPGAs).

[0239] The processor 1301 can execute various functions of the electronic device 1300 by running or executing software programs stored in the memory 1302 and calling data stored in the memory 1302.

[0240] In a specific implementation, as an embodiment, the processor 1301 may include one or more CPUs, such as Figure 13 CPU0 and CPU1 are shown in the figure.

[0241] In a specific implementation, as an embodiment, the electronic device 1300 may also include multiple processors, such as Figure 13 1 and 1304 are shown in FIG. Each of these processors can be a single-CPU or a multi-CPU. A processor herein can refer to one or more communication devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0242] Memory 1302 may be, but is not limited to, a read-only memory (ROM) or other type of static storage communication device capable of storing static information and instructions, a random access memory (RAM) or other type of dynamic storage communication device capable of storing information and instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disk storage, an optical disk storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage communication device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer. Memory 1302 may exist independently or be integrated with processor 1301.

[0243] The memory 1302 is used to store the software program for executing the technical solution provided by the present application, and the execution is controlled by the processor 1301. The above specific implementation can refer to the following method embodiment, which will not be repeated here.

[0244] The transceiver 1303 is used for communicating with other electronic devices, such as a second-hand goods trading cabinet. The transceiver 1303 may include a receiving unit to implement a receiving function and a sending unit to implement a sending function.

[0245] It should be noted that Figure 13 The structure of the electronic device 1300 shown in the figure does not constitute a limitation on the electronic device. The actual electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently, which is not limited by the embodiments of the present application.

[0246] An embodiment of the present application provides a computer-readable storage medium, which stores a program or instruction. When the program or instruction runs on a computer, the computer executes the method described in any implementation manner of the first aspect.

[0247] An embodiment of the present application provides a computer program product, which includes: computer program code, which, when executed on a computer, enables the computer to execute the method described in any one of the implementations of the first aspect.

[0248] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A hotel resource recommendation method, characterized in that: Applicable to a server, the method includes: receiving a hotel reservation request from a terminal, wherein the hotel reservation request carries a destination of the current business trip; Obtaining a first hotel dataset, the first hotel dataset including one or more of the following characteristics of hotels in the area of ​​the destination: location, price, travel distance, travel duration, or type; where the travel distance is the straight-line distance or travel distance between the hotel and the destination, and the travel duration is the commuting time between the hotel and the destination, where both the travel distance and the travel duration correspond to the travel mode; The first hotel dataset is input into an isolation forest model for data screening to obtain a clustered hotel dataset, and a training-screening iterative operation is performed on the first hotel dataset: the clustered hotel dataset output after data screening using isolation forest model one is re-input into isolation forest model one for iterative training to obtain another isolation forest model two, and the first hotel dataset is again screened using isolation forest model two to obtain another clustered hotel dataset, until the number of output clustered hotel datasets meets the set requirements; the isolation forest model is trained based on the hotel sample dataset, the maximum height of the binary tree of the isolation forest model is a fraction of the number of hotel data in the historical hotel sample dataset or the first hotel dataset, and the hotel sample dataset and the first hotel dataset are datasets of the same type; If the number of hotels in the group hotel dataset is greater than or equal to the quantity threshold, determining a first difference standard according to the price of each hotel in the group hotel dataset; Determine the hotels in the group hotel data set whose prices are less than or equal to the first difference standard as recommended hotels, and use the recommended hotels to book hotels for this business trip; A hotel resource recommendation message is sent to the terminal, where the hotel resource recommendation message carries a list of recommended hotels.

2. The method according to claim 1, characterized in that The first hotel dataset is input into the isolation forest model for data screening to obtain a group hotel dataset, including: Inputting the first hotel dataset into the isolation forest model for data screening to obtain a second hotel dataset; If the number of hotels in the second hotel dataset is greater than or equal to the number threshold, the second hotel dataset is determined as the group hotel dataset.

3. The method according to claim 2, characterized in that The first hotel dataset is input into the isolation forest model for data screening to obtain a second hotel dataset, including: Inputting the first hotel dataset into the isolation forest model, traversing each binary tree in the isolation forest model, and determining the average height of each hotel data in the first hotel dataset in the isolation forest model; The second hotel dataset is constructed by filtering hotel data whose average height is greater than or equal to a first height threshold from the first hotel dataset.

4. The method according to claim 3, characterized in that The step of inputting the first hotel dataset into the isolation forest model for data screening to obtain a second hotel dataset further includes: If the number of hotel data in the first hotel data set whose average height is greater than or equal to the first height threshold is less than the number threshold, then the hotel data with the average height greater than or equal to the second height threshold is filtered out from the first hotel data set to construct the second hotel data set, and the second height threshold is less than the first height threshold.

5. The method according to claim 2, characterized in that The inputting the first hotel dataset into the isolation forest model for data screening to obtain a second hotel dataset includes: Inputting the first hotel dataset into the isolation forest model, traversing each binary tree in the isolation forest model, and determining the average height of each hotel data in the first hotel dataset in the isolation forest model; According to the order of the average height from high to low, hotel data with a number greater than or equal to the number threshold is filtered out from the first hotel data set to construct the second hotel data.

6. The method according to claim 1, characterized in that Determining the first difference based on the price of each hotel in the group hotel dataset includes: Determine a second difference standard based on the prices of the hotels in the group hotel dataset, where the second difference standard is the average or median of the lowest prices of each hotel in the group hotel dataset; The first difference mark is determined according to the second difference mark and a difference mark threshold, where the difference mark threshold corresponds to the destination.

7. The method according to claim 6, characterized in that The determining the first difference standard according to the second difference standard and a difference standard threshold comprises: If the second difference is less than or equal to the difference threshold, the second difference is determined as the first difference.

8. The method according to claim 6, characterized in that The determining the first difference standard according to the second difference standard and a difference standard threshold comprises: If the second difference is greater than the difference threshold, and there is a hotel in the group hotel set whose price is less than or equal to the difference threshold, the difference threshold is determined as the first difference.

9. The method according to claim 6, characterized in that The determining the first difference standard according to the second difference standard and a difference standard threshold comprises: If the price of each hotel in the group hotel set is greater than the difference threshold, the second difference is determined as the first difference.

10. The method according to claim 9, characterized in that The determining the second difference standard as the first difference standard includes: If the deviation between the second difference standard and the difference standard threshold is less than or equal to the difference standard threshold, the second difference standard is determined as the first difference standard.

11. The method according to claim 10, characterized in that The obtaining of the first hotel dataset includes: Hotel data in the area where the destination is located is filtered according to a first filtering rule to generate the first hotel data set, where the first filtering rule includes one or more of the following: information about a first geographical area where the destination is located, a first hotel price range, a first distance or a first travel time between the hotel and the destination, and a first hotel type.

12. The method according to claim 11, characterized in that Also includes: If the number of hotels in the group hotel dataset is less than the number threshold, and / or the prices of the hotels in the group hotel dataset are all greater than the difference threshold, then filtering the hotel data in the area where the destination is located according to a second filtering rule to generate the first hotel dataset, where the second filtering rule includes one or more of the following: information about a second geographical area where the destination is located, a second hotel price range, a second distance or a second travel time between the hotel and the destination, and a second hotel type; wherein the second geographical area includes the first geographical area and is larger than the first geographical area; The second hotel price range includes the first hotel price range and is larger than the first hotel price range; The second distance is greater than the first distance; The second passage time is longer than the first passage time; The number of types of the second hotel type is greater than the number of types of the first hotel type.

13. The method according to claim 12, characterized in that The hotel resource recommendation message also carries difference calculation information, and the difference calculation information is used to record the calculation method and result of the first difference.

14. The method according to claim 13, characterized in that Also includes: receiving a hotel reservation confirmation message from the terminal, the hotel reservation confirmation message carrying reservation information for a target hotel from the recommended hotels, the target hotel being used to provide services for this business trip; A hotel reservation completion message is sent to the terminal, wherein the hotel reservation completion message carries reservation success information of the target hotel.

15. The method according to claim 14, characterized in that The difference calculation information and the booking success information are used for this travel reimbursement.

16. An electronic device, characterized in that: include: a processor coupled to the memory; The processor is configured to execute the computer program stored in the memory, so that the electronic device executes the method according to any one of claims 1 to 15.

17. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program or instruction, and when the program or instruction is executed on a computer, the computer is caused to perform the method according to any one of claims 1 to 15.

18. A computer program product, characterized in that The computer program product comprises: a computer program code, and when the computer program code is run on a computer, the computer is caused to perform the method according to any one of claims 1 to 15.

Images