Order receiving type prediction method and device, prediction model training method and device, equipment, storage medium and product

Abstract

The invention relates to an order receiving type prediction method and device, a prediction model training method and device, equipment, a storage medium and a product. The method comprises the steps of obtaining historical order receiving information of a user, current environment information and basic information of the user, inputting the historical order receiving information, the current environment information and the basic information into a preset order receiving prediction model for prediction, and obtaining a prediction result of an order receiving type; the prediction result comprises opening the preset order receiving function and closing the preset order receiving function. According to the method, prediction of the order receiving type is realized through the prediction model, powerful support is provided for development of an online car-hailing service platform for a driver order receiving optimization setting technology, and improvement of the overall service quality of online car-hailing and the vehicle utilization rate of a driver is facilitated.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method for predicting order types, a training method for prediction models, a device, equipment, storage medium and product. Background Art

[0002] In order to meet the different passengers' demands for price, timeliness and service quality, online ride-hailing platforms usually provide a variety of service types of orders, including flat-price orders, economical orders and premium orders. Drivers can open order interfaces of different service types on their own to receive the corresponding orders recommended by the platform.

[0003] However, users have not opened the order-taking interface for the corresponding service type in advance, resulting in delayed orders or failure to receive corresponding orders. For example, some areas may urgently need users who accept fixed-price orders, but users in these areas may not have opened the mode of accepting fixed-price orders in advance, resulting in insufficient fixed-price service capacity and excess drivers. Obviously, reasonable driver order recommendations can effectively alleviate the imbalance between order supply and demand types.

[0004] Therefore, how to reasonably predict the order type has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] Based on this, it is necessary to provide a prediction method for order types, a training method for prediction models, an apparatus, equipment, storage medium and product that can reasonably predict the order types in response to the above-mentioned technical problems.

[0006] In a first aspect, the present application provides a method for predicting an order type, comprising:

[0007] Obtain the user's historical order information, current environment information and basic information of the user;

[0008] The historical order information, current environment information and basic information are input into the preset order prediction model for prediction to obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function.

[0009] In one embodiment, the historical order information, current environment information and basic information are input into a preset order prediction model for prediction, and the prediction result of the order type is obtained, including:

[0010] Perform feature extraction on historical order information, current environment information and basic information to obtain the user's feature vector and the market feature vector corresponding to the user;

[0011] The user's feature vector and the market feature vector are input into a preset order prediction model for prediction to obtain a prediction result of the order type.

[0012] In one embodiment, the method further comprises:

[0013] If the prediction result is to enable the preset order function, the preset order function will be automatically enabled; if the prediction result is to disable the preset order function, the preset order function will be automatically disabled;

[0014] Alternatively, if the prediction result is to turn on the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's opening operation on the operation page, the preset order acceptance is turned on according to the opening operation; if the prediction result is to turn off the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's closing operation on the operation page, the preset order acceptance is turned on according to the closing operation.

[0015] In a second aspect, the present application provides a method for training a prediction model, comprising:

[0016] Obtain the user's basic information and the user's historical sample information within a preset time period; the historical sample information includes order acceptance time information and order acceptance demand information;

[0017] Extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0018] Screening multiple feature vectors to obtain a target feature vector;

[0019] The initial prediction model is trained according to the target feature vector to obtain a preset order prediction model.

[0020] In one embodiment, the above screening of multiple feature vectors to obtain a target feature vector includes:

[0021] Input multiple feature vectors into a preset gradient boosting tree model for identification, and obtain the feature values ​​corresponding to each feature vector;

[0022] According to each eigenvalue, each eigenvector is sorted to obtain a plurality of sorted eigenvectors;

[0023] A feature vector that meets a preset condition is selected from the sorted multiple feature vectors as a target feature vector.

[0024] In one embodiment, the feature extraction of the historical sample information and the basic information is performed to obtain multiple feature vectors within a preset time period, including:

[0025] Based on historical sample information and basic information, user features are extracted to obtain a user feature vector; the user feature vector includes the user's order listening time feature within a preset time period, the user's corresponding vehicle type feature, the user's completed order quantity feature within the preset time period, and the user's rating feature;

[0026] The market characteristics in the historical sample information are extracted to obtain a market characteristic vector; the market characteristic vector includes the quantity characteristics of various types of drivers within a preset time period and the characteristics of various types of orders within the preset time period.

[0027] In a third aspect, the present application further provides an order type prediction device, comprising:

[0028] The acquisition module is used to obtain the user's historical order information, current environment information and basic information of the user;

[0029] The prediction module is used to input historical order information, current environment information and basic information into a preset order prediction model for prediction, and obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function.

[0030] In a fourth aspect, the present application provides a prediction model training device, comprising:

[0031] The acquisition module is used to obtain the user's basic information and the user's historical sample information within a preset time period; the historical sample information includes order acceptance time information and order acceptance demand information;

[0032] An extraction module is used to extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0033] A screening module is used to screen multiple feature vectors to obtain a target feature vector;

[0034] The training module is used to train the initial prediction model according to the target feature vector to obtain a preset order prediction model.

[0035] In a fifth aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0036] Obtain the user's historical order information, current environment information and basic information of the user;

[0037] The historical order information, current environment information and basic information are input into the preset order prediction model for prediction to obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function.

[0038] In a sixth aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0039] Obtain the user's basic information and the user's historical sample information within a preset time period; the historical sample information includes order acceptance time information and order acceptance demand information;

[0040] Extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0041] Screening multiple feature vectors to obtain a target feature vector;

[0042] The initial prediction model is trained according to the target feature vector to obtain a preset order prediction model.

[0043] In a seventh aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0044] Obtain the user's historical order information, current environment information and basic information of the user;

[0045] The historical order information, current environment information and basic information are input into the preset order prediction model for prediction to obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function.

[0046] In an eighth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0047] Obtain the user's basic information and the user's historical sample information within a preset time period; the historical sample information includes order acceptance time information and order acceptance demand information;

[0048] Extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0049] Screening multiple feature vectors to obtain a target feature vector;

[0050] The initial prediction model is trained according to the target feature vector to obtain a preset order prediction model.

[0051] In a ninth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0052] Obtain the user's historical order information, current environment information and basic information of the user;

[0053] The historical order information, current environment information and basic information are input into the preset order prediction model for prediction to obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function.

[0054] In a tenth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0055] Obtain the user's basic information and historical sample information of the user within a preset time period; the historical sample information includes order acceptance time information and order acceptance demand information;

[0056] Extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0057] Screen multiple feature vectors to obtain a target feature vector;

[0058] The initial prediction model is trained according to the target feature vector to obtain a preset order prediction model.

[0059] The above-mentioned order type prediction method, prediction model training method, device, equipment, storage medium and product obtain the user's historical order information, current environment information and basic information of the user, input the historical order information, current environment information and basic information into the preset order prediction model for prediction, and obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function. The above-mentioned method realizes the prediction of the order type through the prediction model, provides strong support for the development of online car-hailing service platform for the optimization setting technology of driver order, and helps to improve the overall service quality of online car-hailing and the vehicle utilization rate of drivers. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0061] Figure 1 A diagram showing an application environment of a method for predicting order types in an embodiment;

[0062] Figure 2 A schematic diagram of a flow chart of a method for predicting order types in one embodiment;

[0063] Figure 3 A schematic diagram of a flow chart of a method for predicting order types in another embodiment;

[0064] Figure 4 A flowchart of a method for training a prediction model in one embodiment;

[0065] Figure 5 A schematic diagram of a flow chart of a training method for a prediction model in another embodiment;

[0066] Figure 6 A schematic diagram of a flow chart of a training method for a prediction model in another embodiment;

[0067] Figure 7 A schematic diagram of a flow chart of a method for predicting order types in one embodiment;

[0068] Figure 8 A flowchart of a method for training a prediction model in one embodiment;

[0069] Fig. 9 is a structural block diagram of a device for predicting order types in one embodiment;

[0070] Fig.10 is a structural block diagram of a training device for a prediction model in one embodiment;

[0071] Fig.11 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0073] In order to meet the different passengers' demands for price, timeliness and service quality, online ride-hailing platforms usually provide a variety of service types of orders, including flat-price orders, economical orders and premium orders. Drivers can open order interfaces of different service types on their own to receive the corresponding orders recommended by the platform.

[0074] However, users have not opened the order-taking interface for the corresponding service type in advance, resulting in delayed orders or failure to receive corresponding orders. For example, some areas may urgently need users who accept fixed-price orders, but users in these areas may not have opened the mode of accepting fixed-price orders in advance, resulting in insufficient fixed-price service capacity and excess drivers. Obviously, reasonable driver order recommendations can effectively alleviate the imbalance between order supply and demand types.

[0075] Therefore, how to reasonably predict the order type has become a technical problem that needs to be solved urgently. This application aims to solve this problem.

[0076] After introducing the background technology of the order type prediction method provided in the embodiment of the present application, the implementation environment involved in the order type prediction method provided in the embodiment of the present application will be briefly described below. The order type prediction method provided in the embodiment of the present application can be applied to Figure 1 In the implementation environment shown, the implementation environment includes a server 104, which can be implemented by an independent server 104 or a server cluster composed of multiple servers 104, and a data storage system 102 can store data that the server 104 needs to process. The data storage system 102 can be integrated on the server 104, or it can be placed on the cloud or other network servers. Among them, the server 104 can obtain the user's historical order information, the user's current environment information and the user's basic information, and predict the user's current order type based on the user's historical order information, current environment information and basic information, and obtain the result of suggesting that the user turn on the preset order function and suggesting that the user turn off the preset order function.

[0077] Those skilled in the art will understand that Figure 1 The structure shown in the figure is only a block diagram of a part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific terminal may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0078] After introducing the application scenarios of the method for predicting the order type provided in the embodiments of the present application, the following focuses on introducing the method for predicting the order type described in the present application.

[0079] In one embodiment, Figure 2 As shown in FIG. 1 , a method for predicting order types is provided, and the method is applied to Figure 1 The server in the example is used to illustrate the following steps:

[0080] S201. Obtain the user's historical order information, current environment information and basic information of the user.

[0081] Among them, the user's historical order information refers to the user's order waiting time within a preset time period, the current environment information refers to the number of tenants who may rent a car in the user's area, and the historical car rental type information of each tenant, etc. The basic information includes the type of vehicle corresponding to the user, such as electric cars, gasoline cars and hybrid cars, and for example, special cars or economy cars. It should be noted that the user in the embodiment of the present application is the driver who provides driving services.

[0082] In an embodiment of the present application, when it is necessary to predict the type of order received by a user, the user's historical order information and the user's basic information can be obtained from a preset database, and the user's current environment information can be obtained from the order receiving platform.

[0083] S202: Input historical order information, current environment information and basic information into a preset order prediction model to perform prediction and obtain a prediction result of the order type.

[0084] Among them, the prediction results include turning on the preset order acceptance function and turning off the preset order acceptance function, for example, turning on the function of the fixed-price order acceptance type, or turning off the function of the fixed-price order acceptance type.

[0085] The preset order prediction model is pre-trained based on historical order information of multiple users and the initial order prediction model. For example, the preset order prediction model can be a pre-trained logistic regression model, and the initial order prediction model is the initial logistic regression model.

[0086] In an embodiment of the present application, after obtaining the user's historical order information, the user's current environment information and the user's basic information as described above, the historical order information, the current environment information and the basic information can be directly input into a preset order prediction model for prediction to obtain a prediction type for turning on the preset order function, or to obtain a prediction type for turning off the preset order function.

[0087] Optionally, after obtaining the historical order information, current environment information and basic information as mentioned above, data cleaning and preprocessing may be performed on the historical order information, current environment information and basic information to obtain processed historical order information, current environment information and basic information, and then the processed historical order information, current environment information and basic information may be input into a preset order prediction model for prediction to obtain a prediction result of the order type.

[0088] The method for predicting the order type provided in the embodiment of the present application obtains the user's historical order information, current environment information and basic information of the user, inputs the historical order information, current environment information and basic information into a preset order prediction model for prediction, and obtains the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function. The above method predicts the order type through the prediction model, provides strong support for the development of the online car-hailing service platform for the optimization setting technology of driver orders, and helps to improve the overall service quality of online car-hailing and the vehicle utilization rate of drivers.

[0089] In one embodiment, Figure 2 Based on the embodiment shown, the process of obtaining the prediction result of the order type can be described as follows: Figure 3As shown, the above S202 "inputting historical order information, current environment information and basic information into a preset order prediction model for prediction, and obtaining a prediction result of the order type" includes:

[0090] S301. Extract features from historical order information, current environment information, and basic information respectively to obtain a feature vector of the user and a market feature vector corresponding to the user.

[0091] Among them, the user's feature vector includes the user's listening time characteristics within the preset time period, the user's corresponding vehicle type characteristics, the number of orders completed by the user within the preset time period, and the user's rating characteristics; the market feature vector includes the number characteristics of various types of drivers within the preset time period and the various types of order characteristics within the preset time period.

[0092] In an embodiment of the present application, after the historical order information, current environment information and basic information are obtained as described above, the features of the historical order information, current environment information and basic information can be extracted respectively to obtain the user's feature vector and the market feature vector corresponding to the user.

[0093] Optionally, user characteristics can be extracted based on historical order information and basic information to obtain characteristics of the user's order-listening time within a preset time period, characteristics of the vehicle type corresponding to the user, characteristics of the number of orders completed by the user within the preset time period, and characteristics of the user's rating.

[0094] Optionally, market characteristics in historical order information may be extracted to obtain characteristics of the number of various types of drivers within a preset time period and characteristics of various types of orders within a preset time period.

[0095] S302: Input the user's feature vector and the market feature vector into a preset order prediction model to perform prediction and obtain a prediction result of the order type.

[0096] In an embodiment of the present application, after the user's feature vector and the market feature vector are extracted as described above, the user's feature vector and the market feature vector may be input into a preset order prediction model for prediction, thereby obtaining a prediction result of the order type.

[0097] The method for predicting the type of order provided in the embodiment of the present application can more accurately and efficiently identify the key variables that affect the type of order a driver accepts by performing feature extraction on the original data, thereby providing a more explanatory and simpler method for predicting the driver's tendency to accept fixed-price orders.

[0098] In one embodiment, Figure 2 or Figure 3 Based on the embodiment shown, the above method also includes the following two cases:

[0099] In the first case, if the prediction result is to turn on the preset order function, the preset order function will be automatically turned on. If the prediction result is to turn off the preset order function, the preset order function will be automatically turned off.

[0100] In an embodiment of the present application, if the above prediction result is to turn on the preset order acceptance function, the preset order will be automatically opened for the user; if the prediction result is to turn off the preset order acceptance function, the preset order will be automatically turned off.

[0101] In the second case, if the prediction result is to turn on the preset order accepting function, the operation page is displayed on the display interface of the order accepting device, and in response to the user's opening operation on the operation page, the preset order accepting is turned on according to the opening operation; if the prediction result is to turn off the preset order accepting function, the operation page is displayed on the display interface of the order accepting device, and in response to the user's closing operation on the operation page, the preset order accepting is turned on according to the closing operation.

[0102] The order-taking device may be a driving device or other electronic devices of the user, such as a mobile phone, a mobile computer, a mobile watch, etc.

[0103] The operation page includes an on control and a off control.

[0104] In an embodiment of the present application, if the above-mentioned prediction result is to turn on the preset order-taking function, an operation page can be displayed on the display interface of the order-taking device, and the user can trigger the turn-on operation on the operation page, so that the server can respond to the user's triggering operation of the turn-on control on the operation page, and turn on the preset order according to the triggering operation of the turn-on control; if the above-mentioned prediction result is to turn off the preset order-taking function, an operation page can be displayed on the display interface of the order-taking device, and the user can trigger the turn-off operation on the operation page, so that the server can respond to the user's triggering operation of the turn-off control on the operation page, and turn on the preset order according to the triggering operation of the turn-off control.

[0105] The method for predicting the order type provided in the embodiments of the present application can realize the function of directly turning on or off the preset order, or realize the recommendation and independent selection of the order type through the interaction between the user and the device interface, thereby providing strong support for the development of online ride-hailing service platform for driver order optimization setting technology, which helps to improve the overall service quality of online ride-hailing and the vehicle utilization rate of drivers.

[0106] Before using the preset order prediction model, the preset order prediction model needs to be trained. In one embodiment, Figure 4 As shown, a prediction model training method is provided, which is applied to Figure 1 The server in the example is used to illustrate the following steps:

[0107] S401: Obtain basic information of a user and historical sample information of the user within a preset time period.

[0108] The historical sample information includes order acceptance time information and order acceptance demand information of multiple users within a preset time period.

[0109] In the embodiment of the present application, before training the prediction model, it is necessary to obtain the user's basic information and the user's historical sample information within a preset event segment, train the initial prediction model, and then obtain the prediction model after the training is completed.

[0110] S402: Extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period.

[0111] The multiple feature vectors include user feature vectors and market feature vectors. The user feature vectors include the user's order listening time characteristics within a preset time period, the user's corresponding vehicle type characteristics, the number of orders completed by the user within the preset time period, and the user's rating characteristics; the market feature vectors include the number of drivers of various types within the preset time period and the order characteristics of various types within the preset time period.

[0112] In the embodiment of the present application, after the historical sample information and the basic information of the user are obtained as described above, feature extraction may be performed on the historical sample information and the basic information to obtain a plurality of feature vectors within a preset time period.

[0113] Optionally, after acquiring the historical sample information and basic information as described above, data cleaning and preprocessing may be performed on the historical sample information and basic information to obtain processed historical sample information and basic information, and then feature extraction may be performed on the processed historical sample information and basic information to obtain multiple feature vectors within a preset time period.

[0114] Optionally, a feature extraction method is also provided below, see Figure 5 The above S402 "extracting features from historical sample information and basic information to obtain multiple feature vectors within a preset time period" includes:

[0115] S501. Extract user features based on historical sample information and basic information to obtain a user feature vector.

[0116] Among them, the user feature vector includes the user's listening time characteristics within the preset time period, the user's corresponding vehicle type characteristics, the number of orders completed by the user within the preset time period, and the user's rating characteristics.

[0117] In the embodiment of the present application, after the historical sample information and basic information are obtained as described above, user features can be extracted based on the historical sample information and basic information to obtain a user feature vector. Optionally, the average and standard deviation of the time each driver listens to orders, and the average and standard deviation of the time for listening to preset orders are obtained based on the historical capacity data in the historical sample information, where the standard deviation shows the stability and volatility of the driver's behavior; optionally, the type of vehicle corresponding to the driver (for example, electric cars, gasoline cars, and hybrid cars) can be extracted based on the basic information, as well as the scoring features of the user's platform; optionally, the number of orders served by the driver per unit time, the average and standard deviation of the number of historical service orders, and the average and standard deviation of service preset orders can be extracted based on the historical sample information.

[0118] S502: Extract market features from historical sample information to obtain a market feature vector.

[0119] The market feature vector includes the quantity characteristics of various types of drivers within a preset time period and the characteristics of various types of orders within a preset time period.

[0120] In the embodiment of the present application, after the historical sample information is obtained as described above, the market characteristics can be extracted based on the historical sample information to obtain a market characteristic vector. Optionally, the number of various types of drivers in a tenant within a preset time period can be counted based on the historical sample information, specifically including the number of drivers whose listening time is greater than a certain threshold and whose listening time is greater than a certain threshold; and the number of various types of orders received by the tenant can be counted, including only checking the fixed price orders, only checking the economic orders, only checking the premium orders, checking the fixed price and economic orders, checking the fixed price and premium orders, checking the economic orders and premium orders, and checking the three types of orders at the same time. These seven types of orders.

[0121] Optionally, other important features may also be extracted, for example, counting the proportion of idle time of the driver in the area to characterize the waiting time of the driver between receiving two adjacent orders.

[0122] S403: Screen multiple feature vectors to obtain a target feature vector.

[0123] In the embodiment of the present application, after the multiple feature vectors within the preset time period are obtained, at least one feature vector can be selected from the multiple feature vectors as a target feature vector.

[0124] Optionally, a feature vector screening method is provided below, see Figure 6 The above S403 “screening multiple feature vectors to obtain a target feature vector” includes:

[0125] S601, inputting a plurality of feature vectors into a preset gradient boosting tree model for identification, and obtaining feature values ​​corresponding to each feature vector.

[0126] Among them, the preset gradient boosting tree model can be pre-trained.

[0127] In the embodiment of the present application, after the multiple feature vectors are obtained as described above, the multiple feature vectors can be input into a preset gradient boosting tree model for identification to obtain the feature values ​​corresponding to each feature vector.

[0128] Optionally, after obtaining multiple feature vectors as mentioned above, the user feature vectors and market feature vectors in the multiple feature vectors can be integrated to obtain the supply and demand quantities and driver characteristics in each data within a preset time period, and the supply and demand quantities and driver characteristics in each data are used as explanatory variables and input into the initial gradient boosting tree model for training to obtain initial training results, and the initial training results are compared with the label results corresponding to the corresponding feature vectors to obtain comparison results, and a loss function (for example, mean square error loss function, mean absolute error loss function, etc.) is predetermined to measure the difference between the initial training results and the label results corresponding to the feature vectors, and when the value of the loss function is less than the preset difference, the preset gradient boosting tree model training is completed.

[0129] S602: Sort the eigenvectors according to the eigenvalues ​​to obtain a plurality of sorted eigenvectors.

[0130] In the embodiment of the present application, after the eigenvalues ​​corresponding to the eigenvectors are obtained, the eigenvectors can be sorted in order from large to small or from small to large according to the eigenvalues ​​to obtain a plurality of sorted eigenvectors.

[0131] S603 , selecting a feature vector that meets a preset condition from the sorted plurality of feature vectors as a target feature vector.

[0132] The preset condition may be a condition that satisfies a preset sorting requirement. For example, when a sorting requirement from large to small is met, the top five feature vectors are used as target feature vectors.

[0133] In the embodiment of the present application, after the sorted multiple feature vectors are obtained as described above, feature vectors that meet preset conditions can be screened from the multiple feature vectors, and the feature vectors that meet the preset conditions can be determined as target feature vectors.

[0134] S404: Train the initial prediction model according to the target feature vector to obtain a preset order prediction model.

[0135] In the embodiment of the present application, after the target feature vector is obtained as described above, the initial prediction model can be trained based on the target feature vector to obtain a trained prediction model, that is, a preset order prediction model.

[0136] Optionally, after obtaining the target feature vector as described above, the target feature vector can be used as an explanatory variable and input into the initial prediction model for training to obtain an initial training result, and the initial training result is compared with the label result corresponding to the feature vector to obtain a comparison result, and a loss function (for example, a mean square error loss function, a mean absolute error loss function, etc.) is predetermined to measure the difference between the initial training result and the label result corresponding to the feature vector, and when the value of the loss function is less than a preset difference, the training of the preset order prediction model is completed.

[0137] The training method of the prediction model provided in the embodiment of the present application trains the order prediction model based on massive supply and demand data, and can accurately and efficiently identify the key variables that affect the driver's preference for accepting preset orders, thereby providing a more interpretable and simpler method for predicting the driver's preference for accepting preset orders; in addition, since the logistic regression prediction model has the advantages of a simple training process and interpretable results, it can provide more convincing prediction results in actual operating scenarios.

[0138] In one embodiment, Figure 7 As shown, a method for predicting order types is provided, including:

[0139] S10, obtaining the user's historical order information, current environment information and basic information of the user;

[0140] S11, extracting features from historical order information, current environment information, and basic information to obtain a feature vector of the user and a market feature vector corresponding to the user;

[0141] S12, inputting the user's feature vector and the market feature vector into a preset order prediction model for prediction, and obtaining a prediction result of the order type;

[0142] S13. If the prediction result is to turn on the preset order receiving function, the preset order receiving function is automatically turned on. If the prediction result is to turn off the preset order receiving function, the preset order receiving function is automatically turned off.

[0143] In one embodiment, Figure 8 As shown, a prediction model training method is provided, including:

[0144] S20, obtaining basic information of the user and historical sample information of the user within a preset time period;

[0145] S21, extracting features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0146] S22. Extract user features based on historical sample information and basic information to obtain a user feature vector;

[0147] S23, extracting market features from historical sample information to obtain a market feature vector;

[0148] S24, sorting the eigenvectors according to the eigenvalues ​​to obtain a plurality of sorted eigenvectors;

[0149] S25, selecting a feature vector that meets a preset condition from the sorted plurality of feature vectors as a target feature vector;

[0150] S26. Train the initial prediction model according to the target feature vector to obtain a preset order prediction model.

[0151] The method for predicting the order type provided in the embodiment of the present application obtains the user's historical order information, current environment information and basic information of the user, inputs the historical order information, current environment information and basic information into a preset order prediction model for prediction, and obtains the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function. The above method predicts the order type through the prediction model, provides strong support for the development of the online car-hailing service platform for the optimization setting technology of driver orders, and helps to improve the overall service quality of online car-hailing and the vehicle utilization rate of drivers.

[0152] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0153] Based on the same inventive concept, the embodiment of the present application also provides a prediction device for order types for implementing the above-mentioned prediction method for order types. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in the embodiments of one or more prediction devices for order types provided below can refer to the limitations of the prediction method for order types above, and will not be repeated here.

[0154] In an exemplary embodiment, Fig. 9 As shown, a device for predicting order types is provided, including: an acquisition module 10 and a prediction module 11, wherein:

[0155] The acquisition module 10 is used to obtain the user's historical order information, current environment information and basic information of the user.

[0156] The prediction module 11 is used to input historical order information, current environment information and basic information into a preset order prediction model for prediction, and obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function.

[0157] In an exemplary embodiment, the prediction module 11 includes: an extraction unit and a prediction unit, wherein:

[0158] The extraction unit is specifically used to extract features from historical order information, current environment information and basic information respectively, to obtain a feature vector of the user and a market feature vector corresponding to the user;

[0159] The prediction unit is specifically used to input the user's feature vector and the market feature vector into a preset order prediction model for prediction, and obtain a prediction result of the order type.

[0160] In an exemplary embodiment, the above device further includes: an automatic opening / closing module, which is used to automatically open the preset order receiving function when the prediction result is to open the preset order receiving function, and automatically close the preset order receiving function when the prediction result is to close the preset order receiving function;

[0161] Alternatively, in a case where the prediction result is to turn on the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's opening operation on the operation page, the preset order acceptance is turned on according to the opening operation; in a case where the prediction result is to turn off the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's closing operation on the operation page, the preset order acceptance is turned on according to the closing operation.

[0162] In an exemplary embodiment, Fig.10As shown, a prediction model training device is provided, comprising: an acquisition module 20, an extraction module 21, a screening module 22 and a training module 23, wherein:

[0163] The acquisition module 20 is used to acquire the basic information of the user and the historical sample information of the user within a preset time period; the historical sample information includes the order receiving time information and the order receiving demand information;

[0164] An extraction module 21 is used to extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0165] A screening module 22, used for screening multiple feature vectors to obtain a target feature vector;

[0166] The training module 23 is used to train the initial prediction model according to the target feature vector to obtain a preset order prediction model.

[0167] In an exemplary embodiment, the screening module 22 includes: an identification unit, a processing unit and a screening unit, wherein:

[0168] The recognition unit is specifically used to input multiple feature vectors into a preset gradient boosting tree model for recognition, and obtain the feature values ​​corresponding to each feature vector;

[0169] A processing unit, specifically used to sort each eigenvector according to each eigenvalue to obtain a plurality of sorted eigenvectors;

[0170] The screening unit is specifically used to screen a feature vector that meets a preset condition from the sorted plurality of feature vectors as a target feature vector.

[0171] In an exemplary embodiment, the extraction module 21, wherein: a first extraction unit and a second extraction unit, wherein:

[0172] The first extraction unit is specifically used to extract user features according to historical sample information and basic information to obtain a user feature vector; the user feature vector includes a feature of the duration of the user's order listening within a preset time period, a feature of the vehicle type corresponding to the user, a feature of the number of orders completed by the user within the preset time period, and a feature of the user's score;

[0173] The second extraction unit is specifically used to extract market features from historical sample information to obtain a market feature vector; the market feature vector includes various types of driver quantity features within a preset time period and various types of order receiving features within a preset time period.

[0174] Each module in the above-mentioned order type prediction device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0175] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig.11 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store historical order data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for predicting an order type is implemented.

[0176] Those skilled in the art will understand that Fig.11 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0177] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0178] Obtain the user's historical order information, current environment information and basic information of the user;

[0179] The historical order information, current environment information and basic information are input into the preset order prediction model for prediction to obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function.

[0180] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0181] Perform feature extraction on historical order information, current environment information and basic information to obtain the user's feature vector and the market feature vector corresponding to the user;

[0182] The user's feature vector and the market feature vector are input into a preset order prediction model for prediction to obtain a prediction result of the order type.

[0183] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0184] If the prediction result is to enable the preset order function, the preset order function will be automatically enabled; if the prediction result is to disable the preset order function, the preset order function will be automatically disabled;

[0185] Alternatively, if the prediction result is to turn on the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's opening operation on the operation page, the preset order acceptance is turned on according to the opening operation; if the prediction result is to turn off the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's closing operation on the operation page, the preset order acceptance is turned on according to the closing operation.

[0186] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0187] Obtain the user's basic information and the user's historical sample information within a preset time period; the historical sample information includes order acceptance time information and order acceptance demand information;

[0188] Extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0189] Screening multiple feature vectors to obtain a target feature vector;

[0190] The initial prediction model is trained according to the target feature vector to obtain a preset order prediction model.

[0191] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0192] Input multiple feature vectors into a preset gradient boosting tree model for identification, and obtain the feature values ​​corresponding to each feature vector;

[0193] According to each eigenvalue, each eigenvector is sorted to obtain a plurality of sorted eigenvectors;

[0194] A feature vector that meets a preset condition is selected from the sorted multiple feature vectors as a target feature vector.

[0195] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0196] Based on historical sample information and basic information, user features are extracted to obtain a user feature vector; the user feature vector includes the user's order listening time feature within a preset time period, the user's corresponding vehicle type feature, the user's completed order quantity feature within the preset time period, and the user's rating feature;

[0197] The market characteristics in the historical sample information are extracted to obtain a market characteristic vector; the market characteristic vector includes the quantity characteristics of various types of drivers within a preset time period and the characteristics of various types of orders within the preset time period.

[0198] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0199] Obtain the user's historical order information, current environment information and basic information of the user;

[0200] The historical order information, current environment information and basic information are input into the preset order prediction model for prediction to obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function.

[0201] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0202] Perform feature extraction on historical order information, current environment information and basic information to obtain the user's feature vector and the market feature vector corresponding to the user;

[0203] The user's feature vector and the market feature vector are input into a preset order prediction model for prediction to obtain a prediction result of the order type.

[0204] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0205] If the prediction result is to enable the preset order function, the preset order function will be automatically enabled; if the prediction result is to disable the preset order function, the preset order function will be automatically disabled;

[0206] Alternatively, if the prediction result is to turn on the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's opening operation on the operation page, the preset order acceptance is turned on according to the opening operation; if the prediction result is to turn off the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's closing operation on the operation page, the preset order acceptance is turned on according to the closing operation.

[0207] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0208] Obtain the user's basic information and historical sample information of the user within a preset time period; the historical sample information includes order acceptance time information and order acceptance demand information;

[0209] Extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0210] Screen multiple feature vectors to obtain a target feature vector;

[0211] The initial prediction model is trained according to the target feature vector to obtain a preset order prediction model.

[0212] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0213] Input multiple feature vectors into a preset gradient boosting tree model for identification, and obtain the feature values ​​corresponding to each feature vector;

[0214] According to each eigenvalue, each eigenvector is sorted to obtain a plurality of sorted eigenvectors;

[0215] A feature vector that meets a preset condition is selected from the sorted multiple feature vectors as a target feature vector.

[0216] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0217] Based on historical sample information and basic information, user features are extracted to obtain a user feature vector; the user feature vector includes the user's order listening time feature within a preset time period, the user's corresponding vehicle type feature, the user's completed order quantity feature within the preset time period, and the user's rating feature;

[0218] The market characteristics in the historical sample information are extracted to obtain a market characteristic vector; the market characteristic vector includes the quantity characteristics of various types of drivers within a preset time period and the characteristics of various types of orders within the preset time period.

[0219] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0220] Obtain the user's historical order information, current environment information and basic information of the user;

[0221] The historical order information, current environment information and basic information are input into the preset order prediction model for prediction to obtain the prediction result of the order type; the prediction result includes turning on the preset order function and turning off the preset order function.

[0222] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0223] Perform feature extraction on historical order information, current environment information and basic information to obtain the user's feature vector and the market feature vector corresponding to the user;

[0224] The user's feature vector and the market feature vector are input into a preset order prediction model for prediction to obtain a prediction result of the order type.

[0225] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0226] If the prediction result is to enable the preset order function, the preset order function will be automatically enabled; if the prediction result is to disable the preset order function, the preset order function will be automatically disabled;

[0227] Alternatively, if the prediction result is to turn on the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's opening operation on the operation page, the preset order acceptance is turned on according to the opening operation; if the prediction result is to turn off the preset order acceptance function, an operation page is displayed on the display interface of the order acceptance device, and in response to the user's closing operation on the operation page, the preset order acceptance is turned on according to the closing operation.

[0228] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0229] Obtain the user's basic information and the user's historical sample information within a preset time period; the historical sample information includes order acceptance time information and order acceptance demand information;

[0230] Extract features from historical sample information and basic information to obtain multiple feature vectors within a preset time period;

[0231] Screening multiple feature vectors to obtain a target feature vector;

[0232] The initial prediction model is trained according to the target feature vector to obtain a preset order prediction model.

[0233] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0234] Input multiple feature vectors into a preset gradient boosting tree model for identification, and obtain the feature values ​​corresponding to each feature vector;

[0235] According to each eigenvalue, each eigenvector is sorted to obtain a plurality of sorted eigenvectors;

[0236] A feature vector that meets a preset condition is selected from the sorted multiple feature vectors as a target feature vector.

[0237] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0238] Based on historical sample information and basic information, user features are extracted to obtain a user feature vector; the user feature vector includes the user's order listening time feature within a preset time period, the user's corresponding vehicle type feature, the user's completed order quantity feature within the preset time period, and the user's rating feature;

[0239] The market characteristics in the historical sample information are extracted to obtain a market characteristic vector; the market characteristic vector includes the quantity characteristics of various types of drivers within a preset time period and the characteristics of various types of orders within the preset time period.

[0240] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0241] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0242] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0243] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for predicting order types, characterized in that: The method comprises: Obtain the user's historical order information, current environment information and basic information of the user; The historical order information, the current environment information and the basic information are input into a preset order prediction model for prediction to obtain a prediction result of the order type; the prediction result includes turning on a preset order function and turning off a preset order function.

2. The method according to claim 1, characterized in that The inputting the historical order information, the current environment information and the basic information into a preset order prediction model for prediction to obtain a prediction result of the order type includes: Extracting features from the historical order information, the current environment information, and the basic information respectively to obtain a feature vector of the user and a market feature vector corresponding to the user; The user's feature vector and the market feature vector are input into the preset order prediction model for prediction to obtain a prediction result of the order type.

3. The method according to claim 1, characterized in that The method further comprises: If the prediction result is the preset order receiving function being turned on, the preset order receiving function is automatically turned on; if the prediction result is the preset order receiving function being turned off, the preset order receiving function is automatically turned off; Alternatively, if the prediction result is to turn on the preset order-taking function, an operation page is displayed on the display interface of the order-taking device, and in response to the user's opening operation of the operation page, the preset order is turned on according to the opening operation; if the prediction result is to turn off the preset order-taking function, an operation page is displayed on the display interface of the order-taking device, and in response to the user's closing operation of the operation page, the preset order is turned on according to the closing operation.

4. A method for training a prediction model, characterized in that: The method comprises: Obtaining basic information of the user and historical sample information of the user within a preset time period; the historical sample information includes order acceptance time information and order acceptance demand information; Performing feature extraction on the historical sample information and the basic information to obtain a plurality of feature vectors within the preset time period; Screening the multiple feature vectors to obtain a target feature vector; The initial prediction model is trained according to the target feature vector to obtain a preset order prediction model.

5. The method according to claim 4, characterized in that The step of screening the plurality of feature vectors to obtain a target feature vector includes: Input the multiple feature vectors into a preset gradient boosting tree model for identification, and obtain the feature values ​​corresponding to each feature vector; According to each of the eigenvalues, each of the eigenvectors is sorted to obtain a plurality of sorted eigenvectors; A feature vector that meets a preset condition is selected from the sorted plurality of feature vectors to serve as the target feature vector.

6. The method according to claim 4 or 5, characterized in that: The extracting features from the historical sample information and the basic information to obtain a plurality of feature vectors within the preset time period includes: Extracting the user features according to the historical sample information and the basic information to obtain a user feature vector; the user feature vector includes a feature of the duration of the user's order reception within the preset time period, a feature of the vehicle type corresponding to the user, a feature of the number of orders completed by the user within the preset time period, and a feature of the user's score; The market characteristics in the historical sample information are extracted to obtain a market characteristic vector; the market characteristic vector includes the quantity characteristics of various types of drivers within the preset time period and the characteristics of various types of orders within the preset time period.

7. A device for predicting order types, characterized in that: The device comprises: An acquisition module is used to acquire the user's historical order information, current environment information and basic information of the user; The prediction module is used to input the historical order information, the current environment information and the basic information into a preset order prediction model for prediction, and obtain a prediction result of the order type; the prediction result includes turning on the order function and turning off the order function.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images