CN112488806A - Method and device for predicting order willingness, computer equipment and computer-readable storage medium - Google Patents
Method and device for predicting order willingness, computer equipment and computer-readable storage medium Download PDFInfo
- Publication number
- CN112488806A CN112488806A CN202011507413.0A CN202011507413A CN112488806A CN 112488806 A CN112488806 A CN 112488806A CN 202011507413 A CN202011507413 A CN 202011507413A CN 112488806 A CN112488806 A CN 112488806A
- Authority
- CN
- China
- Prior art keywords
- order
- driver
- data
- preset
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000012549 training Methods 0.000 claims abstract description 61
- 238000007781 pre-processing Methods 0.000 claims abstract description 18
- 238000010801 machine learning Methods 0.000 claims abstract description 15
- 238000012545 processing Methods 0.000 claims description 21
- 238000004590 computer program Methods 0.000 claims description 10
- 238000009499 grossing Methods 0.000 claims description 7
- 238000005516 engineering process Methods 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 5
- 238000011156 evaluation Methods 0.000 claims description 5
- 230000002596 correlated effect Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 6
- 230000004044 response Effects 0.000 description 6
- 230000000875 corresponding effect Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000013145 classification model Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000003442 weekly effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0633—Lists, e.g. purchase orders, compilation or processing
- G06Q30/0635—Processing of requisition or of purchase orders
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/083—Shipping
Abstract
The application example discloses a method, a device, computer equipment and a computer readable storage medium for predicting order willingness, wherein historical order transaction data are selected as training samples, positive samples and negative samples are selected from the training samples, the positive samples are samples participating in order grabbing of a driver, and the negative samples are samples not participating in order grabbing of the driver; selecting basic information from the training samples and associating; preprocessing the associated data; training, evaluating and updating a machine learning model of the preprocessed data; deploying a trained model to an online environment, the trained model for predicting driver willingness to orders. In an online service link, for each order, N drivers around the order are obtained, the intention of the N drivers to the order is predicted by using a model, the drivers with lower intention are filtered, and the order is pushed to the rest drivers. Therefore, the driver can see more interested orders, personalized pushing is provided for the driver, the use experience of the driver is improved, and the platform efficiency is finally improved.
Description
Technical Field
The embodiment of the application relates to the field of order pushing, in particular to a method and a device for predicting order willingness, computer equipment and a computer-readable storage medium.
Background
The core business of logistics is: the process of ordering by the user, taking order by the driver and performing is strongly dependent on an effective order dispatching system.
The logic of the original system is that new order requirements created by the user are broadcast to more distant drivers step by step over time until the order is responded to by the driver. It may be understood as "push by distance segment".
The original system has a problem that the driver receives too many orders to push, and the efficiency of order watching and decision making is reduced.
Disclosure of Invention
An object of the embodiments of the present application is to provide a method, an apparatus, a computer device and a computer-readable storage medium for predicting order willingness, which are used to solve the problem of order allocation and how to guarantee benefits of three parties through effective matching: for example, the user is responded and satisfied in time, the driver has a single item on the platform, the income is high, the platform is single, and the money is earned.
One aspect of an embodiment of the present application provides a method for predicting order willingness, the method including:
selecting historical order transaction data as training samples, and selecting a positive sample and a negative sample from the training samples, wherein the positive sample is used for participating in order grabbing of a driver, and the negative sample is used for not participating in order grabbing of the driver;
selecting basic information from the training samples and associating;
preprocessing the associated data;
training, evaluating and updating a machine learning model of the preprocessed data;
deploying a trained model to an online environment, the trained model for predicting driver willingness to orders.
An aspect of an embodiment of the present application further provides an apparatus for predicting order willingness, the apparatus including:
the selection module is used for selecting historical order transaction data as training samples, and selecting a positive sample and a negative sample from the training samples, wherein the positive sample is a sample participating in order grabbing of a driver, and the negative sample is a sample not participating in order grabbing of the driver;
the association module is used for selecting basic information from the training samples and associating the basic information;
the preprocessing module is used for preprocessing the associated data;
the updating module is used for training, evaluating and updating the machine learning model of the preprocessed data;
and the sequencing module is used for deploying the trained model to an online environment, and the trained model is used for predicting the intention of the driver to the order.
An aspect of the embodiments of the present application further provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method as described above when executing the computer program.
An aspect of the embodiments of the present application further provides a computer-readable storage medium, including a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method as described above when executing the computer program.
The method, the device, the computer equipment and the computer readable storage medium for predicting the order willingness select historical order transaction data as training samples, and select a positive sample and a negative sample from the training samples, wherein the positive sample is a sample participating in order grabbing of a driver, and the negative sample is a sample not participating in order grabbing of the driver; selecting basic information from the training samples and associating; preprocessing the associated data; training, evaluating and updating a machine learning model of the preprocessed data; deploying a trained model to an online environment, the trained model for predicting driver willingness to orders. In an online service link, for each order, N drivers around the order are obtained, the intention of the N drivers to the order is predicted by using a model, the drivers with lower intention are filtered, and the order is pushed to the rest drivers. Therefore, the driver can see more interested orders, personalized pushing is provided for the driver, the use experience of the driver is improved, and the platform efficiency is finally improved.
Drawings
FIG. 1 schematically illustrates an environmental application diagram according to an embodiment of the present application;
FIG. 2 is a flow chart schematically illustrating a method for predicting order willingness according to an embodiment of the present application;
FIG. 3 is a block diagram schematically illustrating an apparatus for predicting order willingness according to a second embodiment of the present application;
fig. 4 schematically shows a hardware architecture diagram of a computer device suitable for implementing the method for predicting order willingness according to the third embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that the descriptions relating to "first", "second", etc. in the embodiments of the present application are only for descriptive purposes and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In addition, technical solutions between various embodiments may be combined with each other, but must be realized by a person skilled in the art, and when the technical solutions are contradictory or cannot be realized, such a combination should not be considered to exist, and is not within the protection scope of the present application.
Fig. 1 schematically shows an environment application diagram according to an embodiment of the application.
The server 20 is connected to the mobile terminal 10 through the network 9. Each mobile terminal 10 has a client 12 disposed therein, and the client 12 is configured to receive driver information.
Network 9 includes various network devices such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, proxy devices, and/or the like. The network 9 may include physical links such as coaxial cable links, twisted pair cable links, fiber optic links, combinations thereof, and the like. The network 9 may include wireless links such as cellular links, satellite links, Wi-Fi links, etc.
The types of the mobile terminal 10 include: small, medium or large.
The server 20 receives driver's trip information, which includes: the method comprises the following steps: order basic information, vehicle basic information, driver/user basic information and scene information; the order basic information comprises: price, starting and ending point mileage, order type, order placing time, order payment mode, city and/or whether crossing the city; the vehicle basic information includes: vehicle type, size and/or special requirements including: failure to perform if the vehicle fails to meet; the driver/user basic information includes: a member level corresponding to performance capabilities and/or experiences thereof; the scene information includes: order taking distance and weather, the order taking distance includes: and when the driver receives the push of the order, the straight line distance from the starting point of the order is obtained.
The server 20 acquires information sent by the user; searching all the surrounding online drivers by taking the longitude and latitude of the information as a center; acquiring the travel information and characteristic data corresponding to the online drivers, wherein the characteristic data comprises n order pairs, the orders are the same order, the drivers correspond to n drivers, and n is a natural number; inputting feature data of n order pairs into a preset model, wherein the preset model returns n prediction scores, and the n preset scores comprise: the probability of cancellation of the order if the order is taken by each driver; and performing order division decision according to the prediction score, and outputting the driver to be broadcasted corresponding to the order.
Example one
Fig. 2 is a flow chart schematically illustrating a method for predicting order willingness according to an embodiment of the present application. It is understood that the present method embodiment may be performed in the server 20, and the flow chart of the present method embodiment is not used to limit the order in which the steps are performed.
As shown in FIG. 2, the method for predicting order willingness may include steps S200-S208, wherein:
step S200, selecting historical order transaction data as training samples, and selecting a positive sample and a negative sample from the training samples, wherein the positive sample is used for participating in order grabbing of a driver, and the negative sample is used for not participating in order grabbing of the driver;
step S202, selecting basic information from the training samples and associating the basic information;
step S204, preprocessing the associated data;
step S206, training, evaluating and updating a machine learning model for the preprocessed data;
and step S208, deploying a trained model to an online environment, wherein the trained model is used for predicting the intention of a driver to an order.
Wherein the training samples comprise: and order data in a preset area and within a first preset time length, wherein the order data comprises orders responded by drivers.
Illustratively, the basic information includes: price, order vehicle type, order placing time, order payment mode, city and/or whether to cross the city, length of a remark text reserved by the order, and starting and ending point number of the order;
the vehicle basic information includes: vehicle type, size, and/or special needs;
the driver basic information includes: a member level;
the scene information includes: order taking distance, weather, supply and demand current conditions, wherein the order taking distance comprises: when a driver receives the push of the order, the straight line distance from the starting point of the order is obtained; the supply and demand situation includes: the supply-demand ratio of a second preset time length in the past at the current position of the order;
the driver history map information includes: in all preset types of orders pushed to the driver in the past preset days, the percentage of times of the driver participating in the order taking, the quantile of the price of the orders of the driver participating in the order taking in the past preset days, the income of the driver's preset days for completing the order taking and/or the historical punctuation rate of the driver.
For example: the basic information of the order, such as price, starting and ending point mileage, order vehicle type, order placing time, order payment mode, city, whether crossing the city, and the like. The order size of the drala is large, the demands of different users are various, and the preference degrees and the response willingness of different drivers to different orders are different.
Vehicle basic information such as vehicle type, size, special needs. The size of the vehicle is directly related to whether the vehicle can be loaded or unloaded, and if the size or some special requirement of the user cannot be met by the vehicle (such as the user needs a trolley which is not equipped with a trolley), the driver does not choose to respond to the requirement of the order.
Driver/user basis information, such as membership grade. Basic information of the driver is related to performance capability, experience and psychological state of the driver, so that the willingness of the driver to order is influenced; the basic information of the user is related to the type/difficulty degree of the user requirement, and the response willingness of a driver is indirectly influenced.
Scene information:
order pickup distance (straight line distance/navigation distance of driver from order starting point when receiving order push): the further away from the start of the order, the higher the cost/time the driver spends providing the service, and the less will he
Weather: bad weather can affect the response willingness of the driver
The current conditions of supply and demand are as follows: influence the difficulty of each sheet being responded to and influence the driver's mind (such as picking sheets).
Driver/user history profile:
calculating the performance of the past X days: the drivers can reflect the preference of the drivers for different orders by comparing different service dimensions of the orders received in the past; the average response rate of the past orders of the users can reflect the attractiveness of the orders placed by the users to drivers; this all provides a priori knowledge for willingness prediction; where X is 90.
There are mainly two types of feature calculation forms:
"push rate" of a certain type of order, such as "percentage of the reservation order of 90 days in the past that a certain driver participates in pushing the order pk";
statistics of a certain order dimension, such as "order price quantile for past 90 days driver participation in order pk".
Illustratively, the preprocessing the correlated data includes:
if the driver of the order taking is a new driver, no historical portrait information exists, and the data value is null; carrying out control filling on the basic information, and processing by using a preset missing value model; filling the statistical portrait information with the average value of the portrait information of the driver in the preset time of the city; alternatively, the first and second electrodes may be,
acquiring the characteristic quality of a user or a driver under the condition of less historical order data through a Bayesian smoothing technology; alternatively, the first and second electrodes may be,
and carrying out sample unbalance processing, and carrying out down-sampling processing on the negative sample.
Optionally, positive and negative sample markers are selected, the positive sample comprising: selecting order pushing data in a preset period in the whole country, wherein the positive and negative samples are used for training a willingness degree two-classification model;
for any order, timing is started from the order grabbing moment of a first order grabbing driver after broadcasting, all the drivers participating in order grabbing in the preset time can uniformly enter a preset order grabbing rule set, and according to a preset service rule, a preference is selected from the preset order grabbing rule set, and only one driver is selected to successfully take the order;
and removing repeated pushing, and taking the primary pushing data of the driver as a sample of model training.
Specifically, selecting a training sample:
marking positive and negative samples: order push data of nearly 1 week across the country is selected as a sample, with drivers who "participate in the order pk" as positive samples. The method is considered to be capable of well covering all drivers who wish to play the order in the scene of the play order; and the other samples are marked as negative samples, and the positive samples and the negative samples are jointly used for training the willingness degree two-classification model.
Ordering pk: note that the screening is also performed through an intermediate link of a preset order taking rule during the period from "broadcasting a single batch of drivers" to "finally only 1 driver successfully takes an order" rather than the driver preemptively: for any order, timing is started from the order grabbing moment of a first order grabbing driver after broadcasting, all the drivers participating in order grabbing in 8 seconds can uniformly enter a preset order grabbing rule set, then, according to a certain service rule, preference is given to the set, and finally, only one driver is selected to successfully take the order. The link is beneficial to limiting the order grabbing and hanging and also beneficial to improving the granularity of the order control.
Duplicate sample deduplication: the original order system plays orders according to a distance segmentation delay mechanism, so that a certain order in historical data can be repeatedly pushed to a certain driver. The driver is considered to choose a robbery because of the interest in the sheet, which is not as relevant as the number of times the sheet is pushed. In order to avoid mistakenly influencing the sample weight, repeated pushing is carried out to remove the weight, and certain orders only keep once pushing data for certain drivers as samples of model training.
The training, evaluating and updating of the machine learning model for the preprocessed data comprises:
combining the model effect, the training cost, the interpretability and the engineering cost, selecting a preset model as the machine learning model for training, and determining the hyper-parameters of the model in a grid searching mode;
selecting a classic sequencing index for evaluation;
and automatically updating the version model by using the latest data every third preset time.
Optionally, the processing the data includes:
if the driver of the order taking is a new driver, no historical portrait information exists, and the data value is null; for basic information, control filling is not performed, and filling is performed through a model with a processing missing value; filling the statistical portrait information by using the cancellation rate after the average response of the preset time of the corresponding city; alternatively, the first and second electrodes may be,
the Bayesian smoothing technology is adopted to obtain the characteristic quality of the user/driver under the condition of less historical order data; alternatively, the first and second electrodes may be,
if the positive and negative samples are not balanced, performing negative sampling processing on the negative sample to enable the ratio of the positive and negative samples to reach a preset ratio; alternatively, the first and second electrodes may be,
comprehensively considering the model effect, the training cost, the interpretability and the engineering cost, selecting a model with a processing missing value as a training model, and determining the hyper-parameters of the model in a Bayesian parameter adjusting mode; alternatively, the first and second electrodes may be,
selecting a classical sequencing index AUC; alternatively, the first and second electrodes may be,
the version model is automatically updated weekly with the most recent data.
Specifically, data preprocessing:
missing value filling: if some of the drivers are new, no historical picture information exists and the data value is null. For the basic information, control filling is not performed, and the problem is solved by a model (such as xgboost) with processing missing values; for statistical image information, fill-in is performed with the average (after-response) cancellation rate of the past 1 week of the city
Data smoothing: the Bayesian smoothing technology which is commonly used in the industry is adopted, so that the characteristic quality of a user/driver under the condition of less historical order data is improved;
sample imbalance treatment: because the positive and negative samples are not balanced (< 1: 10), negative sampling is performed on the negative samples, so that the ratio of the positive and negative samples is about 1: about 3.
And others:
model training: comprehensively considering model effect, training cost, interpretability and engineering cost, and selecting xgboost as a training model. Determining model hyperparameters in a Bayesian parameter adjusting mode
And (3) off-line evaluation: and selecting the classical ordering index AUC.
Updating the model: a version of the model is automatically updated weekly with the most recent data to ensure that the model provides a representation of the service.
The method, the device, the computer equipment and the computer readable storage medium for predicting the order willingness select historical order transaction data as training samples, and select a positive sample and a negative sample from the training samples, wherein the positive sample is a sample participating in order grabbing of a driver, and the negative sample is a sample not participating in order grabbing of the driver; associating feature information from the data marts; preprocessing the associated data; training, evaluating and updating a machine learning model of the preprocessed data; deploying a trained model to an online environment, the trained model for predicting driver willingness to orders. In an online service link, for each order, N drivers around the order are obtained, the intention of the N drivers to the order is predicted by using a model, the drivers with lower intention are filtered, and the order is pushed to the rest drivers. Therefore, the driver can see more interested orders, personalized pushing is provided for the driver, the use experience of the driver is improved, and the platform efficiency is finally improved.
Example two
Fig. 3 is a block diagram schematically illustrating an apparatus for predicting order willingness according to a second embodiment of the present application, wherein the system for predicting order willingness may be divided into one or more program modules, and the one or more program modules are stored in a storage medium and executed by one or more processors to implement the second embodiment of the present application. The program modules referred to in the embodiments of the present application refer to a series of computer program instruction segments that can perform specific functions, and the following description will specifically describe the functions of the program modules in the embodiments.
As shown in FIG. 3, the apparatus 300 for predicting order willingness may include a selecting module 310, an associating module 320, a preprocessing module 330, an updating module 340, and a sorting module 350, wherein:
the selection module 310 is configured to select historical order transaction data as training samples, and select a positive sample and a negative sample from the training samples, where the positive sample is a sample participating in order grabbing by a driver, and the negative sample is a sample not participating in order grabbing by the driver;
the association module 320 is configured to select basic information from the training samples and perform association;
a preprocessing module 330, configured to preprocess the associated data;
an update module 340, configured to perform machine learning model training, evaluation, and update on the preprocessed data;
a ranking module 350 for deploying a trained model to an online environment, the trained model for predicting driver willingness to orders.
Optionally, the training samples include: the method comprises the steps of presetting order data in an area within a first preset time length, wherein the order data comprises orders responded by a driver;
the basic information includes: price, order vehicle type, order placing time, order payment mode, city and/or whether to cross the city, length of a remark text reserved by the order, and starting and ending point number of the order;
the vehicle basic information includes: vehicle type, size, and/or special needs;
the driver basic information includes: a member level;
the scene information includes: order taking distance, weather, supply and demand current conditions, wherein the order taking distance comprises: when a driver receives the push of the order, the straight line distance from the starting point of the order is obtained; the supply and demand situation includes: the supply-demand ratio of the current position of the order at intervals of a second preset time length;
the driver history map information includes: in all preset types of orders pushed to the driver in the past preset days, the percentage of times of the driver participating in the order taking, the quantile of the price of the orders of the driver participating in the order taking in the past preset days, the income of the driver's preset days for completing the order taking and/or the historical punctuation rate of the driver.
Optionally, the preprocessing module 330 is configured to:
if the driver of the order taking is a new driver, no historical portrait information exists, and the data value is null; carrying out control filling on the basic information, and processing by using a preset missing value model; filling the statistical portrait information with the average value of the portrait information of the driver in the preset time of the city; alternatively, the first and second electrodes may be,
acquiring the characteristic quality of a user or a driver under the condition of less historical order data through a Bayesian smoothing technology; alternatively, the first and second electrodes may be,
and carrying out sample unbalance processing, and carrying out down-sampling processing on the negative sample.
Optionally, the updating module 340 is configured to:
combining the model effect, the training cost, the interpretability and the engineering cost, selecting a preset model as the machine learning model for training, and determining the hyper-parameters of the model in a grid searching mode;
selecting a classic sequencing index for evaluation;
and automatically updating the version model by using the latest data every third preset time.
According to the device for predicting the order willingness, historical order transaction data are selected as training samples, positive samples and negative samples are selected from the training samples, the positive samples are samples participating in order grabbing of a driver, and the negative samples are samples not participating in order grabbing of the driver; selecting basic information from the training samples and associating; preprocessing the associated data; training, evaluating and updating a machine learning model of the preprocessed data; deploying a trained model to an online environment, the trained model for predicting driver willingness to orders. In an online service link, for each order, N drivers around the order are obtained, the intention of the N drivers to the order is predicted by using a model, the drivers with lower intention are filtered, and the order is pushed to the rest drivers. Therefore, the driver can see more interested orders, personalized pushing is provided for the driver, the use experience of the driver is improved, and the platform efficiency is finally improved.
EXAMPLE III
Fig. 4 schematically shows a hardware architecture diagram of a computer device suitable for implementing the method for predicting order willingness according to the third embodiment of the present application.
In this embodiment, the computer device 400 may be used as a provider network or a component part of a provider network, and the computer device 400 may be, for example, a virtual machine host process and one or more virtual machine instances, or a rack server, a blade server, a tower server, or a rack server (including an independent server or a server cluster composed of multiple servers), and the like.
In this embodiment, the computer device 400 may also be used as a mobile terminal or as a component part of a mobile terminal. When the computer device 400 is a mobile terminal or forms part of a mobile terminal, the computer device 400 may be, for example, a smartphone, a computer, a projector, a set-top box, or the like.
In this embodiment, the computer device 400 is a device capable of automatically performing numerical calculation and/or information processing in accordance with a command set in advance or stored. As shown in fig. 4, computer device 400 includes at least, but is not limited to: memory 410, processor 420, and network interface 430 may be communicatively linked to each other via a system bus. Wherein:
the memory 410 includes at least one type of computer-readable storage medium including a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the storage 410 may be an internal storage module of the computer device 400, such as a hard disk or a memory of the computer device 400. In other embodiments, the memory 410 may also be an external storage device of the computer device 400, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), or the like, provided on the computer device 400. Of course, the memory 410 may also include both internal and external memory modules of the computer device 400. In this embodiment, the memory 410 is generally used for storing program codes of various types of application software, such as a method for predicting order willingness, and the like, and an operating system installed in the computer apparatus 400. In addition, the memory 410 may also be used to temporarily store various types of data that have been output or are to be output.
Processor 420 may be a Central Processing Unit (CPU), controller, microcontroller, microprocessor, or other data Processing chip in some embodiments. The processor 420 is generally configured to control overall operation of the computer device 400, such as performing control and processing related to data or communication with the computer device 400. In this embodiment, the processor 420 is used to execute program codes stored in the memory 410 or process data.
Network interface 430 may include a wireless network interface or a wired network interface, and network interface 430 is typically used to establish communication links between computer device 400 and other computer devices. For example, the network interface 430 is used to connect the computer device 400 with an external terminal through a network, establish a data transmission channel and a communication link between the computer device 400 and the external terminal, and the like. The network may be a wireless or wired network such as an Intranet (Intranet), the Internet (Internet), a Global System of Mobile communication (GSM), Wideband Code Division Multiple Access (WCDMA), a 4G network, a 5G network, Bluetooth (Bluetooth), or Wi-Fi.
It should be noted that fig. 4 only shows a computer device having components 410 and 430, but it should be understood that not all of the shown components are required to be implemented, and that more or fewer components may be implemented instead.
In this embodiment, the method for predicting order willingness stored in the memory 410 can be further divided into one or more program modules and executed by one or more processors (in this embodiment, the processor 420) to complete the present application.
Example four
The present embodiments also provide a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the method of predicting order willingness in embodiments.
In this embodiment, the computer-readable storage medium includes a flash memory, a hard disk, a multimedia card, a card type memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a Read Only Memory (ROM), an Electrically Erasable Programmable Read Only Memory (EEPROM), a Programmable Read Only Memory (PROM), a magnetic memory, a magnetic disk, an optical disk, and the like. In some embodiments, the computer readable storage medium may be an internal storage unit of the computer device, such as a hard disk or a memory of the computer device. In other embodiments, the computer readable storage medium may be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like provided on the computer device. Of course, the computer-readable storage medium may also include both internal and external storage devices of the computer device. In this embodiment, the computer-readable storage medium is generally used for storing an operating system and various types of application software installed in a computer device, for example, the program code of the method for predicting order willingness in the embodiment, and the like. Further, the computer-readable storage medium may also be used to temporarily store various types of data that have been output or are to be output.
It will be apparent to those skilled in the art that the modules or steps of the embodiments of the present application described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and alternatively, they may be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, and in some cases, the steps shown or described may be performed in an order different from that described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple ones of them may be fabricated into a single integrated circuit module. Thus, embodiments of the present application are not limited to any specific combination of hardware and software.
The above description is only a preferred embodiment of the present application, and not intended to limit the scope of the present application, and all modifications of equivalent structures and equivalent processes, which are made by the contents of the specification and the drawings of the present application, or which are directly or indirectly applied to other related technical fields, are included in the scope of the present application.
Claims (10)
1. A method of predicting order willingness, the method comprising:
selecting historical order transaction data as training samples, and selecting a positive sample and a negative sample from the training samples, wherein the positive sample is used for participating in order grabbing of a driver, and the negative sample is used for not participating in order grabbing of the driver;
selecting basic information from the training samples and associating;
preprocessing the associated data;
training, evaluating and updating a machine learning model of the preprocessed data;
deploying a trained model to an online environment, the trained model for predicting driver willingness to orders.
2. The method of claim 1, wherein the training samples comprise: and order data in a preset area and within a first preset time length, wherein the order data comprises orders responded by drivers.
3. The method of claim 1, wherein the basic information comprises: the length of a remark text reserved in the order and/or the number of starting and ending points of the order;
the basic information further includes: price, order vehicle type, order placing time, order payment mode, city and/or whether crossing the city;
the vehicle basic information includes: vehicle type, size, and/or special needs;
the driver basic information includes: a member level;
the scene information includes: order taking distance, weather, supply and demand current conditions, wherein the order taking distance comprises: when a driver receives the push of the order, the straight line distance from the starting point of the order is obtained; the supply and demand situation includes: the supply-demand ratio of the current position of the order at intervals of a second preset time length;
the driver history map information includes: in all preset types of orders pushed to the driver in the past preset days, the percentage of times of the driver participating in the order taking, the quantile of the price of the orders of the driver participating in the order taking in the past preset days, the income of the driver's preset days for completing the order taking and/or the historical punctuation rate of the driver.
4. The method of claim 1, wherein preprocessing the correlated data comprises:
if the driver of the order taking is a new driver, no historical portrait information exists, and the data value is null; carrying out control filling on the basic information, and processing by using a preset missing value model; filling the statistical portrait information with the average value of the portrait information of the driver in the preset time of the city; alternatively, the first and second electrodes may be,
acquiring the characteristic quality of a user or a driver under the condition of less historical order data through a Bayesian smoothing technology; alternatively, the first and second electrodes may be,
and carrying out sample unbalance processing, and carrying out down-sampling processing on the negative sample.
5. The method of claim 1, wherein the machine learning model training, evaluating, and updating the pre-processed data comprises:
combining the model effect, the training cost, the interpretability and the engineering cost, selecting a preset model as the machine learning model for training, and determining the hyper-parameters of the model in a grid searching mode;
selecting a classic sequencing index for evaluation;
and automatically updating the version model by using the latest data every third preset time.
6. An apparatus for predicting order willingness, the apparatus comprising:
the selection module is used for selecting historical order transaction data as training samples, and selecting a positive sample and a negative sample from the training samples, wherein the positive sample is a sample participating in order grabbing of a driver, and the negative sample is a sample not participating in order grabbing of the driver;
the association module is used for selecting basic information from the training samples and associating the basic information;
the preprocessing module is used for preprocessing the associated data;
the updating module is used for training, evaluating and updating the machine learning model of the preprocessed data;
and the sequencing module is used for deploying the trained model to an online environment, and the trained model is used for predicting the intention of the driver to the order.
7. The apparatus of claim 6, wherein the training samples comprise: the method comprises the steps of presetting order data in an area within a first preset time length, wherein the order data comprises orders responded by a driver;
the basic information includes: price, order vehicle type, order placing time, order payment mode, city and/or whether to cross the city, length of a remark text reserved by the order, and starting and ending point number of the order;
the vehicle basic information includes: vehicle type, size, and/or special needs;
the driver basic information includes: a member level;
the scene information includes: order taking distance, weather, supply and demand current conditions, wherein the order taking distance comprises: when a driver receives the push of the order, the straight line distance from the starting point of the order is obtained; the supply and demand situation includes: the supply-demand ratio of the current position of the order at intervals of a second preset time length;
the driver history map information includes: in all preset types of orders pushed to the driver in the past preset days, the percentage of times of the driver participating in the order taking, the quantile of the price of the orders of the driver participating in the order taking in the past preset days, the income of the driver's preset days for completing the order taking and/or the historical punctuation rate of the driver.
8. The apparatus of claim 6, wherein the pre-processing module is configured to:
if the driver of the order taking is a new driver, no historical portrait information exists, and the data value is null; carrying out control filling on the basic information, and processing by using a preset missing value model; filling the statistical portrait information with the average value of the portrait information of the driver in the preset time of the city; alternatively, the first and second electrodes may be,
acquiring the characteristic quality of a user or a driver under the condition of less historical order data through a Bayesian smoothing technology; alternatively, the first and second electrodes may be,
and carrying out sample unbalance processing, and carrying out down-sampling processing on the negative sample.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method of any of claims 1 to 5 are implemented by the processor when executing the computer program.
10. A computer-readable storage medium, having stored thereon a computer program, the computer program being executable by at least one processor to cause the at least one processor to perform the steps of the method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011507413.0A CN112488806A (en) | 2020-12-18 | 2020-12-18 | Method and device for predicting order willingness, computer equipment and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011507413.0A CN112488806A (en) | 2020-12-18 | 2020-12-18 | Method and device for predicting order willingness, computer equipment and computer-readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112488806A true CN112488806A (en) | 2021-03-12 |
Family
ID=74914709
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011507413.0A Pending CN112488806A (en) | 2020-12-18 | 2020-12-18 | Method and device for predicting order willingness, computer equipment and computer-readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112488806A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113554387A (en) * | 2021-06-28 | 2021-10-26 | 杭州拼便宜网络科技有限公司 | Driver preference-based e-commerce logistics order allocation method, device, equipment and storage medium |
| CN114331266A (en) * | 2021-12-27 | 2022-04-12 | 深圳依时货拉拉科技有限公司 | Recommendation model training method and device and loading and unloading point recommendation method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060059023A1 (en) * | 2002-08-02 | 2006-03-16 | Alex Mashinsky | Method system and apparatus for providing transportation services |
| CN107133697A (en) * | 2017-05-03 | 2017-09-05 | 百度在线网络技术（北京）有限公司 | Estimate method, device, equipment and the storage medium of driver's order wish |
| CN107357852A (en) * | 2017-06-28 | 2017-11-17 | 镇江五八到家供应链管理服务有限公司 | A kind of determination methods of shipping driver to order wish |
-
2020
- 2020-12-18 CN CN202011507413.0A patent/CN112488806A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060059023A1 (en) * | 2002-08-02 | 2006-03-16 | Alex Mashinsky | Method system and apparatus for providing transportation services |
| CN107133697A (en) * | 2017-05-03 | 2017-09-05 | 百度在线网络技术（北京）有限公司 | Estimate method, device, equipment and the storage medium of driver's order wish |
| CN107357852A (en) * | 2017-06-28 | 2017-11-17 | 镇江五八到家供应链管理服务有限公司 | A kind of determination methods of shipping driver to order wish |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113554387A (en) * | 2021-06-28 | 2021-10-26 | 杭州拼便宜网络科技有限公司 | Driver preference-based e-commerce logistics order allocation method, device, equipment and storage medium |
| CN114331266A (en) * | 2021-12-27 | 2022-04-12 | 深圳依时货拉拉科技有限公司 | Recommendation model training method and device and loading and unloading point recommendation method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111444966B (en) | Media information classification method and device | |
| CN110399550B (en) | Information recommendation method and device | |
| CN111442778A (en) | Travel scheme recommendation method, device and equipment and computer readable storage medium | |
| CN107466469B (en) | Map drawing method, cloud platform and server thereof | |
| CN111612122A (en) | Method and device for predicting real-time demand and electronic equipment | |
| CN105744005A (en) | Client positioning and analyzing method and server | |
| CN112488806A (en) | Method and device for predicting order willingness, computer equipment and computer-readable storage medium | |
| CN111126514A (en) | Image multi-label classification method, device, equipment and medium | |
| CN110782301A (en) | Order combining method and device, electronic equipment and computer readable storage medium | |
| CN110853349A (en) | Vehicle scheduling method, device and equipment | |
| CN111899061A (en) | Order recommendation method, device, equipment and storage medium | |
| CN112488430A (en) | Modeling method and device for predicting order cancellation, computer equipment and computer readable storage medium | |
| CN112884235A (en) | Travel recommendation method, and training method and device of travel recommendation model | |
| CN114821247A (en) | Model training method and device, storage medium and electronic device | |
| CN112989188B (en) | Recommended order determining method, recommended order determining device and server | |
| US20230394552A1 (en) | Method and internet of things system of charging information recommendation for new energy vehicle in smart city | |
| CN111274471B (en) | Information pushing method, device, server and readable storage medium | |
| CN116664250A (en) | Content information recommendation method, device, server and storage medium | |
| CN111866578A (en) | Data processing method and device, electronic equipment and computer readable storage medium | |
| CN111831892A (en) | Information recommendation method, information recommendation device, server and storage medium | |
| CN113128597B (en) | Method and device for extracting user behavior characteristics and classifying and predicting user behavior characteristics | |
| CN112488794A (en) | Order broadcasting method and device, computer equipment and computer readable storage medium | |
| CN111598307A (en) | Optimization method and equipment of bus taking order scheduling system | |
| CN112465602A (en) | Order pushing method and device, computer equipment and computer readable storage medium | |
| CN113741459A (en) | Method for determining training sample and training method and device for automatic driving model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210312 |