CN114463103B - Data processing method and equipment - Google Patents

Data processing method and equipment Download PDF

Info

Publication number
CN114463103B
CN114463103B CN202210367301.2A CN202210367301A CN114463103B CN 114463103 B CN114463103 B CN 114463103B CN 202210367301 A CN202210367301 A CN 202210367301A CN 114463103 B CN114463103 B CN 114463103B
Authority
CN
China
Prior art keywords
shop
time length
information
target
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.)
Active
Application number
CN202210367301.2A
Other languages
Chinese (zh)
Other versions
CN114463103A (en
Inventor
周凯荣
朱麟
王鹏宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang Koubei Network Technology Co Ltd
Original Assignee
Zhejiang Koubei Network Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Zhejiang Koubei Network Technology Co Ltd filed Critical Zhejiang Koubei Network Technology Co Ltd
Priority to CN202210367301.2A priority Critical patent/CN114463103B/en
Publication of CN114463103A publication Critical patent/CN114463103A/en
Application granted granted Critical
Publication of CN114463103B publication Critical patent/CN114463103B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q30/00Commerce
    • G06Q30/06Buying, selling or leasing transactions
    • G06Q30/0601Electronic shopping [e-shopping]
    • G06Q30/0641Shopping interfaces
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00Machine learning
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q10/00Administration; Management
    • G06Q10/08Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083Shipping
    • G06Q10/0838Historical data
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q30/00Commerce
    • G06Q30/06Buying, selling or leasing transactions
    • G06Q30/0601Electronic shopping [e-shopping]
    • G06Q30/0633Lists, e.g. purchase orders, compilation or processing
    • G06Q30/0635Processing of requisition or of purchase orders
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q30/00Commerce
    • G06Q30/06Buying, selling or leasing transactions
    • G06Q30/0601Electronic shopping [e-shopping]
    • G06Q30/0639Item locations

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Physics & Mathematics (AREA)
  • Development Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Human Resources & Organizations (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The application discloses a data processing method and equipment. The method comprises the following steps: the method comprises the steps of obtaining list information characteristics of a target shop, and obtaining a first estimated distribution time length which can be associated with the target shop and is displayed on a first display interface by using a first model according to the list information characteristics of the target shop; the first model is generated according to a first learning sample constructed based on ordering information characteristics of a shop; obtaining order data of a user at the target shop, and obtaining a second expected distribution time length which can be displayed on a second display interface by using the first model according to order placing information characteristics of the order data; wherein a difference between the first estimated delivery duration and the second estimated delivery duration satisfies a preset threshold condition. By adopting the method, the problem that the consistency rate of the first expected distribution time length of the shop and the second expected distribution time length corresponding to the order data of the shop is low is solved.

Description

Data processing method and equipment
Technical Field
The present application relates to the field of computer processing technologies, and in particular, to a data processing method and device.
Background
With the development of electronic commerce, more and more users select online shopping, such as online ordering, online buying of living goods, and the like, and distribution is an important influence factor of online shopping experience of the users. In order to improve user experience, when a shopping platform displays a shop list, a first expected distribution duration of a shop is displayed; when the user enters one of the shops to place an order, the second expected delivery time length of the order data of the shop is displayed on an order information page. If the difference between the first expected delivery duration of the store and the second expected delivery duration of the spot order data is large, the user experience is affected, the order placing may fail, and the conversion rate of potential users to actual users is affected. Therefore, the difference between the first estimated delivery duration and the second estimated delivery duration is particularly important within a certain range.
In the prior art, the characteristic information corresponding to the order data is different from the characteristic information of the shop, and if the difference between the first estimated delivery time length and the second estimated delivery time length is within a certain range, the first estimated delivery time length and the second estimated delivery time length are considered to be consistent. The first estimated delivery duration is generally calculated in the following manner: and counting the distribution time lengths of different shops at different distances in different time periods in history according to different shops, and taking the average value as the first expected distribution time length of the shop at the current time. Actually, many factors affect the distribution time length, and it is not accurate enough to calculate the first estimated distribution time length of the store only by the statistical data of the time interval and the distance. Thus, there may be a large difference between the first estimated delivery time period and the second estimated delivery time period, so that the rate of coincidence therebetween is low.
Therefore, how to increase the matching rate of the first estimated distribution time length of the store and the second estimated distribution time length corresponding to the order data of the store is a problem to be solved.
Disclosure of Invention
The data processing method provided by the embodiment of the application solves the problem that the consistency rate of the first expected distribution time length of the shop and the second expected distribution time length corresponding to the order data of the shop is low.
An embodiment of the present application provides a data processing method, including: the method comprises the steps of obtaining list information characteristics of a target shop, and obtaining a first estimated distribution time length which can be associated with the target shop and is displayed on a first display interface by using a first model according to the list information characteristics of the target shop; the first model is generated according to a first learning sample constructed based on ordering information characteristics of a shop; obtaining order data of a user at the target shop, and obtaining a second expected distribution time length which can be displayed on a second display interface by using the first model according to order placing information characteristics of the order data; wherein a difference between the first estimated delivery duration and the second estimated delivery duration satisfies a preset threshold condition.
Optionally, the method further includes: obtaining order placing information characteristics of shops in a historical calling log and estimated distribution duration corresponding to the order placing information characteristics; taking the ordering information characteristic as an input characteristic, taking the expected delivery duration corresponding to the ordering information characteristic as a corresponding output characteristic, and constructing the first learning sample for generating a first model; generating the first model using the first learning sample; the first model is used for estimating a first estimated distribution time length of the shop according to the list information characteristics of the shop, and/or estimating a second estimated distribution time length of the order data according to ordering information characteristics of the order data associated with the shop.
Optionally, the obtaining, according to the list information feature of the target store, a first expected delivery duration that can be displayed on a first display interface in association with the target store by using a first model includes: inputting the list information characteristic of the target shop into the first model, and taking the output of the first model as the estimated distribution time length to be corrected; acquiring list information characteristics of the target shop, and acquiring a list interface distribution time length correction value of the target shop by using a correction model according to the list information characteristics and the estimated distribution time length to be corrected; the correction model is generated according to a second learning sample constructed based on the list information characteristics of the shop and the corresponding ordering information characteristics; and correcting the estimated distribution time length to be corrected according to the list interface distribution time length correction value of the target shop to obtain the first estimated distribution time length.
Optionally, the method further includes: extracting list information characteristics of the shop and corresponding ordering information characteristics from the historical calling log; obtaining a first estimated distribution time length of the shop according to the first model by using the list information characteristics of the shop; obtaining a second expected distribution time length of the shop according to ordering information characteristics of the shop by using the first model; taking a difference between the first expected delivery time length and the second expected delivery time length as a time length difference label of the shop; and constructing a second learning sample for generating the correction model by using the list information characteristics of the shop and the corresponding first expected delivery time length as input characteristics and the corresponding time length difference label as output characteristics, and generating the correction model by using the second learning sample.
Optionally, the obtaining, according to the list information feature of the target store, a first expected delivery duration that can be displayed on a first display interface in association with the target store by using a first model includes: taking the list information characteristic of the target shop as a first characteristic; determining the missing features of the target shop according to the information of the predetermined missing features; the information of the missing characteristics is determined according to the dimension difference between the list information characteristics of the shops and the ordering information characteristics of the shops; filling the first characteristics by using the missing characteristics of the target shop to obtain second characteristics of the target shop; and inputting the second characteristics of the target shop into the first model, and obtaining the first expected delivery time according to the output result of the first model.
Optionally, the information of the missing feature includes feature information of at least the following dimensions: the method comprises the following steps of (1) ordering object quantity dimension, price dimension, shop meal-out duration dimension and user payment time dimension; the determining the missing feature of the target shop according to the predetermined information of the missing feature comprises the following steps: taking the average value of the number of objects in the historical time period of the target shop as the missing characteristic data of the target shop corresponding to the dimension of the number of the objects; and/or taking the price average value of the target shop in the historical time period as the missing characteristic data of the target shop corresponding to the price dimension; and/or taking the average meal serving duration of a target store adjacent to the last specified time period at the current moment as missing characteristic data of the target store corresponding to the meal serving duration dimension; and/or taking the visit time of the user visiting the shop list information containing the target shop as the missing characteristic data of the target shop corresponding to the payment time dimension of the user; the populating the first feature with the missing features of the targeted store to obtain a second feature of the targeted store includes: and filling the first characteristic by using the at least one missing characteristic data to obtain the second characteristic.
Optionally, the order placing information features of the shop include at least the following dimension features corresponding to the order placing information: the system comprises a store, a distribution area, a distribution time length dimension, a distribution pressure dimension, a distribution distance dimension, a weather dimension, an object quantity dimension of an order, a price dimension of the order, a store meal length dimension and a user payment time dimension.
Optionally, the method further includes: and calling the first model to calculate the first expected delivery time length and the second expected delivery time length through the same module calling link.
Optionally, the method includes: receiving a request of a client used by a target user for obtaining a shop list, and sending the shop list containing the target shop and the first expected delivery time length to the client so as to enable the target shop and the first expected delivery time length to be related to a first display interface displayed on the client; and receiving a settlement request of the target user for the order data, and sending the second expected delivery duration to the client so that the second expected delivery duration and the settlement information of the order data are displayed on a second display interface of the client.
The present embodiment further provides a data processing method, including: displaying basic shop information of one or more shops and a first expected delivery time length of the shops on a first display interface; receiving order data of a user in a target shop, which is selected by the user from the shops and triggered, and determining the order data of the user in the target shop according to the order information of the target shop; obtaining a second estimated distribution time length corresponding to the order data, and displaying the second estimated distribution time length on a second display interface; the first estimated distribution time length and the second estimated distribution time length are obtained according to a first model respectively, and the difference between the first estimated distribution time length and the second estimated distribution time length meets a preset threshold condition; the first model is generated according to a first learning sample constructed based on ordering information characteristics of the shop.
Optionally, the first expected delivery duration is obtained according to a first output result obtained by inputting the list information characteristics of the store into the first model; or, the first estimated distribution time length is obtained by taking the list information characteristic of the store as a first characteristic, filling the first characteristic with the missing characteristic of the store to obtain a second characteristic, inputting the second characteristic into the first model to obtain a first output result, and obtaining the first output result according to the first output result; the missing feature of the shop is determined according to information of the predetermined missing feature, and the information of the missing feature is determined according to the dimension difference between the list information feature of the shop and the order information feature of the shop.
Optionally, the second expected delivery duration is obtained according to a second output result obtained by inputting ordering information characteristics of the ordering data into the first model.
Optionally, the first expected delivery duration of the store is: modifying the value obtained after the first output result is modified according to the list interface distribution time length modification value of the shop; the list interface distribution duration correction value is obtained by using a correction model for estimation according to the list information characteristics of the shop and the first output result; and the correction model is generated according to a second learning sample constructed based on the list information characteristics of the shop and the corresponding ordering information characteristics.
Optionally, the displaying, on the first display interface, the store basic information of the store and the first expected delivery time length of the store includes: obtaining basic information of each shop in a shop list to be displayed on the first display interface and a first expected delivery time of each shop; and displaying the shops in the shop list according to the basic shop information on the first display interface, and displaying the first expected delivery time of the shop in association with the shop.
An embodiment of the present application further provides an electronic device, including: a memory, and a processor; the memory is used for storing a computer program, and the computer program is executed by the processor to execute the method provided by the embodiment of the application.
The embodiment of the present application further provides a storage device, in which a computer program is stored, and the computer program is executed by the processor to perform the method provided in the embodiment of the present application.
Compared with the prior art, the method has the following advantages:
according to the data processing method, the data processing device and the data processing equipment, through acquiring the list information characteristics of a target store, according to the list information characteristics of the target store, a first expected delivery time length which can be associated with the target store and is displayed on a first display interface is obtained by using a first model; the first model is generated according to a first learning sample constructed based on order placing information characteristics of a shop; obtaining order data of a user at the target shop, and obtaining a second expected distribution time length which can be displayed on a second display interface by using the first model according to order placing information characteristics of the order data; wherein a difference between the first estimated delivery duration and the second estimated delivery duration satisfies a preset threshold condition. Since the first estimated delivery time length and the second estimated delivery time length are obtained by using the same model which is generated according to ordering information characteristic learning, the difference between the two estimated delivery time lengths is reduced as much as possible. Furthermore, the features used for calculating the two estimated delivery durations are consistent as much as possible through missing feature filling, and the same module is adopted to call the link, so that the problem that the consistency rate of the two estimated delivery durations is low is solved.
According to the data processing method, the data processing device and the data processing equipment, basic shop information of one or more shops and first expected distribution time of the shops are displayed on a first display interface; receiving order data of a user at a target shop, which is selected by the user from the shops and triggered, and determining the order data of the user at the target shop according to the order information of the target shop; obtaining a second expected distribution time length corresponding to the order data, and displaying the second expected distribution time length on a second display interface; the first estimated distribution time length and the second estimated distribution time length are obtained according to a first model respectively, and the difference between the first estimated distribution time length and the second estimated distribution time length meets a preset threshold condition; the first model is generated according to a first learning sample constructed based on ordering information characteristics of the shop. Since the first estimated delivery time length and the second estimated delivery time length are obtained by using the same model which is generated according to ordering information characteristic learning, the difference between the two estimated delivery time lengths is reduced as much as possible. Furthermore, the characteristics used for calculating the two estimated delivery durations are consistent as much as possible through missing characteristic filling, so that the problem that the consistency rate of the two estimated delivery durations is low is solved.
Drawings
Fig. 1 is a processing flow chart of a data processing method according to a first embodiment of the present application.
Fig. 2 is a processing flow chart of another data processing method according to a second embodiment of the present application.
Fig. 3 is a schematic diagram illustrating a display of a predicted delivery duration according to a second embodiment of the present application.
Fig. 4 is a schematic diagram of a data processing apparatus according to a third embodiment of the present application.
Fig. 5 is a schematic diagram of another data processing apparatus according to a fourth embodiment of the present application.
Fig. 6 is a schematic diagram of an electronic device provided in the present application.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. This application is capable of implementation in many different ways than those herein set forth and of similar import by those skilled in the art without departing from the spirit and scope of this application, and thus this application is not limited to the specific implementations disclosed below.
The embodiment of the application provides two data processing methods and devices, electronic equipment and storage equipment. The following examples are individually set forth.
For ease of understanding, a scenario of the data processing method provided in the embodiment of the present application is first given. In practical application, a user uses a client to perform online shopping through a shopping platform, and a display interface of the client displays a shop list for the user to select a target shop. The first estimated delivery time length of each store in the store list and the corresponding store association are displayed on the store list display interface for the user to refer to in order to make a selection, and the first estimated delivery time length can be understood as the approximate delivery time length of the store which provides objects and delivers to the user at the current time. The object may be a commodity object or a service object. For example, dishes in a take-away scene are an example of an object. When the user selects a target shop and enters the target shop selection object to place an order, a second expected delivery time length corresponding to the current ordering data is displayed on an ordering information display interface, and the second expected delivery time length can be generally used as the committed delivery time length of the shop for the current ordering data. And when the difference between the first estimated distribution time length and the second estimated distribution time length meets a preset threshold value condition, the first estimated distribution time length and the second estimated distribution time length are considered to be consistent. Since the first estimated delivery duration gives a certain psychological expectation to the user, so that the user enters the shop for purchasing the object, if the two are inconsistent or the difference exceeds a preset threshold condition, the user experience may be poor, and even the ordering fails, which affects the conversion rate from the potential user to the actual user. The embodiment of the application provides a data processing method, a first estimated delivery time length and a second estimated delivery time length are estimated by using a shared first model, the difference between the first estimated delivery time length and the second estimated delivery time length is reduced as much as possible, and the consistency rate between the first estimated delivery time length and the second estimated delivery time length is improved.
The data processing method provided in the first embodiment of the present application is described below with reference to fig. 1. The data processing method shown in fig. 1 includes: step S101 to step S102.
Step S101, obtaining list information characteristics of a target shop, and obtaining a first expected delivery time length which can be associated with the target shop and is displayed on a first display interface by using a first model according to the list information characteristics of the target shop; the first model is generated according to a first learning sample constructed based on order placing information characteristics of the shop.
The list information feature may be feature data in which store basic information that can be acquired by a store is displayed in a store list. The store list may be a set of stores queried under a specific condition, or may be a set of stores of a specific store type. For example, the characteristics of the basic information of one or more stores searched for according to the location condition, or the characteristics of the basic information of one or more stores displayed under a certain store category. The basic information of the shop may include, but is not limited to: the system comprises a store identification, a store cover image, a store position, store business hours, and a distance between the store position and a position of a user loading a client; the position of the user can be the current position of the large user obtained under the permission of the user or the position specified by the user. The client is a user side use client corresponding to a shopping platform for providing a transaction link for the shop. Of course, the list information feature may be a feature in contrast to the order placing information feature, and may be feature data of store information that can be acquired when the order placing information is not acquired.
The order placing information feature is feature data related to order placing information of the store. The order placing behavior information and/or order data can be obtained from the historical call logs, the order data can be obtained from the historical order records, and the order placing information characteristics of the shop are determined according to the order placing information and/or order data. Further, the current shop information and/or the current distribution environment information corresponding to the order placing behavior information and/or the order data can be obtained, and the order placing information characteristic of the shop can be determined according to the order placing information and/or the order data and the corresponding current shop information and/or the current distribution environment information. The estimated distribution time length of the shop according to the list information characteristics of the shop can provide reference for the first estimated distribution time length of the shop, so that the user can make a decision whether to order at the shop or not. The first expected delivery time period of the store may be understood as an estimated delivery time period for the view scene of the store. It can be understood that the first expected delivery time length displayed on the first display interface in association with the target store refers to the delivery time length of the target store estimated according to the list information characteristics of the target store, and is displayed in association with the target store, so that a user can conveniently make whether to enter or make a purchase in the target store.
The first display interface, in this embodiment, refers to a display interface on which the client can display the store and the first expected delivery duration of the store in an associated manner. Preferably, the first display interface is a display interface of a shop list page. Of course, the first display interface may also refer to a top page of a store after the user enters one of the stores (for example, the target store) in the store list, and the basic store information and the first expected distribution time of the store are displayed on the top page of the store. Therefore, when the user browses the shop list or browses the shop information, the user can timely know the expected distribution time of the shop and assist in shopping decision.
The first model is a model generated by learning order information characteristics of the shop, and is particularly generated by learning a first learning sample. Wherein the first learning sample is constructed by: obtaining order information characteristics of shops in a historical call log and predicted delivery duration corresponding to the order information characteristics; taking the ordering information characteristic as an input characteristic, taking the expected delivery duration corresponding to the ordering information characteristic as a corresponding output characteristic, and constructing the first learning sample for generating a first model; further, generating the first model using the first learning sample; the first model may be used to estimate a first expected delivery duration for a store based on list information characteristics of the store. In implementation, a Deep fm (depth recommendation) machine learning model is constructed as an initial model, wherein a Deep part is preferably a 3-layer feedforward neural network, an adam optimizer (namely an optimization algorithm) is adopted as a learning algorithm, MAE (Mean Absolute Error) is used as a loss function, and a relu activation function is adopted as a learning algorithm. And inputting the input features of the first learning sample into the Deep part and the input layer of the FM part, training the initial model, and taking the trained target model as a first model. Wherein the Deep portion and the FM portion share input features; the adam optimizer is used for updating variables according to the oscillation condition of the historical gradients in the machine learning model training process and the real historical gradients after filtering oscillation, so that gradient oscillation in the machine learning process is relieved. During training, iteration training is carried out by taking a preset number of samples (4096, for example) as a batch (batch), and the target model is obtained after the preset iteration times (20000 batches, for example) or a loss function meet a convergence condition. Preferably, the related information of the orders of the shop is obtained, and the related information of a plurality of orders can be extracted from a history calling log; order information features are extracted for each order from multiple dimensions. For example, from the transport capacity dimension, transport capacity characteristics such as transport capacity tension degree, grid rider number, current total back order quantity of grid riders, peripheral rider number, peripheral rider assignment probability and back order quantity, and current order degree of peripheral riders are extracted; the grid rider refers to a distribution resource of a distribution area, namely one or more grid distribution areas are obtained after the areas are divided according to grid, and the distribution resource for executing an order distribution task in the distribution area is the grid rider. The peripheral rider means a grid rider of a peripheral delivery area adjacent to each delivery area, which delivers resources around the delivery area. For another example, store busy characteristics such as historical meal delivery duration, current store meal delivery order quantity, store busy, and rider waiting data are extracted from the store dimension. For another example, the order features such as distance, unit price of the customer, dishes, etc. are extracted from the order dimension. For another example, from the dimensions of the distribution environment, the characteristics of the distribution environment such as the weather severity and the road congestion degree are extracted. For another example, delivery difficulty characteristics such as delivery difficulty degree are extracted from the dimension of the delivery position. Determining the order placing information characteristic according to at least one of the capacity characteristic, the shop busyness characteristic, the order characteristic, the delivery environment characteristic and the delivery difficulty characteristic, generating a first learning sample according to the order placing information characteristic and the expected delivery duration corresponding to the order placing information characteristic, and training the initial model by using the first learning sample to obtain a first model.
This step is to obtain a first estimated delivery duration for the targeted store using the first model. In practice, a list of stores is obtained according to specific conditions and/or by store type, and for each store in the list of stores, a first expected delivery time length of the store is obtained using a first model based on list information characteristics of the store. One of the stores in the store list is a target store. When the method is applied to a client, the method comprises the following steps: the method comprises the steps of receiving a request of a client used by a target user for obtaining a shop list, sending the shop list containing the target shop and the first expected delivery time length to the client, and displaying the target shop and the first expected delivery time length on a first display interface of the client in an associated mode. Since the shop list can include a plurality of shops, the shop basic information of each shop can be obtained through a batch data interface, the shop basic information is stored by using a map data structure, the first expected delivery time length of each shop is calculated, the shop basic information of the corresponding shop is obtained from the map, the list information characteristic of the corresponding shop is generated, the first model is input, and the result is output as the first expected delivery time length of the corresponding shop. Since different users may browse the stores at different locations, at least the characteristic data related to the delivery location among the list information characteristics may be different for different users, and it is necessary to calculate the first expected delivery time period of the stores for different users. And the client displays the shops in the shop list according to the basic shop information on the first display interface, and displays the first expected delivery time of the shops in association with the shops.
Step S102, acquiring order data of a user in the target shop, and acquiring a second expected distribution time length which can be displayed on a second display interface by using the first model according to order placing information characteristics of the order data; wherein a difference between the first estimated delivery duration and the second estimated delivery duration satisfies a preset threshold condition.
The order data is data generated by determining the number and attribute information of objects of the selected store by the user entering the store and settling the account. The second expected delivery duration is a delivery duration estimated for the ordering data, and may be understood as a delivery duration estimated for an ordering scene of a user ordering in a shop. The ordering information characteristics which can be used by estimating the second expected delivery time length are richer and more sufficient, so that the ordering information characteristics are closer to the actual delivery time length, and the ordering information characteristics can be displayed on a second display interface of the client as the promised delivery time length in implementation. The second display interface may be a display interface of a menu information page of the client.
In this embodiment, the first model is configured to estimate a first expected delivery time length of a store according to list information characteristics of the store, and/or estimate a second expected delivery time length of order data associated with the store according to order information characteristics of the order data. Because the first estimated delivery time length and the second estimated delivery time length are calculated by using the common model, and the list information characteristic and the order information characteristic which are input into the common model during calculation have characteristic data which are similar or identical to each other to a certain extent, the first estimated delivery time length and the second estimated delivery time length can be relatively close to each other, namely, the difference between the first estimated delivery time length and the second estimated delivery time length meets the preset threshold condition. When the difference between the first estimated delivery time length and the second estimated delivery time length meets a preset threshold condition, the first estimated delivery time length and the second estimated delivery time length can be judged to be consistent. For example, if the preset threshold is 5 minutes, the difference between the two is within 5 minutes, and the two are considered to be consistent.
In practice, although the list information characteristic and the order information characteristic have similar or identical characteristic data to some extent, there is a certain difference. The ordering information feature is generally richer than the list information feature because the ordering information feature can be obtained according to the basic information of the shop and the personalized ordering data of the user when ordering, and the personalized ordering data of the user is not included when the list information feature is generated. Compared with the ordering information characteristic, the list information characteristic has characteristic loss with a certain dimensionality, so that missing characteristic data is adopted for filling the missing characteristic, the input characteristics of the first model are consistent, and the first estimated delivery time length and the second estimated delivery time length which are closer to each other are estimated. Specifically, the obtaining, according to the list information feature of the target store, a first expected delivery duration that can be associated with the target store and shown on a first display interface by using a first model includes: taking the list information characteristic of the target shop as a first characteristic; determining the missing features of the target shop according to the information of the predetermined missing features; the information of the missing features is determined according to the dimension difference between the list information features of the stores and the ordering information features of the stores; filling the first characteristics by using the missing characteristics of the target shop to obtain second characteristics of the target shop; and inputting the second characteristics of the target shop into the first model, and obtaining the first expected delivery time according to the output result of the first model. The step of determining the information of the missing features according to the dimension difference between the list information features of the stores and the order placing information features of the stores refers to analyzing a history calling log of a shopping platform, extracting the list information features of the stores and the order placing information features of order data corresponding to the list information features of the corresponding stores, comparing the dimension difference between the corresponding list information features and the order placing information features, and determining the information of the missing features. And determining the dimension characteristic data of the missing characteristic of the target store according to different dimensions corresponding to the information of the missing characteristic, filling the dimension characteristic data to the list information characteristic, expanding the dimension characteristic data to be the ordering information characteristic of the target store, inputting the ordering information characteristic into a first model, and estimating to obtain the first predicted delivery time length of the target store.
Preferably, the order placing information characteristics of the shop include at least the following dimensional characteristics corresponding to the order placing information: the system comprises a store, a distribution area, a distribution time length dimension, a distribution pressure dimension, a distribution distance dimension, a weather dimension, an object quantity dimension of an order, a price dimension of the order, a store meal length dimension and a user payment time dimension. Compared with the order information characteristic, the information of the missing characteristic corresponding to the list information characteristic of the shop comprises characteristic information of at least the following dimensions: the method comprises the following steps of (1) ordering object quantity dimension, price dimension, shop meal-out duration dimension and user payment time dimension; correspondingly, the determining the missing feature of the target shop according to the predetermined information of the missing feature comprises: taking the average value of the number of objects in the historical time period of the target shop as the missing characteristic data of the target shop corresponding to the dimension of the number of the objects; and/or taking the price average value of the target shop in the historical time period as the missing characteristic data of the price dimension corresponding to the target shop; and/or taking the average meal serving duration of a target store adjacent to the last specified time period at the current moment as missing characteristic data of the target store corresponding to the meal serving duration dimension; and/or taking the visit time of the user visiting the shop list information containing the target shop as the missing characteristic data of the target shop corresponding to the payment time dimension of the user; the populating the first feature with the missing features of the targeted store to obtain a second feature of the targeted store includes: and filling the first characteristic by using the at least one missing characteristic data to obtain the second characteristic. The missing characteristic data can be as close as possible to the value of the dimension characteristic in the average sense, so that the second characteristic is as close as possible to the ordering information characteristic, and the first estimated distribution time length obtained by inputting the first model is consistent with the second estimated distribution time length.
Referring to table 1, the missing feature filling value is shown in table 1, where table 1 shows an example of missing feature data of a restaurant a, which is used to fill the list information feature of the restaurant, and extend the list information feature to be consistent with the dimension of the next order information feature, and the missing feature data is close to the average value of the corresponding dimension feature of the restaurant, so that the extended list information feature is consistent with the next order information feature in terms of both the dimension and the feature value. A restaurant is an example of a store.
TABLE 1 missing feature fill values
Absence feature Fill value
Number of dishes Restaurant A History 7 balance average
Price Restaurant A History 7 balance
Length of meal delivery Restaurant A average meal length in past 10 minutes at the current time
User payment time User access time
Of course, when the first model is generated according to the first learning sample, the dimension of the ordering information feature may be processed according to the available list information features of the stores to select a more reasonable ordering information feature to construct the first learning sample, so that the dimension of the input feature of the generated first model is more reasonable. For example, part of the ordering information feature is deleted to ensure that the list information feature is consistent with the common feature of the ordering information feature to some extent as much as possible.
In this embodiment, the method further includes: and calling the first model to calculate the first expected delivery time length and the second expected delivery time length through the same module calling link. The same link and the shared first model are used, so that the estimated first estimated delivery time length is consistent with the estimated second delivery time length, and further, reasonable missing characteristic values are used for filling the list information characteristics to enable the list information characteristics to be close to the ordering information characteristics, so that the estimated consistency degree of the first estimated delivery time length and the second estimated delivery time length can be improved.
In this embodiment, in order to ensure consistency between the first estimated delivery duration and the second estimated delivery duration, the first model is further modified according to the estimated first estimated delivery duration of the store by using a modification model, and the modified value is used as the estimated delivery duration displayed on the first display interface in association with the store. Specifically, the obtaining, by using a first model according to the list information feature of the target store, a first expected delivery duration that can be displayed on a first display interface in association with the target store includes: inputting the list information characteristic of the target shop into the first model, and taking the output of the first model as the estimated distribution time length to be corrected; acquiring list information characteristics of the target shop, and acquiring a list interface distribution time length correction value of the target shop by using a correction model according to the list information characteristics and the estimated distribution time length to be corrected; the correction model is generated according to a second learning sample constructed based on the list information characteristics of the shop and the corresponding ordering information characteristics; and correcting the estimated distribution time length to be corrected according to the list interface distribution time length correction value of the target shop to obtain the first estimated distribution time length. Wherein the correction model is obtained by: extracting list information characteristics of the shop and corresponding ordering information characteristics from a historical calling log; obtaining a first estimated distribution time length of the shop according to the first model by using the list information characteristics of the shop; obtaining a second expected distribution time length of the shop according to ordering information characteristics of the shop by using the first model; taking a difference between the first expected delivery time length and the second expected delivery time length as a time length difference label of the shop; and constructing a second learning sample for generating the correction model by using the list information characteristics of the shop and the corresponding first expected distribution time length as input characteristics and the corresponding time length difference label as output characteristics, and generating the correction model by using the second learning sample. The modified model can be understood as a post-processing model of the first model. Compared with the ordering information characteristic, the list information characteristic has the characteristic loss problem, and even if the missing characteristic filling is carried out, the estimated delivery time length estimated by using the first model has certain difference in two scenes of shop browsing and ordering. In order to further eliminate the difference, the correction model is added when the first estimated distribution time length of each store is calculated, so that the first estimated distribution time length of each store is more accurate and is closer to the second estimated distribution time length, and the consistency rate of the first estimated distribution time length and the second estimated distribution time length of each store is improved.
For example, a second learning sample of the correction model is generated to extract information from a history call log of a first estimated delivery time period in which the store exists 14 days before the current time and a second estimated delivery time period of the order data. The label is determined using the following formula:
label = first expected delivery duration-second expected delivery duration;
and the list information characteristic of the shop and the first expected delivery time length form an input characteristic, and the correction model is obtained by learning the label and is used for estimating the difference between the first expected delivery time length estimated by the first model and the second expected delivery time length estimated by the first model of the shop. The first expected delivery duration displayed on the first display interface is:
the displayed first estimated delivery duration = the first estimated delivery duration estimated by the first model + the difference estimated by the correction model;
so that the first predicted delivery time length displayed on the first display interface is consistent with the second predicted delivery time length displayed on the second display interface to a higher degree.
In this embodiment, the consistency of the first estimated delivery duration and the second estimated delivery duration of each store of the shopping platform as a whole can be measured by the consistency rate. The method provided by the embodiment can improve the consistency rate, so that the user experience of the shopping platform can be improved, and the user viscosity is increased. The consistency ratio is specifically defined by the following formula:
the rate of agreement = the difference between the first and second expected delivery durations is less than 5 minutes of order quantity/total order quantity.
The method provided by this embodiment may be applied to a server, and if the method is applied to the server, the method further includes: receiving a request of a client used by a target user for obtaining a shop list, and sending the shop list containing the target shop and the first expected delivery time length to the client so as to enable the target shop and the first expected delivery time length to be displayed on a first display interface of the client in an associated manner; and receiving a settlement request of the target user for the order data, and sending the second expected delivery duration to the client so that the second expected delivery duration and the settlement information of the order data are displayed on a second display interface of the client. The first expected distribution time period represents an approximate distribution time period in the general sense of the store, and is not specific to the order data of a certain user. The second estimated delivery time length represents a more specific estimated delivery time length of the store, and the order data specific to a certain user can be understood as an estimated delivery time length of the user of the store. The first estimated delivery duration is of reference to the purchase decision of the user, and the second estimated delivery duration can be generally used as a committed delivery duration of the order for the shop after ordering by the user.
It is understood that the first model of the present embodiment may be generated by extracting order information features based on the order information log. In a ordering scene of a user, a second expected delivery time length is estimated by using the first model according to ordering information characteristics of ordering data, and the second expected delivery time length is displayed on a second display interface of the client and can be specifically used as committed delivery time for ordering data to be displayed. In a scene that a user browses a shop, a first expected delivery time length is output by using a first model according to the list information characteristics of the shop. Preferably, the list information characteristic and the first predicted distribution time length output by the first model are input into the correction model to obtain a correction value, the correction value is added to the first predicted distribution time length output by the first model to obtain a corrected first predicted distribution time length, the corrected first predicted distribution time length is displayed on the first display interface of the client, and the corrected first predicted distribution time length can be closer to the second predicted distribution time length, so that the consistency rate of the two is improved. The shared first model can ensure the consistency rate of the first estimated delivery time length and the second estimated delivery time length to a certain extent, and the consistency rate can be improved by modifying the model, so that the user experience is improved, the purchase willingness of a user is increased, and the user conversion rate from a potential user to an actual user is improved.
It should be noted that, in the case of no conflict, the features given in this embodiment and other embodiments of the present application may be combined with each other, and the steps S101 and S102 or similar terms do not limit the steps to be executed sequentially.
Thus, the method provided in this embodiment is described, in which the first predicted delivery time length and the second predicted delivery time length are obtained by using the same model, and the model is generated by learning according to the ordering information characteristics, so that the difference between the two predicted delivery time lengths is reduced as much as possible. Furthermore, the features used for calculating the two estimated delivery durations are consistent as much as possible through missing feature filling, and the same module is adopted to call the link, so that the problem that the consistency rate of the two estimated delivery durations is low is solved.
Based on the foregoing embodiment, a second embodiment of the present application provides another data processing method, which is described below with reference to fig. 2 and 3, and related portions may refer to the corresponding portions for description. Referring to fig. 2, the data processing method shown in the figure includes: step S201 to step S203.
Step S201, store basic information of one or more stores and a first expected delivery time of the stores are displayed on a first display interface.
The method provided by the embodiment is applied to the client. This step is to display the first expected delivery duration. Specifically, when the shop is displayed on the shop list page, the first expected delivery time corresponding to the shop can be displayed in association with the shop. The store list page is a scene of store browsing and can assist the user in making a decision whether to enter the store. If the user selects one store and enters the store page of the store, the first expected distribution time length can be displayed in the store page, so that the user can know the general distribution time length of the store conveniently, and a decision of shopping in the store is assisted. Wherein, show the shop basic information in the shop at first display interface to and the first expected delivery duration in shop, include: obtaining basic information of each shop in a shop list to be displayed on the first display interface and a first expected delivery time of each shop; and displaying the shops in the shop list according to the shop basic information on the first display interface, and displaying the first expected delivery time length of the shop in a manner of being associated with the shop.
In this embodiment, the first expected delivery duration is obtained according to a first output result obtained by inputting the list information characteristics of the store into a first model; or, the first expected delivery duration is obtained according to a first output result, wherein the first expected delivery duration takes the list information characteristic of the store as a first characteristic, a second characteristic is obtained by filling the first characteristic with the missing characteristic of the store, and the second characteristic is input into the first model to obtain the first output result; the missing feature of the shop is determined according to information of the predetermined missing feature, and the information of the missing feature is determined according to the dimension difference between the list information feature of the shop and the order information feature of the shop. The first model is generated according to a first learning sample constructed based on ordering information characteristics of the shop.
Further, the first estimated delivery duration of the store is: modifying the first output result according to the list interface distribution time length modification value of the shop to obtain a value; the list interface distribution duration correction value is obtained by estimating through a correction model according to the list information characteristics of the shop and the first output result; and the correction model is generated according to a second learning sample constructed based on the list information characteristics of the shop and the corresponding ordering information characteristics. The corrected first estimated distribution time length can be closer to a second estimated distribution time length displayed in a subsequent step, so that the consistency rate of the first estimated distribution time length and the second estimated distribution time length is improved, and the user experience is improved.
Step S202, receiving order data of a user in a trigger target shop selected from the shops by the user, and determining the order data of the user in the target shop according to the order information aiming at the target shop.
The method comprises the steps of obtaining ordering data of a user in a target shop. In the implementation process, the user can select and trigger the target shop to enter the target shop according to the requirement of the user by referring to the first estimated distribution time length of the shop shown on the first display interface. And receiving point single line information of a user aiming at the target shop on a point single page of the target shop displayed by the client, and determining the point single data.
Step S203, obtaining a second expected delivery duration corresponding to the order data, and displaying the second expected delivery duration on a second display interface; the first estimated distribution time length and the second estimated distribution time length are obtained according to a first model respectively, and the difference between the first estimated distribution time length and the second estimated distribution time length meets a preset threshold condition; the first model is generated according to a first learning sample constructed based on ordering information characteristics of the shop.
This step is to display a second expected delivery duration corresponding to the order data. When the system is implemented, the user adds the object selected by the order to the shopping cart, the settlement is triggered, a settlement request is sent to the server, the server returns the second expected distribution duration to the client, and the client displays the second expected distribution duration on the second display interface. The server may also provide the second expected delivery duration and the information of the confirmation order data to the client, display the order data and the corresponding second expected delivery duration on the order information page, and display the expected delivery duration as the promised delivery duration of the order. Preferably, the second expected delivery duration is obtained by inputting order placing information characteristics of the order placing data into the first model to obtain a second output result and according to the second output result.
In this embodiment, the first expected delivery duration and the second expected delivery duration are estimated by using the first model in common, so that the consistency rate of the first expected delivery duration and the second expected delivery duration can be ensured to a certain extent. The first estimated delivery time length estimated by the first model is further corrected through the correction model, and the corrected first estimated delivery time length is used as the displayed first estimated delivery time length, so that the consistency rate of the two displayed estimated delivery time lengths can be improved, the purchase willingness of a user can be increased, and the user conversion rate from a potential user to an actual user is improved.
Referring to fig. 3, a schematic diagram of a display of a first expected delivery duration and a second expected delivery duration is shown, which includes: a first estimated delivery duration 301, and a second estimated delivery duration 302. The first estimated delivery time 301 is a time period between the time of the user using the client and the time of the store, and is a time period between the time of the user using the client and the time of the store. The user location refers to the current location of the user obtained under the condition that the user allows, or the location or the delivery location specified by the user. The store list page is a scene of store browsing and can assist the user in making a decision whether to enter the store. If the user selects one shop and enters the shop page of the shop, the first expected delivery time length can be displayed in the shop page, so that the user can know the general delivery time length of the shop and assist in making a decision of shopping in the shop. A second estimated delivery duration 302, which is an estimated delivery duration for the ordering data, displayed on the ordering information page of the user, and may be used as a committed delivery duration of the order.
It is understood that the interface layout and the size, shape, appearance style, etc. of each interface element shown in fig. 3 are illustrative examples, and are not intended to limit the methods provided by the embodiments of the present application.
Thus, the method provided in this embodiment is described, in which the first estimated delivery time length and the second estimated delivery time length are obtained by using the same model, and the model is generated by learning according to the ordering information characteristics, so that the difference between the two estimated delivery time lengths is reduced as much as possible. Furthermore, the characteristics used for calculating the two estimated distribution time lengths are consistent as much as possible through missing characteristic filling, so that the problem that the consistency rate of the two estimated distribution time lengths is low is solved.
Corresponding to the first embodiment, a third embodiment of the present application provides a data processing apparatus, and please refer to the description of the corresponding method embodiment for related portions. Referring to fig. 4, the data processing apparatus shown in the figure includes:
the first estimation unit 401 is configured to obtain list information characteristics of a target store, and obtain, according to the list information characteristics of the target store, a first estimated delivery time length which can be displayed on a first display interface in association with the target store by using a first model; the first model is generated according to a first learning sample constructed based on ordering information characteristics of a shop;
a second estimation unit 402, configured to obtain order data of a user at the target store, and obtain, according to order placing information characteristics of the order data, a second estimated delivery duration that can be displayed on a second display interface by using the first model; wherein a difference between the first estimated delivery duration and the second estimated delivery duration satisfies a preset threshold condition.
Optionally, the apparatus further includes a first model generating unit, where the first model generating unit is configured to: obtaining order placing information characteristics of shops in a historical calling log and estimated distribution duration corresponding to the order placing information characteristics; taking the ordering information characteristic as an input characteristic, taking the expected delivery duration corresponding to the ordering information characteristic as a corresponding output characteristic, and constructing the first learning sample for generating a first model; generating the first model using the first learning sample; the first model is used for predicting a first predicted delivery time length of a shop according to the list information characteristics of the shop and/or predicting a second predicted delivery time length of the order data according to the order information characteristics of the order data associated with the shop.
Optionally, the first estimating unit is specifically configured to: inputting the list information characteristics of the target shop into the first model, and taking the output of the first model as the estimated distribution time length to be corrected; obtaining list information characteristics of the target shop, and obtaining a list interface distribution time length correction value of the target shop by using a correction model according to the list information characteristics and the estimated distribution time length to be corrected; the correction model is generated according to a second learning sample constructed based on the list information characteristics of the shop and the corresponding ordering information characteristics; and correcting the estimated distribution time length to be corrected according to the list interface distribution time length correction value of the target shop to obtain the first estimated distribution time length.
Optionally, the apparatus further includes a modified model generating unit, where the modified model generating unit is configured to: extracting list information characteristics of the shop and corresponding ordering information characteristics from a historical calling log; obtaining a first estimated distribution time length of the shop according to the first model by using the list information characteristics of the shop; obtaining a second expected distribution time length of the shop according to ordering information characteristics of the shop by using the first model; taking a difference between the first expected delivery duration and the second expected delivery duration as a duration difference tag for the store; and constructing a second learning sample for generating the correction model by using the list information characteristics of the shop and the corresponding first expected distribution time length as input characteristics and the corresponding time length difference label as output characteristics, and generating the correction model by using the second learning sample.
Optionally, the first estimating unit is specifically configured to: taking the list information characteristic of the target shop as a first characteristic; determining the missing features of the target shop according to the information of the predetermined missing features; the information of the missing characteristics is determined according to the dimension difference between the list information characteristics of the shops and the ordering information characteristics of the shops; filling the first characteristics by using the missing characteristics of the target shop to obtain second characteristics of the target shop; and inputting the second characteristics of the target shop into the first model, and obtaining the first expected delivery time length according to the output result of the first model.
Optionally, the information of the missing features includes feature information of at least the following dimensions: the method comprises the steps of ordering, wherein the order comprises an object quantity dimension, a price dimension, a shop meal-serving time length dimension and a user payment time dimension; the first estimating unit is specifically configured to: taking the average value of the number of objects in the historical time period of the target shop as the missing characteristic data of the target shop corresponding to the dimension of the number of the objects; and/or taking the price average value of the target shop in the historical time period as the missing characteristic data of the price dimension corresponding to the target shop; and/or taking the average meal serving duration of a target store adjacent to the last specified time period at the current moment as missing characteristic data of the target store corresponding to the meal serving duration dimension; and/or taking the visit time of the user visiting the shop list information containing the target shop as the missing characteristic data of the target shop corresponding to the payment time dimension of the user; and filling the first characteristic by using the at least one missing characteristic data to obtain the second characteristic.
The order information characteristics of the shop comprise at least the following dimension characteristics corresponding to the order information: the system comprises a store, a distribution area, a distribution time length dimension, a distribution pressure dimension, a distribution distance dimension, a weather dimension, an object quantity dimension of orders, a price dimension of orders, a store meal length dimension and a user payment time dimension.
Optionally, the apparatus further includes a calling unit, where the calling unit is configured to: and calling the first model to calculate the first expected delivery time length and the second expected delivery time length through the same module calling link.
Optionally, the apparatus further includes a communication unit, where the communication unit is configured to receive a request from a client used by a target user to obtain a store list, and send the store list including the target store and the first expected delivery time length to the client, so that the target store and the first expected delivery time length are displayed on a first display interface of the client in association with each other; and receiving a settlement request of the target user for the order data, and sending the second expected delivery duration to the client so that the second expected delivery duration and the settlement information of the order data are displayed on a second display interface of the client.
A fourth embodiment of the present application provides a data processing apparatus corresponding to the second embodiment, and please refer to the description of the corresponding method embodiment for related parts. Referring to fig. 5, the data processing apparatus shown in the figure includes:
a first display unit 501, configured to display, on a first display interface, basic store information of one or more stores and a first expected delivery time of the stores;
a receipt unit 502, configured to receive receipt data that a user selects a trigger target store from the stores, and determines that the user is at the target store according to a receipt line for information of the target store;
a second display unit 503, configured to obtain a second estimated distribution duration corresponding to the order data, and display the second estimated distribution duration on a second display interface;
the first estimated distribution time length and the second estimated distribution time length are obtained according to a first model respectively, and the difference between the first estimated distribution time length and the second estimated distribution time length meets a preset threshold condition; the first model is generated according to a first learning sample constructed based on ordering information characteristics of the shop.
Optionally, the first expected delivery duration is obtained according to a first output result obtained by inputting the list information characteristics of the store into the first model; or, the first expected delivery duration is obtained according to a first output result, wherein the first expected delivery duration takes the list information characteristic of the store as a first characteristic, a second characteristic is obtained by filling the first characteristic with the missing characteristic of the store, and the second characteristic is input into the first model to obtain the first output result; the missing characteristics of the shops are determined according to the information of the predetermined missing characteristics, and the information of the missing characteristics is determined according to the dimension difference between the list information characteristics of the shops and the ordering information characteristics of the shops.
Optionally, the second expected delivery duration is obtained according to a second output result obtained by inputting ordering information characteristics of the order data into the first model.
Optionally, the first expected delivery duration of the store is: modifying the value obtained after the first output result is modified according to the list interface distribution time length modification value of the shop; the list interface distribution duration correction value is obtained by estimating through a correction model according to the list information characteristics of the shop and the first output result; and the correction model is generated according to a second learning sample constructed based on the list information characteristics of the shop and the corresponding ordering information characteristics.
Optionally, the first display unit 501 is specifically configured to: obtaining basic information of each shop in a shop list to be displayed on the first display interface and a first expected delivery time of each shop; and displaying the shops in the shop list according to the basic shop information on the first display interface, and displaying the first expected delivery time of the shop in association with the shop.
Based on the above embodiments, a fifth embodiment of the present application provides an electronic device, and please refer to the corresponding description of the above embodiments for related parts. Referring to fig. 6, the electronic device shown in the figure includes: a memory 601, and a processor 602; the memory is used for storing a computer program, and the computer program is executed by the processor to execute the method provided by the embodiment of the application.
Based on the foregoing embodiments, a sixth embodiment of the present application provides a storage device, and please refer to the corresponding description of the foregoing embodiments for related parts. The schematic diagram of the storage device is similar to fig. 6. The storage device stores a computer program, and the computer program is executed by the processor to execute the method provided by the embodiment of the application.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
1. Computer-readable media, including both permanent and non-permanent, removable and non-removable media, may implement the information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include non-transitory computer readable media (transient media), such as modulated data signals and carrier waves.
2. As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
Although the present application has been described with reference to the preferred embodiments, it is not intended to limit the present application, and those skilled in the art can make variations and modifications without departing from the spirit and scope of the present application, therefore, the scope of the present application should be determined by the claims that follow.

Claims (16)

1. A data processing method, comprising:
the method comprises the steps of obtaining list information characteristics of a target shop, filling the list information characteristics according to dimension difference between the list information characteristics of the shop and ordering information characteristics of the shop, and obtaining a first estimated distribution time length which can be associated with the target shop and is shown on a first display interface by using a first model according to the filled list information characteristics of the target shop; the first model is generated according to a first learning sample constructed based on ordering information characteristics of a shop;
obtaining order data of a user in the target shop, and obtaining second predicted delivery time length which can be displayed on a second display interface by using the first model according to order information characteristics of the order data;
wherein a difference between the first estimated delivery duration and the second estimated delivery duration satisfies a preset threshold condition.
2. The method of claim 1, further comprising:
obtaining order information characteristics of shops in a historical call log and predicted delivery duration corresponding to the order information characteristics;
taking the ordering information characteristic as an input characteristic, taking the expected delivery duration corresponding to the ordering information characteristic as a corresponding output characteristic, and constructing the first learning sample for generating a first model;
generating the first model using the first learning sample; the first model is used for predicting a first predicted delivery time length of a shop according to the list information characteristics of the shop and/or predicting a second predicted delivery time length of the order data according to the order information characteristics of the order data associated with the shop.
3. The method as claimed in claim 1, wherein the obtaining a first expected delivery duration which can be shown in a first display interface in association with the target shop by using a first model according to the populated list information characteristics of the target shop comprises:
inputting the list information characteristics of the target shop into the first model, and taking the output of the first model as the estimated distribution time length to be corrected;
obtaining list information characteristics of the target shop, and obtaining a list interface distribution time length correction value of the target shop by using a correction model according to the list information characteristics and the estimated distribution time length to be corrected; the correction model is generated according to a second learning sample constructed based on the list information characteristics of the shop and the corresponding ordering information characteristics;
and correcting the estimated distribution time length to be corrected according to the list interface distribution time length correction value of the target shop to obtain the first estimated distribution time length.
4. The method of claim 3, further comprising:
extracting list information characteristics of the shop and corresponding ordering information characteristics from a historical calling log;
obtaining a first estimated distribution time length of the shop according to the first model by using the list information characteristics of the shop;
obtaining a second expected distribution time length of the shop according to ordering information characteristics of the shop by using the first model;
taking a difference between the first expected delivery time length and the second expected delivery time length as a time length difference label of the shop;
and constructing a second learning sample for generating the correction model by using the list information characteristics of the shop and the corresponding first expected delivery time length as input characteristics and the corresponding time length difference label as output characteristics, and generating the correction model by using the second learning sample.
5. The method of claim 1, wherein populating list information features of a store according to dimensional differences between the list information features and order information features of a store comprises:
taking the list information characteristic of the target shop as a first characteristic;
determining the missing features of the target shop according to the information of the predetermined missing features; the information of the missing features is determined according to the dimension difference between the list information features of the stores and the ordering information features of the stores;
filling the first characteristics with the missing characteristics of the target shop to obtain second characteristics of the target shop;
and taking the second characteristic of the target shop as the filled list information characteristic of the target shop.
6. The method of claim 5, wherein the missing feature information comprises feature information in at least the following dimensions:
the method comprises the following steps of (1) ordering object quantity dimension, price dimension, shop meal-out duration dimension and user payment time dimension;
the determining the missing feature of the target shop according to the predetermined information of the missing feature comprises the following steps:
taking the average value of the number of objects in the historical time period of the target shop as the missing characteristic data of the target shop corresponding to the dimension of the number of the objects; and/or the presence of a gas in the atmosphere,
taking the price average value of the target store in the historical time period as the missing characteristic data of the target store corresponding to the price dimension; and/or the presence of a gas in the atmosphere,
taking the average meal serving duration of a target store adjacent to a last specified time period at the current moment as missing characteristic data of the target store corresponding to the meal serving duration dimension; and/or the presence of a gas in the atmosphere,
taking the visit time of the user visiting the shop list information containing the target shop as the missing characteristic data of the target shop corresponding to the payment time dimension of the user;
the populating the first feature with the missing features of the targeted store to obtain a second feature of the targeted store, comprising:
and filling the first characteristic by using the at least one missing characteristic data to obtain the second characteristic.
7. The method of claim 1, wherein the order information characteristics of the store comprise characteristics corresponding to order information in at least the following dimensions:
the system comprises a store, a distribution area, a distribution time length dimension, a distribution pressure dimension, a distribution distance dimension, a weather dimension, an object quantity dimension of orders, a price dimension of orders, a store meal length dimension and a user payment time dimension.
8. The method of claim 1, further comprising:
and calling the first model to calculate the first expected delivery time length and the second expected delivery time length through the same module calling link.
9. The method of claim 1, comprising:
receiving a request of a client used by a target user for obtaining a shop list, and sending the shop list containing the target shop and the first expected delivery time length to the client so as to enable the target shop and the first expected delivery time length to be displayed on a first display interface of the client in an associated manner;
and receiving a settlement request of the target user for the order data, and sending the second expected distribution time length to the client so that the second expected distribution time length and the settlement information of the order data are displayed on a second display interface of the client.
10. A data processing method, comprising:
displaying basic shop information of one or more shops and a first expected delivery time length of the shops on a first display interface;
receiving order data of a user in a target shop, which is selected by the user from the shops and triggered, and determining the order data of the user in the target shop according to the order information of the target shop;
obtaining a second estimated distribution time length corresponding to the order data, and displaying the second estimated distribution time length on a second display interface;
the first estimated distribution time length and the second estimated distribution time length are obtained according to a first model respectively, and the difference between the first estimated distribution time length and the second estimated distribution time length meets a preset threshold condition; the first model is generated according to a first learning sample constructed based on ordering information characteristics of a shop;
wherein obtaining the first expected delivery duration according to a first model comprises:
filling the list information characteristics of the stores according to the dimension difference between the list information characteristics of the stores and the ordering information characteristics of the stores, inputting the filled list information characteristics of the stores into the first model to obtain a first output result, and obtaining the first estimated distribution time length according to the first output result.
11. The method of claim 10, wherein populating list information features of a store according to dimensional differences between the list information features of the store and ordering information features of a store, inputting populated list information features of the store into the first model to obtain a first output comprising:
taking the list information characteristic of the shop as a first characteristic, filling the first characteristic with the missing characteristic of the shop to obtain a second characteristic, and inputting the second characteristic into the first model to obtain a first output result; the missing feature of the shop is determined according to information of the predetermined missing feature, and the information of the missing feature is determined according to the dimension difference between the list information feature of the shop and the order information feature of the shop.
12. The method of claim 10, wherein the second expected delivery duration is obtained based on a second output result obtained by inputting ordering information characteristics of the ordering data into the first model.
13. The method of claim 10, wherein the first estimated delivery duration for the store is:
modifying the first output result according to the list interface distribution time length modification value of the shop to obtain a value;
the list interface distribution duration correction value is obtained by estimating through a correction model according to the list information characteristics of the shop and the first output result; and the correction model is generated according to a second learning sample constructed based on the list information characteristics of the shop and the corresponding ordering information characteristics.
14. The method of claim 10, wherein the presenting, on a first display, store basic information for a store and a first estimated delivery duration for the store comprises:
obtaining basic information of each shop in a shop list to be displayed on the first display interface and a first expected delivery time of each shop;
and displaying the shops in the shop list according to the shop basic information on the first display interface, and displaying the first expected delivery time length of the shop in a manner of being associated with the shop.
15. An electronic device, comprising:
a memory, and a processor; the memory is adapted to store a computer program which, when executed by the processor, performs the method of any one of claims 1 to 14.
16. A storage device, characterized in that a computer program is stored which, when being executed by a processor, performs the method of any one of claims 1 to 14.
CN202210367301.2A 2022-04-08 2022-04-08 Data processing method and equipment Active CN114463103B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210367301.2A CN114463103B (en) 2022-04-08 2022-04-08 Data processing method and equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210367301.2A CN114463103B (en) 2022-04-08 2022-04-08 Data processing method and equipment

Publications (2)

Publication Number Publication Date
CN114463103A CN114463103A (en) 2022-05-10
CN114463103B true CN114463103B (en) 2022-07-15

Family

ID=81418006

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210367301.2A Active CN114463103B (en) 2022-04-08 2022-04-08 Data processing method and equipment

Country Status (1)

Country Link
CN (1) CN114463103B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2014207813A1 (en) * 2013-06-24 2014-12-31 楽天株式会社 List presentation device, list presentation method, and program
CN110490356A (en) * 2019-07-02 2019-11-22 北京星选科技有限公司 Interface display method, device, electronic equipment
CN112036800A (en) * 2020-09-07 2020-12-04 上海明略人工智能(集团)有限公司 Meal distribution method and device
CN112862133A (en) * 2019-11-12 2021-05-28 北京三快在线科技有限公司 Order processing method and device, readable storage medium and electronic equipment
CN113129053A (en) * 2021-03-29 2021-07-16 北京沃东天骏信息技术有限公司 Information recommendation model training method, information recommendation method and storage medium
CN113888229A (en) * 2021-10-19 2022-01-04 拉扎斯网络科技(上海)有限公司 Store data processing and order processing method and device
CN114239977A (en) * 2021-12-21 2022-03-25 北京三快在线科技有限公司 Method, device, equipment and storage medium for determining estimated delivery time length

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110415077A (en) * 2019-07-24 2019-11-05 万翼科技有限公司 Data processing method, server, terminal and storage medium based on the date
CN112770187B (en) * 2020-12-23 2022-11-11 口碑(上海)信息技术有限公司 Shop data processing method and device

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2014207813A1 (en) * 2013-06-24 2014-12-31 楽天株式会社 List presentation device, list presentation method, and program
CN110490356A (en) * 2019-07-02 2019-11-22 北京星选科技有限公司 Interface display method, device, electronic equipment
CN112862133A (en) * 2019-11-12 2021-05-28 北京三快在线科技有限公司 Order processing method and device, readable storage medium and electronic equipment
CN112036800A (en) * 2020-09-07 2020-12-04 上海明略人工智能(集团)有限公司 Meal distribution method and device
CN113129053A (en) * 2021-03-29 2021-07-16 北京沃东天骏信息技术有限公司 Information recommendation model training method, information recommendation method and storage medium
CN113888229A (en) * 2021-10-19 2022-01-04 拉扎斯网络科技(上海)有限公司 Store data processing and order processing method and device
CN114239977A (en) * 2021-12-21 2022-03-25 北京三快在线科技有限公司 Method, device, equipment and storage medium for determining estimated delivery time length

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Customer Satisfaction Evaluation of Food Delivery Platforms-Taking Meituan as an Example;Chuanpeng Wang;《2020 International Conference on Big Data Economy and Information Management (BDEIM)》;20201213;全文 *
O2O外卖配送预计送达时间决策模式的选择策略;赵道致等;《工业工程与管理》;20181010(第05期);全文 *
基于BP神经网络的网店销售预测模型研究;翁莹晶等;《闽江学院学报》;20160925(第05期);全文 *

Also Published As

Publication number Publication date
CN114463103A (en) 2022-05-10

Similar Documents

Publication Publication Date Title
CN107180371B (en) Method, system and computer-readable storage medium for purchasing goods using coupons
US20200065750A1 (en) Inventory management system and method thereof
US20110302030A1 (en) System and Method for Incorporating Packaging and Shipping Ramifications on Net Profit/Loss When Up-Selling
CN111612594A (en) Localized merchant retail system
CN113095893A (en) Method and device for determining sales of articles
CN112948521B (en) Object handling method and device
US20210109906A1 (en) Clustering model analysis for big data environments
CN111325587A (en) Method and apparatus for generating information
CN111861605A (en) Business object recommendation method
CN109711917A (en) Information-pushing method and device
CN113204712A (en) Information pushing method, device, medium and program product based on community service
CN110347887B (en) Method and device for acquiring time sequence data of service scene
CN113793081A (en) Storage monitoring method and device, computer readable medium and electronic equipment
CN114663015A (en) Replenishment method and device
JP5794881B2 (en) Information processing apparatus, information processing method, and information processing program
CN109978421B (en) Information output method and device
US20230367768A1 (en) Anonymization of query information while retaining query structure and sizing information
CN113298610A (en) Information recommendation and acquisition method, equipment and storage medium
CN114463103B (en) Data processing method and equipment
US20170249697A1 (en) System and method for machine learning based line assignment
CN116595390A (en) Commodity information processing method and electronic equipment
CN112017000B (en) Commodity information pushing method, device, equipment and storage medium
CN110033292A (en) Information output method and device
US11282126B1 (en) Learning staple goods for a user
CN110515946B (en) Data extraction method, device, equipment and computer readable storage medium

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
GR01 Patent grant
GR01 Patent grant