WO2019218668A1 - 配送范围的确定 - Google Patents

配送范围的确定 Download PDF

Info

Publication number
WO2019218668A1
WO2019218668A1 PCT/CN2018/122085 CN2018122085W WO2019218668A1 WO 2019218668 A1 WO2019218668 A1 WO 2019218668A1 CN 2018122085 W CN2018122085 W CN 2018122085W WO 2019218668 A1 WO2019218668 A1 WO 2019218668A1
Authority
WO
WIPO (PCT)
Prior art keywords
merchant
merchants
blocks
geographic
geographical
Prior art date
Application number
PCT/CN2018/122085
Other languages
English (en)
French (fr)
Inventor
丁雪涛
张润丰
贾东
何仁清
郭振刚
郝小菠
Original Assignee
北京三快在线科技有限公司
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 北京三快在线科技有限公司 filed Critical 北京三快在线科技有限公司
Priority to BR112020023257-2A priority Critical patent/BR112020023257A2/pt
Priority to US17/055,930 priority patent/US20210312486A1/en
Publication of WO2019218668A1 publication Critical patent/WO2019218668A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G06Q30/0204—Market segmentation
    • G06Q30/0205—Location or geographical consideration
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/04—Inference or reasoning models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083—Shipping
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083—Shipping
    • G06Q10/0838—Historical data

Definitions

  • the present application relates to a method and apparatus for determining a distribution range in the field of network technology, an electronic device, and a storage medium.
  • each merchant has its own distribution range.
  • the scope of distribution of a merchant is an area of geographical concept.
  • the merchant is only visible to users located within the delivery range of the merchant.
  • the order relationship only generates users within the scope of the merchant and distribution.
  • the scope of merchant distribution can affect the business volume, distribution efficiency and user experience. If the distribution range is set too small, the potential user group is small, the merchant quantity, platform gross GMV (gross merchandise volume) will be small; if the distribution range is set too large, although the potential user group is large, resulting The quantity of the order may be improved to some extent, but the overall distribution efficiency may be reduced, thereby affecting the user experience.
  • the embodiment of the present application provides a method, a device, an electronic device, and a storage medium for determining a distribution range, which can improve the distribution efficiency while ensuring the overall benefit in the area.
  • a method for determining a delivery range comprising:
  • a delivery range is determined for each of the merchants based on the target merchant set of each of the geographic blocks.
  • a delivery range determining device comprising:
  • a data acquisition module configured to acquire historical behavior data in multiple geographical blocks and historical order data of multiple merchants
  • a target merchant set obtaining module configured to determine, according to historical behavior data in the plurality of geographical blocks and historical order data of the plurality of merchants, a target merchant set of each of the geographic blocks;
  • a delivery scope determining module configured to determine a delivery range for each of the merchants based on the target merchant set of each of the geographic blocks.
  • a computer apparatus comprising a processor and a memory, the memory storing executable instructions loaded by the processor and causing the processor to perform the dispensing Range determination method.
  • a computer readable storage medium having stored therein executable instructions loaded by a processor and causing the processor to perform the distribution range determination method described above.
  • the technical solution provided by the embodiment of the present application finds that the geographical block can be made for the geographical block from the perspective of the geographical block by making full use of the historical behavior data in the geographical block and the historical order data of the merchant, and by an automated method.
  • the collection of merchants with higher overall internal returns can not only ensure the overall income of the geographical blocks, but also improve the distribution efficiency.
  • FIG. 1 is a flowchart of a method for determining a delivery range according to an embodiment of the present application
  • FIG. 2 is a flowchart of a method for determining a delivery range according to an embodiment of the present application
  • FIG. 3 is a schematic diagram of a prediction process provided by an embodiment of the present application.
  • FIG. 4 is a schematic diagram of a process for determining a delivery range according to an embodiment of the present application.
  • FIG. 5 is a schematic diagram of optimization of a delivery range according to an embodiment of the present application.
  • FIG. 6 is a schematic structural diagram of a distribution range determining apparatus according to an embodiment of the present application.
  • FIG. 7 is a schematic structural diagram of a computer device according to an embodiment of the present application.
  • FIG. 1 is a flowchart of a method for determining a delivery range according to an embodiment of the present application.
  • the method can be applied to a server of an instant delivery application platform. Referring to Figure 1, the method includes steps 101-103:
  • step 101 historical behavior data in a plurality of geographical blocks and historical order data of a plurality of merchants are acquired.
  • a target merchant set of each of the geographic blocks is determined according to historical behavior data in the plurality of geographic blocks and historical order data of the plurality of merchants.
  • a delivery range is determined for each of the merchants based on the target merchant set of each of the geographic blocks.
  • determining, according to historical behavior data in the plurality of geographic blocks and historical order data of the plurality of merchants, the target merchant set of each of the geographic blocks includes:
  • predicting, according to historical behavior data in the plurality of geographic blocks and historical order data of the plurality of merchants, a conversion rate or an order quantity of each of the merchants on each of the geographic blocks includes: :
  • the historical order data of the plurality of merchants and the historical behavior data of the plurality of geographical blocks are input into the prediction model, and the conversion rate or the order quantity of each merchant in each geographical block is output.
  • the training process of the predictive model includes:
  • the first feature includes at least one of an exposure amount and a click amount of the merchant dimension, and a conversion rate or an order quantity of the merchant dimension;
  • the second feature includes at least one of an exposure amount and a click amount of the geographic block dimension And a conversion rate or order quantity of the geographic block dimension;
  • the third feature includes at least one of exposure and click volume of the cross-dimension of the merchant and the geographic block, and conversion of the cross-dimension of the merchant and the geographic block Rate or order quantity.
  • historical order data of the plurality of merchants, and historical behavior data in the plurality of geographic blocks Determining, by the target merchant set of each of the geographic blocks, the following:
  • historical order data of the plurality of merchants, and historical behavior data in the plurality of geographic blocks Determining, by the target merchant set of each of the geographic blocks, the following:
  • the plurality of merchants are combined and optimized according to the number of orders of each merchant on each of the geographical blocks and the average customer unit price of each of the merchants, to obtain a target merchant set of each of the geographic blocks.
  • the plurality of merchants are combined according to the conversion rate of each merchant on the geographic block, the exposure amount of each geographic block, and the average customer unit price of each merchant.
  • Optimizing, obtaining the target merchant set of each of the geographic blocks includes:
  • the first target optimization function is the first target optimization function
  • g is the geographic block index
  • M is the number of the geographical blocks
  • p is the merchant index
  • N is the number of the merchants
  • pv g is the exposure amount of the geographic block g
  • cvr p, g is the merchant p
  • the conversion rate on the geographic block g, order p, g is the predicted order quantity of the merchant p on the geographic block g
  • Price p is the average customer unit price of the merchant p
  • C p, g is whether the geographical area is allocated to the merchant p
  • Block g is the 0-1 identifier of the block in its distribution range; when C p, g is 1, it means that the merchant p is allocated the geographical block g, and C p, g is 0, indicating that it is not a merchant.
  • p allocates the geographic block g.
  • determining a delivery range for the merchant based on the target merchant set of each of the geographic blocks includes:
  • the connected areas of each of the merchants are processed to obtain the distribution range of each merchant.
  • the connected area of the merchant is processed, and the distribution range of the merchant is obtained:
  • the connected areas of the merchants are combined and/or the hole spurs are processed to obtain the distribution range of the merchant.
  • the method further includes:
  • the compressed area data is stored.
  • FIG. 2 is a flowchart of a method for determining a delivery range according to an embodiment of the present application. Referring to Figure 2, the method specifically includes steps 201-207.
  • step 201 the server performs feature extraction on historical order data of multiple merchants and historical behavior data in multiple geographical blocks to obtain multiple sets of first features, second features, and third features.
  • the historical order data of the merchant may include information such as an order placing address, an order amount, and an order delivery time.
  • the server can perform statistics on the merchant's historical order data to obtain the average customer price of the merchant, the average delivery time of the merchant to a certain geographical block, and the number of orders.
  • the historical behavior data in multiple geographical blocks may include exposures, clicks, exposures of merchants, clicks, etc. in the geographic block.
  • the server can also collect historical behavior data in multiple geographic blocks to obtain exposures and clicks of different dimensions, such as the exposure and click volume of the merchant, the exposure and click volume of the geographic block, and Exposure and clicks of a business in a geographic block.
  • the conversion rate of different dimensions can also be obtained, which is the ratio between the order quantity and the exposure amount or the click amount.
  • the server may extract multiple sets of the first feature, the second feature, and the third feature respectively based on the above data when performing feature extraction.
  • the extracted first feature includes at least one of the exposure amount and the click amount of the merchant dimension, and the merchant dimension.
  • Conversion rate includes at least one of an exposure amount and a click amount of the geographic block dimension, and a conversion rate of the geographic block dimension
  • the third characteristic includes an exposure amount of a cross-dimension of the merchant and the geographic block At least one of the clicks, as well as the conversion rate for the cross-dimension of the business and geographic blocks.
  • the extracted first feature includes at least one of an exposure amount and a click amount of the merchant dimension, and an order quantity of the merchant dimension
  • the second feature includes at least one of an exposure amount and a click amount of the geographic block dimension, and an order quantity of the geographic block dimension
  • the third feature includes at least one of an exposure amount and a click amount of the intersection dimension of the merchant and the geographic block. , and the number of orders for the cross-dimension of the merchant and geographic blocks.
  • the foregoing statistical process and the feature extraction process may be performed by using at least one of an exposure amount and a click amount, which is not specifically limited in the embodiment of the present application.
  • the server trains the prediction model based on the plurality of sets of first features, the second features, and the third features.
  • the data corresponding to the plurality of sets of features is used as training data, and model training can be performed based on any machine learning method to obtain a predictive model. It is assumed that the prediction model is used to predict the conversion rate of the merchant in any geographical block according to the historical order data of the merchant (the flow diagram is shown in FIG. 3).
  • the machine learning method can employ a regression algorithm to construct a predictive model that can be used to characterize conversion rate by exposure and/or click volume and different dimensional conversion rates.
  • the model training process of the foregoing steps 201-202 may be performed at any time, and only needs to be completed before the delivery scope determination is performed, which is not specifically limited in the embodiment of the present application.
  • the foregoing training process and the subsequent delivery range determining process may be performed by one server or by different servers. In the embodiment of the present application, only the same server execution is taken as an example for description.
  • the server invokes the predictive model.
  • the server may invoke the prediction model based on the plurality of sets of the first feature, the second feature, and the third feature, so as to be based on the merchant dimension, the geographic block dimension, and the cross-dimension of any of the merchants. Characteristics to predict the conversion rate of any merchant on any geographic block.
  • the predictive model is used to predict the order quantity, then the number of orders for any merchant on any geographic block can be predicted by the call to the predictive model.
  • step 204 the server inputs historical order data of the plurality of merchants and historical behavior data of the plurality of geographical blocks into the prediction model, and outputs a conversion rate of each merchant in each geographical block.
  • the prediction model is used to predict the order quantity, the historical order data of the plurality of merchants and the historical behavior data of the plurality of geographical blocks may be input into the prediction model to predict any merchant in any area. The number of orders on the block.
  • each merchant is located in each geographic block based on historical order data of multiple merchants and historical behavior data of the plurality of geographical blocks. Conversion rate. The conversion rate based on the actual data can provide real data support in combination optimization, so that the results obtained by the combination optimization are more accurate.
  • step 205 the server optimizes multiple merchants according to the conversion rate of each merchant in each geographic block, the exposure amount of each geographic block, and the average customer unit price of each merchant. A set of target merchants for each geographic block.
  • the merchant set that makes the income of the geographic block higher may be obtained based on each geographical block.
  • the following first objective optimization function (1) can be designed:
  • the second target optimization function (2) may be applied to perform the combined optimization on the plurality of merchants to obtain the target merchant set of each of the geographic blocks;
  • order p,g is the predicted order quantity of the merchant p on the geographic block g, and other parameters are the same as the parameters of the first target optimization function.
  • the constraint condition may be that the average delivery time (or distance) is less than a preset threshold.
  • the time p, g is the average delivery time of the merchant p to the geographic block g
  • T is the preset threshold of the average delivery time, that is, the solution result needs to satisfy the average delivery time is less than the threshold.
  • the constraint can also be constrained by the average distance, that is, the above constraint can be expressed by formula (4):
  • Distance p, g is the average delivery distance of the merchant p to the regional block g
  • Distance is the limit threshold of the preset average delivery distance
  • the target merchant set of each geographical block can be obtained, and the historical order data of the merchant included in the target merchant set satisfies the constraint condition and can ensure the high return in the regional block.
  • the above step 205 is a process of recommending a suitable merchant set for each geographical block by using a combination optimization method (such as the second process in FIG. 4).
  • a combination optimization method such as the second process in FIG. 4.
  • References such as average passenger unit price and average delivery time or average delivery distance are introduced as reference, and actual exposure conditions are introduced, which can improve the accuracy of combination optimization.
  • step 206 the server generates a connected area of each merchant according to at least one geographic block corresponding to each merchant.
  • the target merchant set of each geographical block is obtained.
  • at least one geographical block corresponding to each merchant is obtained, so that it can be determined based on the merchant angle and targeted based on the merchant. Its distribution range.
  • the corresponding at least one geographical block appears as a separate block on the map, and based on the blocks, a connected area of the polygon of the merchant can be generated, as shown in FIG. 5( a ). The polygon shown is outsourced.
  • step 207 the server performs a merge process and/or a hole spur process on the connected area of each merchant according to the three-level road network, and obtains the distribution range of each merchant.
  • the connected areas of each merchant may be combined and processed in combination with the geographic information of the residential area and/or the office area in the three-level road network. For example, when the boundary of the connected area is located in any of the residential areas and/or the office area, the residential area and/or the office area are deleted from the connected area based on the geographic information of the residential area and/or the office area, as shown in the figure.
  • Figure 5 (b) shows.
  • the processed connected area may also have holes and spurs, and the holes may refer to some areas of the connected area that are not covered by the connected area, and the spurs may refer to irregular edge portions, in order to make the distribution range more It is reasonable to fill the hole portion (as shown in Fig. 5(c)), and delete the spur portion as shown in Fig. 5(d), and the final processed area is used as the distribution range of the merchant.
  • the above steps 206 to 207 are processes for generating and optimizing the merchant distribution range (such as the third process in FIG. 5).
  • the geographical area of the merchant needs to be optimized from the perspective of the merchant.
  • the composition of the distribution range, the optimization may include the above-mentioned merger, hole spur processing and the like.
  • the distribution range of each merchant may be saved, the distribution range of the plurality of merchants may be compressed to store the compressed area data.
  • the compressed area data may also be transmitted to reduce the data storage amount of the terminal.
  • the geographic block involved in the foregoing implementation process may be a geographic block based on a geographic hash (geohash granularity), or may be a geographical block based on any granularity of the regional dividing manner, for example, the map may be divided into multiple
  • the hexagonal block or other shape block and the like are not limited in this embodiment of the present application.
  • the method provided by the embodiment of the present application finds that the geographical block can be made in the geographical block by utilizing the historical behavior data in the geographical block and the historical order data of the merchant through an automated method from the perspective of the geographical block.
  • the set of merchants with higher overall returns can not only ensure the overall income of the geographical block, but also improve the distribution efficiency, and can also improve the accuracy of distribution and the efficiency of distribution.
  • the accuracy of the distribution can be improved, and the user experience can be guaranteed.
  • the distribution area is processed, the actual distribution of the road network is also taken into consideration to further rationalize the delivery area, and the accuracy of the distribution can also be improved.
  • FIG. 6 is a schematic structural diagram of a delivery range determining apparatus according to an embodiment of the present application.
  • the apparatus includes:
  • the data obtaining module 601 is configured to acquire historical behavior data in multiple geographical blocks and historical order data of multiple merchants;
  • a target merchant set obtaining module 602 configured to determine, according to historical behavior data in the plurality of geographical blocks and historical order data of the plurality of merchants, a target merchant set of each of the geographic blocks;
  • the delivery range determining module 603 is configured to determine a delivery range for each of the merchants based on the target merchant set of each of the geographic blocks.
  • the target merchant collection acquisition module includes:
  • a prediction submodule configured to predict, according to historical behavior data in the plurality of geographic blocks and historical order data of the plurality of merchants, a conversion rate or an order quantity of each of the merchants on each of the geographic blocks;
  • the predicting sub-module is used to:
  • the historical order data of the plurality of merchants and the historical behavior data of the plurality of geographical blocks are input into the prediction model, and the conversion rate or the order quantity of each merchant in each geographical block is output.
  • the apparatus further includes a training module, the training module for:
  • the first feature includes at least one of an exposure amount and a click amount of the merchant dimension, and a conversion rate or an order quantity of the merchant dimension;
  • the second feature includes at least one of an exposure amount and a click amount of the geographic block dimension And a conversion rate or order quantity of the geographic block dimension;
  • the third feature includes at least one of exposure and click volume of the cross-dimension of the merchant and the geographic block, and conversion of the cross-dimension of the merchant and the geographic block Rate or order quantity.
  • the target merchant collection acquisition module is configured to:
  • the conversion rate of each merchant in each geographical block the exposure amount of each geographical block, and the average customer unit price of each merchant, the plurality of merchants are combined and optimized to obtain each of the geographical regions.
  • the target merchant collection of blocks or,
  • the plurality of merchants are combined and optimized according to the order quantity of each merchant in each geographical block and the average customer unit price of each of the merchants, and the target merchant set of each geographical block is obtained.
  • the target merchant set obtaining module is configured to: apply a first target optimization function to perform combined optimization on the plurality of merchants to obtain a target merchant set of each of the geographic blocks;
  • the first target optimization function is the first target optimization function
  • the target merchant set obtaining module is configured to: apply a second target optimization function to perform combined optimization on the plurality of merchants to obtain a target merchant set of each of the geographic blocks;
  • g is the geographic block index
  • M is the number of the geographical blocks
  • p is the merchant index
  • N is the number of the merchants
  • pv g is the exposure amount in the geographic block g
  • cvr p, g is The conversion rate of the merchant p on the geographic block g, order p, g is the predicted order quantity of the merchant p on the geographic block g
  • Price p is the average customer unit price of the merchant p
  • C p, g is the distribution of the merchant p
  • the geographic block g is used as the 0-1 identifier of the block in the distribution range; when C p, g is 1, the geographical block g is assigned to the merchant p, and C p, g is 0, indicating no Assign a geographic block g to the merchant p.
  • the delivery range determining module comprises:
  • An area generation submodule configured to generate, according to at least one geographic block corresponding to each merchant, a connected area of each of the merchants;
  • the processing submodule is configured to process the connected area of each merchant to obtain a distribution range of each merchant.
  • the processing sub-module is configured to perform a combining process and/or a hole spurting process on the connected areas of each of the merchants according to the three-level road network, to obtain a distribution range of each of the merchants.
  • the apparatus further includes: a compression module, configured to compress a distribution range of the plurality of merchants, and store the compressed area data.
  • the delivery range determining device provided by the above embodiment is only illustrated by the division of each functional module. In actual applications, the function distribution may be completed by different functional modules as needed. The internal structure of the device is divided into different functional modules to complete all or part of the functions described above.
  • the distribution range determining apparatus and the distribution range determining method embodiment provided by the foregoing embodiments are in the same concept, and the specific implementation process is described in detail in the method embodiment, and details are not described herein again.
  • FIG. 7 is a schematic structural diagram of a computer device according to an embodiment of the present disclosure.
  • the computer device 700 may have a large difference due to different configurations or performances, and may include one or more central processing units (CPUs) 701. And one or more memories 702, wherein the memory 702 stores executable instructions that are loaded by the processor 701 and cause the processor 701 to execute to implement the various methods described above.
  • the server may also have components such as a wired or wireless network interface, a keyboard, and an input/output interface for input and output.
  • the server may also include other components for implementing the functions of the device, and details are not described herein.
  • a computer readable storage medium such as a memory comprising instructions executable by a processor in a terminal to perform the computer device method of the embodiments described below.
  • the computer readable storage medium can be a non-transitory computer readable storage medium, a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device.
  • a person skilled in the art may understand that all or part of the steps of implementing the above embodiments may be completed by hardware, or may be instructed by a program to execute related hardware, and the program may be stored in a computer readable storage medium.
  • the storage medium mentioned may be a read only memory, a magnetic disk or an optical disk or the like.

Landscapes

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

Abstract

一种配送范围确定方法，包括：获取多个地域区块内的历史行为数据和多个商户的历史订单数据（101）；根据所述多个地域区块内的历史行为数据和所述多个商户的历史订单数据，获取每个所述地域区块的目标商户集合（102）；和基于每个所述地域区块的目标商户集合，为每个所述商户确定配送范围（103）。

Description

配送范围的确定
相关申请的交叉引用
本专利申请要求于2018年05月17日提交的、申请号为201810475812.X、发明名称为“配送范围确定方法、装置、电子设备以及存储介质”的中国专利申请的优先权，该申请的全文以引用的方式并入本文中。
技术领域
本申请涉及网络技术领域的配送范围确定方法和装置、电子设备以及存储介质。
背景技术
在即时配送场景中，每个商家都有独自的配送范围。商家的配送范围是地理概念上的一块区域。在即时配送应用平台上，商家只对位于该商家的配送范围内的用户可见。也就是说，订单关系只产生在商家和配送范围内的用户。由此可见，商家配送范围可影响商家单量、配送效率和用户体验。若配送范围设定得过小，则潜在用户群体小，商家单量、平台GMV(gross merchandise volume，成交总额)将较小；若配送范围设定得过大，虽然潜在用户群体较大，产生的单量可能有一定程度提升，但整体的配送效率可能降低，进而影响用户的体验。
发明内容
本申请实施例提供了一种配送范围确定方法、装置、电子设备以及存储介质，能够在保证区域内的整体收益的同时，提高配送效率。
一方面，提供了一种配送范围确定方法，所述方法包括：
获取多个地域区块内的历史行为数据和多个商户的历史订单数据；
根据所述多个地域区块内的历史行为数据和多个商户的历史订单数据，确定所述每个地域区块的目标商户集合；和
基于所述每个地域区块的目标商户集合，为所述每个商户确定配送范围。
一方面，提供了一种配送范围确定装置，所述装置包括：
数据获取模块，用于获取多个地域区块内的历史行为数据和多个商户的历史订单数据；
目标商户集合获取模块，用于根据所述多个地域区块内的历史行为数据和多个商户的历史订单数据，确定所述每个地域区块的目标商户集合；和
配送范围确定模块，用于基于所述每个地域区块的目标商户集合，为所述每个商户确定配送范围。
一方面，提供了一种计算机设备，所述计算机设备包括处理器和存储器，所述存储器中存储有可执行指令，所述可执行指令由所述处理器加载并促使所述处理器执行上述配送范围确定方法。
一方面，提供了一种计算机可读存储介质，所述存储介质中存储有可执行指令，所述可执行指令由处理器加载并促使所述处理器执行上述配送范围确定方法。
本申请实施例提供的技术方案，通过充分利用地域区块内的历史行为数据以及商户的历史订单数据，通过自动化的方法，从地域区块的角度，为地域区块找到能够使得该地域区块内时整体收益较高的商户集合，不仅能够保证地域区块的整体收益，还能够提高配送效率。
附图说明
为了更清楚地说明本申请实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本申请实施例提供的一种配送范围确定方法的流程图；
图2是本申请实施例提供的一种配送范围确定方法的流程图；
图3是本申请实施例提供的一种预测流程示意图；
图4是本申请实施例提供的一种配送范围确定过程的示意图；
图5是本申请实施例提供的一种配送范围优化示意图；
图6是本申请实施例提供的一种配送范围确定装置的结构示意图；
图7是本申请实施例提供的一种计算机设备的结构示意图。
具体实施方式
为使本申请的目的、技术方案和优点更加清楚，下面将结合附图对本申请实施方式作进一步地详细描述。
图1是本申请实施例提供的一种配送范围确定方法的流程图。该方法可应用在即时配送应用平台的服务器上。参见图1，该方法包括步骤101-103：
在步骤101、获取多个地域区块内的历史行为数据和多个商户的历史订单数据。
在步骤102、根据所述多个地域区块内的历史行为数据和多个商户的历史订单数据，确定所述每个地域区块的目标商户集合。
在步骤103、基于所述每个地域区块的目标商户集合，为所述每个商户确定配送范围。
在一实施例中，根据所述多个地域区块内的历史行为数据和多个商户的历史订单数据，确定所述每个地域区块的目标商户集合包括：
根据所述多个地域区块内的历史行为数据和多个商户的历史订单数据，预测所述每个商户在所述每个地域区块上的转化率或订单数量；
根据所述每个商户在所述每个地域区块上的转化率或订单数量、所述多个商户的历史订单数据以及所述多个地域区块内的历史行为数据，获取所述每个地域区块的目标商户集合。
在一实施例中，根据所述多个地域区块内的历史行为数据和多个商户的历史订单数据，预测所述每个商户在所述每个地域区块上的转化率或订单数量包括：
调用预测模型；
将所述多个商户的历史订单数据和所述多个地域区块的历史行为数据输入该预测模型，输出每个商户在每个地域区块的转化率或订单数量。
在一实施例中，所述预测模型的训练过程包括：
对所述多个商户的历史订单数据和所述多个地域区块内的历史行为数据进行特征提取，得到多组第一特征、第二特征和第三特征；
基于所述多组第一特征、所述第二特征和所述第三特征，训练得到所述预测模型；
其中，所述第一特征包括商户维度的曝光量和点击量中至少一项，以及商户维度的 转化率或订单数量；所述第二特征包括地域区块维度的曝光量和点击量中至少一项，以及所述地域区块维度的转化率或订单数量；所述第三特征包括商户以及地域区块交叉维度的曝光量和点击量中至少一项，以及商户以及地域区块交叉维度的转化率或订单数量。
在一实施例中，根据所述每个商户在所述每个地域区块上的转化率或订单数量、所述多个商户的历史订单数据以及所述多个地域区块内的历史行为数据，确定所述每个地域区块的目标商户集合包括：
根据所述每个商户在所述每个地域区块上的转化率、每个地域区块的曝光量以及所述每个商户的平均客单价，对多个商户进行组合优化，得到所述每个地域区块的目标商户集合。
在一实施例中，根据所述每个商户在所述每个地域区块上的转化率或订单数量、所述多个商户的历史订单数据以及所述多个地域区块内的历史行为数据，确定所述每个地域区块的目标商户集合包括：
根据所述每个商户在所述每个地域区块上的订单数量以及所述每个商户的平均客单价，对多个商户进行组合优化，得到所述每个地域区块的目标商户集合。
在一实施例中，根据所述每个商户在所述每个地域区块上的转化率、每个地域区块的曝光量以及所述每个商户的平均客单价，对多个商户进行组合优化，得到所述每个地域区块的目标商户集合包括：
应用第一目标优化函数对所述多个商户进行组合优化，得到所述每个地域区块的目标商户集合；
所述第一目标优化函数：
Figure PCTCN2018122085-appb-000001
C p,g∈(0,1)
根据所述每个商户在所述每个地域区块上的订单数量以及所述每个商户的平均客单价，对多个商户进行组合优化，得到所述每个地域区块的目标商户集合包括：
应用第二目标优化函数对所述多个商户进行组合优化，得到所述每个地域区块的目标商户集合；
所述第二目标优化函数：
Figure PCTCN2018122085-appb-000002
C p,g∈(0,1)
其中，g为地域区块索引；M为所述地域区块的数量；p为商家索引；N为所述商家的数量；pv g为地域区块g的曝光量；cvr p,g为商户p在地域区块g上的转化率，order p,g为商户p在地域区块g上的预测订单数量；Price p为商户p的平均客单价；C p,g为是否为商户p分配地域区块g作为其配送范围内的区块的0-1标识；C p,g取值为1时，表示为商户p分配地域区块g，C p,g取值为0时，表示不为商户p分配地域区块g。
在一实施例中，基于所述每个地域区块的目标商户集合，为所述商户确定配送范围包括：
基于每个所述地域区块的目标商户集合，确定所述商户对应的至少一个地域区块；
根据每个商户对应的至少一个地域区块，生成所述每个商户的连通区域；
对所述每个商户的连通区域进行处理，得到每个商户的配送范围。
在一实施例中，对所述商户的连通区域进行处理，得到所述商户的配送范围包括：
根据三级路网对所述商户的连通区域进行合并处理和/或孔洞突刺处理，得到所述商户的配送范围。
在一实施例中，基于所述每个地域区块的目标商户集合，为所述每个商户确定配送范围之后，所述方法还包括：
对所述每个商户的配送范围进行压缩，得到要说后的区域数据；和
存储所述压缩后的区域数据。
上述所有可选技术方案，可以采用任意结合形成本公开的可选实施例，在此不再一一赘述。
图2是本申请实施例提供的一种配送范围确定方法的流程图。参见图2，该方法具体包括步骤201-207。
在步骤201、服务器对多个商户的历史订单数据和多个地域区块内的历史行为数据进行特征提取，得到多组第一特征、第二特征和第三特征。
其中，商户的历史订单数据可以包括订单的下单地址、订单金额、订单配送时长等信息。服务器可以对商户的历史订单数据进行统计，以得到商户的平均客单价、商家配送到某个地域区块的平均配送时长以及订单数量。而多个地域区块内的历史行为数据可 以包括地域区块内的曝光量、点击量、商户的曝光量、点击量等。服务器还可以对多个地域区块内的历史行为数据进行统计，得到不同维度的曝光量、点击量，例如，商户的曝光量和点击量，地域区块的曝光量和点击量，以及，某个商户在某个地域区块内的曝光量和点击量等。基于上述统计得到的数据，还可以得到不同维度的转化率，该转化率是指从订单数量与曝光量或点击量之间的比例。为了从不同维度来确定转化规律，服务器在进行特征提取时，可以基于上述数据，分别提取多组第一特征、第二特征和第三特征。
需要说明的是，如果预测模型用于预测商户在地域区块的转化率，则在进行特征提取时，所提取的第一特征包括商户维度的曝光量和点击量中至少一项，以及商户维度的转化率；该第二特征包括地域区块维度的曝光量和点击量中至少一项，以及该地域区块维度的转化率；该第三特征包括商户以及地域区块交叉维度的曝光量和点击量中至少一项，以及商户以及地域区块交叉维度的转化率。
需要说明的是，如果预测模型用于预测订单数量，则在进行特征提取时，所提取的第一特征包括商户维度的曝光量和点击量中至少一项，以及商户维度的订单数量；该第二特征包括地域区块维度的曝光量和点击量中至少一项，以及该地域区块维度的订单数量；该第三特征包括商户以及地域区块交叉维度的曝光量和点击量中至少一项，以及商户以及地域区块交叉维度的订单数量。
需要说明的是，上述统计过程和特征提取过程，可以通过曝光量和点击量中至少一项进行，本申请实施例对此不做具体限定。
在步骤202、服务器基于所述多组第一特征、所述第二特征和所述第三特征，训练得到所述预测模型。
将上述多组特征对应的数据作为训练数据，可以基于任一种机器学习方法，来进行模型训练，以得到预测模型。假设所述预测模型用于根据商户的历史订单数据对所述商户在任一个地域区块的转化率进行预测(该流程示意可参见图3)。例如，该机器学习方法可以采用回归算法，以构建能够用于表征转化率受到曝光量和/或点击量以及不同维度转化率影响的预测模型。
需要说明的是，对于服务器来说，上述步骤201-202的模型训练过程可以是在任一时刻进行，只需在进行配送范围确定之前训练完成即可，本申请实施例对此不做具体限定。且，上述训练过程和后续的配送范围确定过程可以由一个服务器执行，也可以由不 同服务器执行，在本申请实施例中，仅以采用同一个服务器执行为例进行说明。
在步骤203、服务器调用预测模型。
当确定配送范围时，服务器可以调用该基于多组第一特征、第二特征和第三特征训练得到的预测模型，从而能够基于任一商户的商户维度、地域区块维度以及二者交叉维度上的特征，来预测任一商户在任一地域区块上的转化率。
当然，如果该预测模型用于预测订单数量，则可以通过对该预测模型的调用来预测任一商户在任一地域区块上的订单数量。
在步骤204、服务器将所述多个商户的历史订单数据和所述多个地域区块的历史行为数据输入该预测模型，输出每个商户在每个地域区块的转化率。
由于之前所训练的预测模型，可以提供一种转化率受到各方面影响的规律，因此，基于该规律，可以对任一个商户和任一个地域区块来说，预测出该商户在该地域区块的转化率。当然，如果该预测模型用于预测订单数量，则可以通过将所述多个商户的历史订单数据和所述多个地域区块的历史行为数据输入该预测模型，来预测任一商户在任一地域区块上的订单数量。
上述步骤201至204实际上提供了组合优化过程所需的数据。参见图4中的第一过程，在该第一过程中，主要是基于多个商户的历史订单数据和所述多个地域区块的历史行为数据，来得到每个商户在每个地域区块的转化率。基于实际数据所得到的转化率，能够在组合优化时提供真实的数据支持，使得组合优化所得到的结果更加准确。
在步骤205、服务器根据所述每个商户在每个地域区块的转化率、每个地域区块的曝光量以及所述每个商户的平均客单价，对多个商户进行组合优化，得到所述每个地域区块的目标商户集合。
在确定了商户在该地域区块的转化率后，可以基于每个地域区块，来获取使得该地域区块的收益较高的商户集合。以提高收益为目的，可以设计如下的第一目标优化函数(1)：
Figure PCTCN2018122085-appb-000003
其中，C p,g∈(0,1)；g为地域区块索引；M为所述地域区块的数量；p为商家索引；N为所述商家的数量；pv g为地域区块g的曝光量；cvr p,g为商户p在地域区块g上的转化率；Price p为商户p的平均客单价；C p,g为是否为商户p分配地域区块g作为其配送 范围内的区块的0-1标识；其中，C p,g取值为1时，表示为商户p分配地域区块g，C p,g取值为0时，表示不为商户p分配地域区块g。
如果预测模型是用于预测订单数量的模型，则还可以应用下述第二目标优化函数(2)对所述多个商户进行组合优化，得到所述每个地域区块的目标商户集合；
Figure PCTCN2018122085-appb-000004
其中，C p,g∈(0,1)，order p,g为商户p在地域区块g上的预测订单数量，其他参数与上述第一目标优化函数的参数同理。
在以地域区块内收益最大化为优化目标求解的同时，还需要基于一定约束条件，以保证用户体验，该约束条件可以为平均配送时长(或距离)小于预设阈值。
该约束条件可以用公式(3)表示：
Figure PCTCN2018122085-appb-000005
其中，Time p,g为商户p配送至地域区块g的平均配送时长，T为预设的平均配送时长的限制阈值，即求解结果需要满足平均配送时长小于该阈值。需要说明的是，该约束条件还可以采用平均距离进行约束，也即是，上述约束条件可以用公式(4)表示：
Figure PCTCN2018122085-appb-000006
其中，Distance p,g为商户p配送至地域区块g的平均配送距离，Distance为预设的平均配送距离的限制阈值。
基于上述目标优化函数和约束条件进行求解，则可以得到每个地域区块的目标商户集合，目标商户集合所包含商户的历史订单数据满足约束条件且能够保证该地域区块内的收益较高。
需要说明的是，上述仅是组合优化的一种示例，在实际场景中，还可以采用其他组合优化算法以及其他约束条件来进行生成地域区块集合，本申请实施例对此不做具体限定。
上述步骤205是利用组合优化的方法为每个地域区块推荐合适的商户集合的过程(如图4中的第二过程)，在该第二过程中，为了使得用户体验和收益能够得到保证， 引入了诸如平均客单价以及平均配送时长或者平均配送距离等作为参考，还引入了实际的曝光情况，能够提高组合优化的准确性。
在步骤206、服务器根据每个商户对应的至少一个地域区块，生成所述每个商户的连通区域。
上述求解时得到了每个地域区块的目标商户集合，实际上，也即是得到了每个商户对应的至少一个地域区块，因此，可以基于商户角度，有针对性的再基于商户来确定其配送范围。对于每个商户来说，其对应的至少一个地域区块在地图上表现为一个一个独立的区块，基于这些区块，可以生成该商户的多边形的连通区域，如图5中(a)图所示的多边形外包。
在步骤207、服务器根据三级路网对所述每个商户的连通区域进行合并处理和/或孔洞突刺处理，得到所述每个商户的配送范围。
基于上述连通区域，可以结合三级路网中的居住区和/或办公区的地理信息，对每个商户的连通区域进行合并处理。例如，当连通区域的边界位于任一个居住区和/或办公区内时，基于该居住区和/或办公区的地理信息，将该居住区和/或办公区从连通区域中删除，如图5中(b)图所示。
当然，处理后的连通区域还可能有孔洞和突刺部分，孔洞可以是指连通区域中有一些未被该连通区域覆盖的地域区块，突刺可以是指不规则的边缘部分，为了使得配送范围更加合理，可以对孔洞部分进行填充(如图5中(c)图所示)，对突刺部分进行删除如图5中(d)图所示，将最终处理得到的区域作为商户的配送范围。
需要说明的是，在上述服务器对所述每个商户的连通区域进行处理，得到每个商户的配送范围的过程中，可以基于连通区域的不同情况进行不同的处理，而无需对于每个商户的连通区域均进行上述合并处理、孔洞处理以及突刺处理，以避免对服务器计算资源的浪费。
上述步骤206至步骤207是进行商户配送范围的生成和优化的过程(如图5中的第三过程)，该第三过程中，需要从商户的角度，来整体优化该商户的各个地域区块所组成的配送范围，该优化可以包括上述涉及的合并、孔洞突刺处理等等过程。进一步地，在保存各个商户的配送范围时，还可以对所述多个商户的配送范围进行压缩，存储压缩后的区域数据。而在将各个商户的配送范围发送给商户终端时，也可以发送压缩后的区域数据，以减少终端的数据存储量。
在上述实现过程中所涉及的地域区块可以是基于地理哈希(geohash)粒度的地域区块，还可以是基于任一种区域划分方式粒度的地域区块，如，可以将地图分割成多个六边形区块或其他形状区块等，本申请实施例对此不做限定。
本申请实施例提供的方法，通过充分利用地域区块内的历史行为数据以及商户的历史订单数据，通过自动化的方法，从地域区块的角度，为地域区块找到能够使得该地域区块内时整体收益较高的商户集合，不仅能够保证地域区块的整体收益，还能够提高配送效率，还能够提高分配的准确性和分配的效率。进一步地，在为地域区块获取商户集合时，不仅考虑到了订单的转化情况，还考虑到了配送情况，能够提高分配的准确性，且能够保证用户体验。进一步地，在对配送区域进行处理时，还考虑到了实际的路网分布情况，来进一步对配送区域进行合理化，也能够提高分配的准确性。
图6是本申请实施例提供的一种配送范围确定装置的结构示意图。参见图6，所述装置包括：
数据获取模块601，用于获取多个地域区块内的历史行为数据和多个商户的历史订单数据；
目标商户集合获取模块602，用于根据所述多个地域区块内的历史行为数据和多个商户的历史订单数据，确定所述每个地域区块的目标商户集合；
配送范围确定模块603，用于基于所述每个地域区块的目标商户集合，为所述每个商户确定配送范围。
在一实施例中，所述目标商户集合获取模块包括：
预测子模块，用于根据所述多个地域区块内的历史行为数据和多个商户的历史订单数据，预测所述每个商户在所述每个地域区块上的转化率或订单数量；
获取子模块，用于根据所述每个商户在所述每个地域区块上的转化率或订单数量、所述多个商户的历史订单数据以及所述多个地域区块内的历史行为数据，获取所述每个地域区块的目标商户集合。
在一实施例中，所述预测子模块用于：
调用预测模型；
将所述多个商户的历史订单数据和所述多个地域区块的历史行为数据输入该预测模型，输出每个商户在每个地域区块的转化率或订单数量。
在一实施例中，所述装置还包括训练模块，所述训练模块用于：
对所述多个商户的历史订单数据和所述多个地域区块内的历史行为数据进行特征提取，得到多组第一特征、第二特征和第三特征；
基于每组第一特征、所述第二特征和所述第三特征，训练得到所述预测模型；
其中，所述第一特征包括商户维度的曝光量和点击量中至少一项，以及商户维度的转化率或订单数量；所述第二特征包括地域区块维度的曝光量和点击量中至少一项，以及所述地域区块维度的转化率或订单数量；所述第三特征包括商户以及地域区块交叉维度的曝光量和点击量中至少一项，以及商户以及地域区块交叉维度的转化率或订单数量。
在一实施例中，所述目标商户集合获取模块用于：
根据所述每个商户在每个地域区块的转化率、每个地域区块的曝光量以及所述每个商户的平均客单价，对多个商户进行组合优化，得到所述每个地域区块的目标商户集合；或，
根据所述每个商户在每个地域区块的订单数量以及所述每个商户的平均客单价，对多个商户进行组合优化，得到所述每个地域区块的目标商户集合。
在一实施例中，所述目标商户集合获取模块用于：应用第一目标优化函数对所述多个商户进行组合优化，得到所述每个地域区块的目标商户集合；
所述第一目标优化函数：
Figure PCTCN2018122085-appb-000007
或，
所述目标商户集合获取模块用于：应用第二目标优化函数对所述多个商户进行组合优化，得到所述每个地域区块的目标商户集合；
所述第二目标优化函数：
Figure PCTCN2018122085-appb-000008
其中，g为地域区块索引；M为所述地域区块的数量；p为商家索引；N为所述商家的数量；pv g为在地域区块g上的曝光量；cvr p,g为商户p在地域区块g上的转化率，order p,g为商户p在地域区块g上的预测订单数量；Price p为商户p的平均客单价；C p,g为是否为商户p分配地域区块g作为其配送范围内的区块的0-1标识；C p,g取值为1时，表示为商户p分配地域区块g，C p,g取值为0时，表示不为商户p分配地域区块g。
在一实施例中，所述配送范围确定模块包括：
区域生成子模块，用于根据每个商户对应的至少一个地域区块，生成所述每个商户的连通区域；
处理子模块，用于对所述每个商户的连通区域进行处理，得到每个商户的配送范围。
在一实施例中，所述处理子模块用于根据三级路网对所述每个商户的连通区域进行合并处理和/或孔洞突刺处理，得到所述每个商户的配送范围。
在一实施例中，所述装置还包括：压缩模块，用于对所述多个商户的配送范围进行压缩，存储压缩后的区域数据。
需要说明的是：上述实施例提供的配送范围确定装置在配送范围确定时，仅以上述各功能模块的划分进行举例说明，实际应用中，可以根据需要而将上述功能分配由不同的功能模块完成，即将设备的内部结构划分成不同的功能模块，以完成以上描述的全部或者部分功能。另外，上述实施例提供的配送范围确定装置与配送范围确定方法实施例属于同一构思，其具体实现过程详见方法实施例，这里不再赘述。
图7是本申请实施例提供的一种计算机设备的结构示意图，该计算机设备700可因配置或性能不同而产生比较大的差异，可以包括一个或一个以上处理器(central processing units，CPU)701和一个或一个以上的存储器702，其中，所述存储器702中存储有可执行指令，所述可执行指令由所述处理器701加载并促使处理器701执行以实现上述各个方法。当然，该服务器还可以具有有线或无线网络接口、键盘以及输入输出接口等部件，以便进行输入输出，该服务器还可以包括其他用于实现设备功能的部件，在此不做赘述。
在示例性实施例中，还提供了一种计算机可读存储介质，例如包括指令的存储器，上述指令可由终端中的处理器执行以完成下述实施例中的计算机设备方法。例如，所述计算机可读存储介质可以是非易失性计算机可读存储介质、ROM、随机存取存储器(RAM)、CD-ROM、磁带、软盘和光数据存储设备等。
本领域普通技术人员可以理解实现上述实施例的全部或部分步骤可以通过硬件来完成，也可以通过程序来指令相关的硬件完成，所述的程序可以存储于一种计算机可读存储介质中，上述提到的存储介质可以是只读存储器，磁盘或光盘等。
以上所述仅为本申请的较佳实施例，并不用以限制本申请，凡在本申请的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本申请的保护范围之内。

Claims (14)

  1. 一种配送范围确定方法，包括：
    获取多个地域区块内的历史行为数据和多个商户的历史订单数据；
    根据所述多个地域区块内的历史行为数据和所述多个商户的历史订单数据，确定每个所述地域区块的目标商户集合；
    基于每个所述地域区块的目标商户集合，为每个所述商户确定配送范围。
  2. 根据权利要求1所述的方法，其特征在于，根据所述多个地域区块内的历史行为数据和所述多个商户的历史订单数据，确定每个所述地域区块的目标商户集合包括：
    根据所述多个地域区块内的历史行为数据和所述多个商户的历史订单数据，预测每个所述商户在每个所述地域区块上的转化率或订单数量；
    根据每个所述商户在每个所述地域区块上的所述转化率或订单数量、所述多个商户的历史订单数据以及所述多个地域区块内的历史行为数据，确定每个所述地域区块的目标商户集合。
  3. 根据权利要求2所述的方法，其特征在于，根据所述多个地域区块内的历史行为数据和所述多个商户的历史订单数据，预测每个所述商户在每个所述地域区块上的转化率或订单数量包括：
    调用预测模型；
    将所述多个商户的历史订单数据和所述多个地域区块的历史行为数据输入该预测模型，输出每个所述商户在每个所述地域区块的转化率或订单数量。
  4. 根据权利要求3所述的方法，其特征在于，所述预测模型的训练过程包括：
    对所述多个商户的历史订单数据和所述多个地域区块内的历史行为数据进行特征提取，得到多组第一特征、第二特征和第三特征；
    基于每组所述第一特征、所述第二特征和所述第三特征，训练得到所述预测模型；
    其中，所述第一特征包括商户维度的曝光量和点击量中至少一项，以及商户维度的转化率或订单数量；
    所述第二特征包括地域区块维度的曝光量和点击量中至少一项，以及所述地域区块维度的转化率或订单数量；
    所述第三特征包括商户与地域区块交叉维度的曝光量和点击量中至少一项，以及商户与地域区块交叉维度的转化率或订单数量。
  5. 根据权利要求2所述的方法，其特征在于，根据每个所述商户在每个所述地域区块上的所述转化率或订单数量、所述多个商户的历史订单数据以及所述多个地域区块 内的历史行为数据，确定每个所述地域区块的目标商户集合包括：
    根据每个所述商户在每个所述地域区块上的所述转化率、每个所述地域区块的曝光量以及每个所述商户的平均客单价，对所述多个商户进行组合优化，得到每个所述地域区块的目标商户集合。
  6. 根据权利要求5所述的方法，其特征在于，根据每个所述商户在每个所述地域区块上的所述转化率、每个所述地域区块的曝光量以及每个所述商户的平均客单价，对所述多个商户进行组合优化，得到每个所述地域区块的目标商户集合包括：
    应用第一目标优化函数对所述多个商户进行组合优化，得到每个所述地域区块的目标商户集合；
    所述第一目标优化函数用公式(1)表达：
    Figure PCTCN2018122085-appb-100001
    其中，所述第一目标优化函数的约束条件用公式(2)表达：
    Figure PCTCN2018122085-appb-100002
    其中，g为地域区块索引；
    M为所述地域区块的数量；
    p为商家索引；
    N为所述商家的数量；
    pv g为地域区块g的曝光量；
    cvr p,g为商户p在地域区块g上的转化率；
    Price p为商户p的平均客单价；
    C p,g为是否为商户p分配地域区块g作为其配送范围内的区块的0-1标识；
    C p,g取值为1时，表示为商户p分配地域区块g；
    C p,g取值为0时，表示不为商户p分配地域区块g；
    Time p,g为商户p配送至地域区块g的平均配送时长；
    T为预设的平均配送时长阈值。
  7. 根据权利要求2所述的方法，其特征在于，根据每个所述商户在每个所述地域区块上的所述转化率或订单数量、所述多个商户的历史订单数据以及所述多个地域区块内的历史行为数据，确定每个所述地域区块的目标商户集合包括：
    根据每个所述商户在每个所述地域区块上的所述订单数量以及每个所述商户的平均客单价，对所述多个商户进行组合优化，得到每个所述地域区块的目标商户集合。
  8. 根据权利要求7所述的方法，其特征在于，根据所述每个商户在所述每个地域区块上的订单数量以及所述多个商户的平均客单价，对所述多个商户进行组合优化，得到所述每个地域区块的目标商户集合包括：
    应用第二目标优化函数对所述多个商户进行组合优化，得到每个所述地域区块的目标商户集合；
    所述第二目标优化函数用公式(3)表达：
    Figure PCTCN2018122085-appb-100003
    其中，所述第二目标优化函数的约束条件用公式(4)表达：
    Figure PCTCN2018122085-appb-100004
    其中，g为地域区块索引；
    M为所述地域区块的数量；
    p为商家索引；
    N为所述商家的数量；
    order p,g为商户p在地域区块g上的订单数量；
    Price p为商户p的平均客单价；
    C p,g为是否为商户p分配地域区块g作为其配送范围内的区块的0-1标识；
    C p,g取值为1时，表示为商户p分配地域区块g；
    C p,g取值为0时则表示不为商户p分配地域区块g；
    Distance p,g为商户p配送至地域区块g的平均配送距离，
    Distance为预设的平均配送距离阈值。
  9. 根据权利要求1所述的方法，其特征在于，基于每个所述地域区块的目标商户集合，为所述商户确定配送范围包括：
    基于每个所述地域区块的目标商户集合，确定所述商户对应的至少一个地域区块；
    根据所述商户对应的所述至少一个地域区块，生成所述商户的连通区域；
    对所述商户的连通区域进行处理，得到所述商户的配送范围。
  10. 根据权利要求9所述的方法，其特征在于，对所述商户的连通区域进行处理， 得到所述商户的配送范围包括：
    根据三级路网对所述商户的连通区域进行合并处理和/或孔洞突刺处理，得到所述商户的配送范围。
  11. 根据权利要求1所述的方法，其特征在于，所述方法还包括：
    对每个所述商户的配送范围进行压缩，得到压缩后的区域数据；和
    存储所述压缩后的区域数据。
  12. 一种配送范围确定装置，包括：
    数据获取模块，用于获取多个地域区块内的历史行为数据和多个商户的历史订单数据；
    目标商户集合获取模块，用于根据所述多个地域区块内的历史行为数据和所述多个商户的历史订单数据，确定每个所述地域区块的目标商户集合；
    配送范围确定模块，用于基于每个所述地域区块的目标商户集合，为每个所述商户确定配送范围。
  13. 一种计算机设备，所述计算机设备包括处理器和存储器，所述存储器中存储有可执行指令，所述可执行指令由所述处理器加载并促使所述处理器执行权利要求1-11任一项所述的配送范围确定方法。
  14. 一种计算机可读存储介质，所述存储介质中存储有可执行条指令，所述指令由处理器加载并促使所述处理器执行权利要求1-11任一项所述的配送范围确定方法。
PCT/CN2018/122085 2018-05-17 2018-12-19 配送范围的确定 WO2019218668A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
BR112020023257-2A BR112020023257A2 (pt) 2018-05-17 2018-12-19 método e aparelho para determinar uma faixa de entrega, dispositivo de computador, e, mídia de armazenamento legível por computador.
US17/055,930 US20210312486A1 (en) 2018-05-17 2018-12-19 Determining delivery range

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810475812.X 2018-05-17
CN201810475812.XA CN110570136B (zh) 2018-05-17 2018-05-17 配送范围确定方法、装置、电子设备以及存储介质

Publications (1)

Publication Number Publication Date
WO2019218668A1 true WO2019218668A1 (zh) 2019-11-21

Family

ID=68539418

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/122085 WO2019218668A1 (zh) 2018-05-17 2018-12-19 配送范围的确定

Country Status (4)

Country Link
US (1) US20210312486A1 (zh)
CN (1) CN110570136B (zh)
BR (1) BR112020023257A2 (zh)
WO (1) WO2019218668A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112836965B (zh) * 2021-02-02 2023-04-25 拉扎斯网络科技（上海）有限公司 压力调控方案的选取方法和装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104851006A (zh) * 2015-05-21 2015-08-19 北京京东尚科信息技术有限公司 配送范围确定方法以及装置
CN105574014A (zh) * 2014-10-13 2016-05-11 北京明略软件系统有限公司 一种商圈划分方法及系统
CN105825360A (zh) * 2016-03-31 2016-08-03 北京小度信息科技有限公司 商户配送范围的调整方法和装置
CN106920056A (zh) * 2017-02-27 2017-07-04 广州云移信息科技有限公司 多个门店配送范围智能设定方法及终端

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2012034105A2 (en) * 2010-09-10 2012-03-15 Turnkey Intelligence, Llc Systems and methods for generating prospect scores for sales leads, spending capacity scores for sales leads, and retention scores for renewal of existing customers
US8880420B2 (en) * 2011-12-27 2014-11-04 Grubhub, Inc. Utility for creating heatmaps for the study of competitive advantage in the restaurant marketplace
US9923979B2 (en) * 2013-06-27 2018-03-20 Google Llc Systems and methods of determining a geographic location based conversion
US20150186869A1 (en) * 2013-12-05 2015-07-02 Cfph, Llc Examples of delivery and/or referral service sms ordering
US20160148226A1 (en) * 2014-11-21 2016-05-26 Oracle International Corporation System and method for forecasting and managing returned merchanidse in retail
CN105160503A (zh) * 2015-07-16 2015-12-16 绍兴越淘网络技术有限公司 一种网站运营商直接实现同城当天配送的方法
US20170300948A1 (en) * 2016-04-18 2017-10-19 Mastercard International Incorporated Systems and Methods for Predicting Purchase Behavior Based on Consumer Transaction Data in a Geographic Location
CN107437122B (zh) * 2016-05-26 2019-03-26 北京三快在线科技有限公司 确定商家配送范围的方法及装置
CN107633680B (zh) * 2016-07-12 2021-05-04 阿里巴巴集团控股有限公司 出行数据的获取方法、装置、设备和系统
CN107392513B (zh) * 2017-01-26 2018-12-28 北京小度信息科技有限公司 订单处理方法和装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105574014A (zh) * 2014-10-13 2016-05-11 北京明略软件系统有限公司 一种商圈划分方法及系统
CN104851006A (zh) * 2015-05-21 2015-08-19 北京京东尚科信息技术有限公司 配送范围确定方法以及装置
CN105825360A (zh) * 2016-03-31 2016-08-03 北京小度信息科技有限公司 商户配送范围的调整方法和装置
CN106920056A (zh) * 2017-02-27 2017-07-04 广州云移信息科技有限公司 多个门店配送范围智能设定方法及终端

Also Published As

Publication number Publication date
US20210312486A1 (en) 2021-10-07
BR112020023257A2 (pt) 2021-02-23
CN110570136A (zh) 2019-12-13
CN110570136B (zh) 2021-11-02

Similar Documents

Publication Publication Date Title
CN110033314B (zh) 广告数据处理方法及装置
CN109902708A (zh) 一种推荐模型训练方法及相关装置
CN107633257B (zh) 数据质量评估方法及装置、计算机可读存储介质、终端
CN111242165A (zh) 商户聚类方法、装置、计算机设备和存储介质
CN112507212A (zh) 智能回访方法、装置、电子设备及可读存储介质
CN112700017A (zh) 物业维修资源分配方法、装置、计算机设备及存储介质
CN113568759B (zh) 一种基于云计算的大数据处理方法及其系统
WO2019218668A1 (zh) 配送范围的确定
CN112950079B (zh) 绿地供需数据处理方法、系统、计算机设备及存储介质
CN107844496B (zh) 统计信息输出方法及装置
CN112465371A (zh) 一种资源数据分配方法、装置及设备
CN107395663A (zh) 数据获取方法及装置
CN114116528B (zh) 一种内存访问地址预测方法、装置、存储介质及电子设备
CN113890948B (zh) 基于语音外呼机器人对话数据的资源分配方法及相关设备
CN114581130A (zh) 基于客户画像的银行网点派号方法及装置、存储介质
CN114490719A (zh) 一种数据查询方法、装置、电子设备以及存储介质
CN113822455A (zh) 一种时间预测方法、装置、服务器及存储介质
WO2020128449A1 (en) Data aggregation
CN110781314A (zh) 用户关系图的分层显示方法、装置及电子设备
CN110083517A (zh) 一种用户画像置信度的优化方法及装置
TWI657393B (zh) 行銷客群預測系統及其方法
CN114936851B (zh) 一种基于区块链的idc数据协同方法及系统
US20110047113A1 (en) Method and apparatus for the prediction of order turnaround time in an information verification system
CN112465629A (zh) 风险预测方法、装置、电子设备及存储介质
CN112035067B (zh) 基于事件驱动的云存储容量预测方法、系统及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18919104

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

REG Reference to national code

Ref country code: BR

Ref legal event code: B01A

Ref document number: 112020023257

Country of ref document: BR

ENP Entry into the national phase

Ref document number: 112020023257

Country of ref document: BR

Kind code of ref document: A2

Effective date: 20201113

122 Ep: pct application non-entry in european phase

Ref document number: 18919104

Country of ref document: EP

Kind code of ref document: A1