CN114971761A - Commodity recommendation method based on machine learning - Google Patents

Commodity recommendation method based on machine learning Download PDF

Info

Publication number
CN114971761A
CN114971761A CN202210259968.0A CN202210259968A CN114971761A CN 114971761 A CN114971761 A CN 114971761A CN 202210259968 A CN202210259968 A CN 202210259968A CN 114971761 A CN114971761 A CN 114971761A
Authority
CN
China
Prior art keywords
user
determining
recommendation
commodity
recommended
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210259968.0A
Other languages
Chinese (zh)
Inventor
贾信明
林昱洲
杨宏
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.)
Hua Analysis Technology Shanghai Co ltd
Original Assignee
Hua Analysis Technology Shanghai 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 Hua Analysis Technology Shanghai Co ltd filed Critical Hua Analysis Technology Shanghai Co ltd
Priority to CN202210259968.0A priority Critical patent/CN114971761A/en
Publication of CN114971761A publication Critical patent/CN114971761A/en
Pending legal-status Critical Current

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]
    • G06Q30/0631Item recommendations
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00Machine learning

Landscapes

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

Abstract

An embodiment of the present specification provides a commodity recommendation method based on machine learning, including: judging whether a user is in an isolated state or not based on the consumption condition and the position information of the user in the first time; and determining recommended commodities pushed to the user through a recommendation model in response to the user being in the isolation state.

Description

Commodity recommendation method based on machine learning
Technical Field
The specification relates to the technical field of computers, in particular to a commodity recommendation method based on machine learning.
Background
With the continuous development of electronic technology and network technology, the shopping platform continuously collects user data and can recommend commodities to users according to commodity characteristic information and user characteristic information. However, in a special period (such as a flu outbreak period), the user often lives at home for a longer time, so that the shopping preference of the user changes, and the recommended goods do not accord with the preference of the user. Therefore, how to identify the home state of the user to perform more reasonable commodity recommendation is a problem to be solved urgently.
Disclosure of Invention
One embodiment of the present specification provides a commodity recommendation method based on machine learning. The commodity recommendation method comprises the following steps: judging whether a user is in an isolated state or not based on the consumption condition and the position information of the user in the first time; and determining recommended commodities pushed to the user through a recommendation model in response to the user being in the isolation state.
One of the embodiments of the present specification provides a commodity recommendation system based on machine learning, the system including: the judgment module is used for judging whether the user is in an isolation state or not based on the consumption condition and the position information of the user in the first time. And the recommending module is used for responding to the condition that the user is in the isolation state and determining the commodities recommended to the user through a recommending model.
One of the embodiments of the present specification provides a commodity recommendation device based on machine learning, which includes a processor, where the processor is configured to execute the commodity recommendation method based on machine learning described in the embodiments of the present specification. A method.
One of the embodiments of the present specification provides a computer-readable storage medium, where the storage medium stores computer instructions, and when the computer reads the computer instructions in the storage medium, the computer executes the method for recommending commodities based on machine learning according to the embodiments of the present specification.
Drawings
The present description will be further explained by way of exemplary embodiments, which will be described in detail by way of the accompanying drawings. These embodiments are not intended to be limiting, and in these embodiments, like numbers indicate like structures, where:
FIG. 1 is a schematic diagram of an application scenario of a merchandise recommendation system according to some embodiments of the present description;
FIG. 2 is a block diagram of a machine learning based merchandise recommendation system according to some embodiments of the present description;
FIG. 3 is an exemplary flow diagram of a method for machine learning based merchandise recommendation in accordance with some embodiments of the present description;
FIG. 4 is an exemplary flow diagram of an isolation status determination method according to some embodiments described herein;
FIG. 5 provides a schematic diagram of a method of determining recommended merchandise, according to some embodiments of the present description;
FIG. 6 provides a flow chart of another method of determining recommended merchandise according to some embodiments of the present description.
Detailed Description
In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings used in the description of the embodiments will be briefly described below. It is obvious that the drawings in the following description are only examples or embodiments of the present description, and that for a person skilled in the art, the present description can also be applied to other similar scenarios on the basis of these drawings without inventive effort. Unless otherwise apparent from the context, or otherwise indicated, like reference numbers in the figures refer to the same structure or operation.
It should be understood that "system", "apparatus", "unit" and/or "module" as used herein is a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, other words may be substituted by other expressions if they accomplish the same purpose.
As used in this specification and the appended claims, the terms "a," "an," "the," and/or "the" are not intended to be inclusive in the singular, but rather are intended to be inclusive in the plural, unless the context clearly dictates otherwise. In general, the terms "comprising" and "comprises" merely indicate that steps and elements are included which are explicitly identified, that the steps and elements do not form an exclusive list, and that a method or apparatus may comprise other steps or elements.
Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of the present specification. It should be understood that the preceding or following operations are not necessarily performed in the exact order in which they are performed. Rather, the various steps may be processed in reverse order or simultaneously. Meanwhile, other operations may be added to the processes, or a certain step or several steps of operations may be removed from the processes.
Fig. 1 is a schematic diagram of an application scenario of a product recommendation system according to some embodiments of the present specification. As shown in fig. 1, an application scenario of the machine learning-based goods recommendation system may include a server 110, a processor 120, a storage device 130, a user terminal 140, a network 150, and the like.
The commodity recommendation system can be used for a sales service platform. In some embodiments, the system comprises a sales service platform for retail goods. Such as e-commerce platforms, vending machines, etc. The commodity recommendation system can recommend commodities to users by the commodity recommendation method disclosed in the specification.
In some embodiments, the sales service platform may provide the product exhibition service for the offline merchant (also referred to as store), and the user may access the sales service platform through the user terminal 140 and purchase the offline merchant according to the push information.
The server 110 may communicate with the processor 120, the storage device 130, and the user terminal 140 through the network 150 to provide various functions of merchandise recommendation. In some embodiments, the user terminal 140 may send the consumption situation and location information of the current user within the first time to the server 110, and receive the commodity recommendation information sent by the server 110. The server 110 may obtain the store location information and the store epidemic prevention evaluation information, and perform processing to determine recommended goods to be sent to the nearby user terminal 140. The above information transfer relationship between the devices is only an example, and in some cases, the information transfer between the devices may also include other forms, which may be determined according to actual situations.
The server 110 may be used to manage historical consumption data of the user and to process data and/or information from at least one component of the present system or an external data source (e.g., a cloud data center). In some embodiments, the server 110 may be a single server or a group of servers. The set of servers can be centralized or distributed (e.g., the servers 110 can be a distributed system), can be dedicated, or can be served simultaneously by other devices or systems. In some embodiments, the server 110 may be regional or remote. In some embodiments, the server 110 may be implemented on a cloud platform, or provided in a virtual manner.
Processor 120 refers to a device or system having computing capabilities. In some embodiments, processor 120 may include one or more processing engines (e.g., a single core processing engine or a multi-core processor). By way of example only, the processing engine may include one or more combinations of central processing units, application specific instruction set processors (ASIPs), Digital Signal Processors (DSPs), Reduced Instruction Set Computers (RISCs), microprocessors, and the like. In some embodiments, the processor 120 may invoke data and/or instructions stored in the storage device 130 to implement the merchandise recommendation methods provided herein. In some embodiments, the processor 120 may determine whether the user is in the isolated state based on the consumption and location information of the user within the first time. In some embodiments, the processor 120 may recommend the commodity to the user based on the quarantine status of the user.
Storage device 130 may be used to store data and/or instructions. Storage device 130 may include one or more storage components, each of which may be a separate device or part of another device. In some embodiments, the storage device 130 may further store related information generated in the process of selling goods, for example, the related information may include store information and consumption information generated by the user, and for example, the related information may include user related information such as consumption situation, location information, and consumption habits of the user; for another example, the related information may also be store-related information such as store location information of a store, store epidemic prevention evaluation information, store inventory information, and product characteristics. In some embodiments, the storage device 130 may store related data generated by the sales service platform, for example, related data generated when the platform pushes the pushed goods to the user, the quarantine status of the user, and the recommendation model perform the goods recommendation method. In some embodiments, storage 130 may include Random Access Memory (RAM), Read Only Memory (ROM), mass storage, removable storage, volatile read and write memory, and the like, or any combination thereof. In some embodiments, the storage device 130 may be implemented on a cloud platform.
User terminal 140 refers to one or more terminal devices or software used by a user. The user terminal 140 may include at least a processing unit, a display unit, an input/output unit, a storage unit, and the like. In some embodiments, the user may make a purchase of goods through the user terminal, and upload information/data to the server 110 or the storage device 130 through the user terminal. In some embodiments, the user terminal 140 may be one or more users, may include users who directly use the service, and may also include other related users.
The network 150 may connect the various components of the system and/or connect the system with external resource components. The network 150 allows communication between the various components, as well as with other components outside the system. For example, the processor 120 may obtain the location information of the user from the user terminal 140 through the network 150. As another example, the processor 120 may obtain user-related information (e.g., consumption situation, location information, consumption habits, etc. of the user) and store-related information (e.g., store location information of the store, store epidemic prevention assessment information, store inventory information, commodity characteristics, etc.) from the storage device 130 via the network 150. In some embodiments, the processor 120 may also send the goods recommendation to the user terminal 140 via the network 150.
FIG. 2 is a block diagram of a machine learning based merchandise recommendation system according to some embodiments of the present description.
In some embodiments, the machine learning based item recommendation system module 200 may include a determination module 210, a recommendation module 220.
In some embodiments, the determining module 210 is configured to determine whether the user is in the quarantine state based on the consumption and location information of the user within the first time. For more on determining whether the user is in the isolated state, see fig. 3 and its associated description.
In some embodiments, the determining module 210 is further configured to determine the evaluation score of the user based on a difference between the purchasing habits of the user in the first time period and the purchasing habits of the user in a second time period, wherein the second time period is a time period before the first time period and has the same length; determining whether a user is in an isolated state according to the evaluation score of the user. For more on the evaluation score and the difference of purchasing habits, see FIG. 4 and its related description
In some embodiments, the recommendation module 220 is configured to determine, via a recommendation model, goods to recommend to the user in response to the user being in the quarantine state. See figure 3 and its associated description for more about the goods recommended to the user.
In some embodiments, the recommendation module 220 is further configured to determine a candidate recommended good based on the store location information and the store epidemic prevention assessment information; determining the recommended good from the candidate recommended goods through a recommendation model. For more on store related information, see fig. 5 and its associated description.
In some embodiments, recommendation module 220 is further for the recommendation module to further determine potential isolated users within a proximity based on the location information of the user; determining the consumption condition of the potential isolation user, and determining a reference commodity based on the consumption condition of the potential isolation user; and determining the recommended commodity through a recommendation model based on the reference commodity. For more on the potential isolated users and reference goods, see FIG. 6 and its associated description.
It should be noted that the above descriptions of the merchandise recommendation system and the modules thereof are only for convenience of description, and the description is not limited to the scope of the illustrated embodiments. It will be appreciated by those skilled in the art that, given the teachings of the present system, any combination of modules or sub-system configurations may be used to connect to other modules without departing from such teachings. In some embodiments, the determining module 210 and the recommending module 220 disclosed in fig. 2 may be different modules in a system, or may be a module that implements the functions of two or more modules described above. For example, each module may share one memory module, or each module may have its own memory module. Such variations are within the scope of the present disclosure.
FIG. 3 is an exemplary flow diagram of a method for machine learning based merchandise recommendation in accordance with some embodiments of the present description. In some embodiments, flow 300 may be performed by a processing device (e.g., processor 120). For example, flow 300 may be stored in a storage device in the form of a program or instructions, which when executed by a server or the modules shown in FIG. 2, may implement flow 300. As shown in fig. 3, the process 300 may include the following steps:
step 310, determining whether the user is in the isolated state based on the consumption condition and the location information of the user in the first time. In some embodiments, step 310 may be performed by the decision module 210.
A user may refer to a consumer selling a service platform. The user may view and/or purchase the items of the off-line store at the sales service platform. In some embodiments, the User may be determined by a User Identification (UID), which may be a unique and unalterable account tag generated by the sales service platform when the sales service platform registers for an account. For example, the sales service platform, upon receiving the user, may determine the particular user from the UID in the user order.
The first time may refer to a period of time before the present. In some embodiments, the first time may be determined according to user behavior, for example, the first time may be a period of time since the location of the user changed, and for example, when the user returned from tianjin to beijing 10 days ago, the first time of the user may be from 10 days ago to now. In some embodiments, the first time may be determined based on the user's purchasing behavior, for example, the first time may be a time period up to date when the user first placed an order within the user's quarantine or home hours.
The consumption situation of the user may refer to a commodity consumption situation of the user, for example, the consumption situation may be a commodity list successfully transacted by the user, where the commodity list may include related information such as name, quantity, price, time, and the like of commodities purchased by the user. In some embodiments, the consumption profile of the user may include the consumption profile of the goods of the user at the sales service platform and the consumption profile of the goods of the user at the store. For example, when a user purchases an item at a store, an account of the purchased item may be paid for by a cashier system of the sales service platform.
The location information may be current geographic information of the user when purchasing the goods, for example, when the user purchases the goods through the sales service platform, the location information may be location information corresponding to the location information of the user terminal. For another example, when the user purchases a product at a store, the location information may be location information of the store.
The consumption condition and the location information of the user in the first time may be determined according to the consumption order of the user, for example, when the user purchases a commodity through the sales service platform, the consumption order may be generated according to the consumption condition of the user, that is, obtaining the consumption condition and the location information of the user in the first time may be achieved by obtaining and analyzing the consumption order of the user in the first time. For another example, when a user purchases a commodity in a store, the banking system of the store may be in communication connection with the sales service platform, and when the user pays a bill, the sales service platform may generate a corresponding consumption order according to the consumption condition of the user.
An isolated state may refer to a state in which a user is primarily active indoors, with little or no egress, for a period of time. For example, an isolation status may refer to a status where an infected or suspected patient lives prohibitively out at a particular location (e.g., an isolation point). As another example, an isolation state may also refer to a user actively living during an influenza outbreak
In some embodiments, whether the user is in the isolated state may be determined based on the isolated commodity and a non-isolated commodity, wherein the non-isolated commodity may refer to a commodity that the user cannot purchase during isolation, such as a fresh meat commodity, a large appliance, an outdoor sports commodity, and the like. The isolated commodity may refer to a commodity which a user has a strong desire to purchase during isolation, such as a mask, instant noodles, milk, sealed meat products, and the like.
In some embodiments, the user may be determined to be in the quarantine state based on the consumer condition of the quarantined goods and the non-quarantined goods by the user reflecting the quarantine likelihood of the user. For example, the quarantined merchandise may reflect the user's home wishes, and the user may be determined to be in the quarantined state when the user purchases more quarantined merchandise. For example, a user may be considered to be in the quarantine state or about to enter the quarantine state when the user places an order for a commodity belonging to only the quarantine commodity from the first time period.
In some embodiments, the aforementioned user is considered to be in the quarantine state if the type and/or quantity of the quarantine duration item in the item data is greater than the first threshold. The first threshold may be determined according to actual conditions, for example, the first threshold may be a preset numerical value, and for example, the first threshold may be 2 kinds, 10 kinds, that is, the kind of the isolation period goods in the goods data is greater than 2, and if the number is greater than 10, it may be determined that the user is in the isolation state. In some embodiments, a user is considered not to be in the quarantine state when the number or type of non-quarantined goods in the user's consumption and location information is greater than a threshold.
In some embodiments, the evaluation score of the user may be determined according to the consumption condition of the user, and then whether the user is in the isolation state may be determined based on the evaluation score of the user. Reference may be made to fig. 4 and its associated description for specific content regarding determining whether a user is in an isolated state based on an evaluation score.
In response to the user being in the quarantine state, a recommended good to be pushed to the user is determined via the recommendation model, step 320. In some embodiments, step 320 may be performed by recommendation module 220.
The recommended commodity is a commodity recommended to the user, and may also be referred to as a recommended commodity. The recommended goods may be goods related to the user isolation status and the user preference, which are determined from goods other than the non-isolated goods, for example, the recommended goods may be goods whose user shopping preference or shopping demand is not affected by the isolation status, such as skin care products, tobacco and wine goods, and the like. Also, for example, the recommended commodity may be an isolated commodity such as mineral water, instant noodles, or the like. In some embodiments, the recommended merchandise may also be merchandise purchased by other users in the quarantine state.
In some embodiments, the items that need to be recommended to the user (i.e., recommended items) may be determined by the first recommendation model. The input of the first recommendation model may include the user-related information in an isolated state, and the output of the first recommendation model may be the recommended goods. The user-related information may include basic information of the user and/or consumption conditions of the user, for example, the user-related information may include location information, gender, age, and the like of the user. For another example, the user-related information may include information about the type, quantity, price, time of purchase, location of purchase, etc. of the goods in each trade order that the user successfully trades in the sales service platform.
In some embodiments, the first recommendation model may be a Convolutional Neural Network (CNN), a Deep Neural Network (DNN), a Long Short-Term Memory (LSTM) model, or the like.
In some embodiments, the first recommendation model may be other models, and the recommendation model may be machine learning algorithms or models such as a support vector machine model, a Logistic regression model, a naive bayes classification model, a gaussian distribution bayes classification model, a decision tree model, a random forest model, and a KNN classification model
The first recommendation model may be trained based on historical data. The training sample with the label can be input into the initial first recommendation model, parameters of the initial first recommendation model are trained and updated through optimization methods such as random gradient descent, when the trained model meets preset conditions, the training is finished, and the trained recommendation model is obtained.
The training sample may be historical user-related data in the historical data, for example, the historical user-related data may include basic information of the user and historical consumption situation of the user, and for example, the historical consumption situation may include information of commodity type, quantity, price, purchase time, purchase location, and the like in each trade order in which the user successfully trades in the sales service platform. In some embodiments, the historical users may include users in the historical data that are in an isolated state. The trained labels are commodities historically purchased by the user.
In some embodiments, the input of the first recommendation model may further include store-related data, where the store-related data may include related information such as inventory, types of items, prices of items, sales promotion of stores, and information on epidemic prevention evaluation of stores. In some embodiments, the store may be a store selected by the user at the sales service platform and/or a nearby store determined from the user's location information.
FIG. 4 is an exemplary flow diagram of an isolation status determination method according to some embodiments described herein. In some embodiments, the flow 400 may be performed by a processing device (e.g., the processor 120) or a merchandise recommendation system module (e.g., the determination module 210). For example, the process 400 may be stored in a storage device in the form of a program or instructions, which when executed by a server or the modules shown in fig. 2, may implement the process 400. As shown in fig. 4, the process 400 may include the following steps:
at step 410, an evaluation score of the user is determined based on a difference between the user's consumption habits at the first time and the user's consumption habits at the second time. In some embodiments, step 410 may be performed by decision module 210.
The second time may be a period of time before the first time, and the length of the second time may be the same as the first time. In some embodiments, the second time may be a time when the user is in a non-isolated state. In some embodiments, to avoid that the user in the second time is also in the isolated state, the time distance between the second time and the first time may be greater than a preset threshold, for example, the last day of the second time is at least 14 days before the first day of the first time.
The consumption habits can be used for describing shopping demands and shopping preferences of the user. The shopping demand may refer to an item that a user needs to shop for, and the shopping preference may refer to a shopping tendency of the user for a characteristic of the item when purchasing the item, for example, the shopping preference of the user may include a tendency to purchase an item with a smaller packing size.
In some embodiments, consumption habits may be described by a consumption feature vector of the user. In some embodiments, the consumption feature vector may include consumption of the user under each commodity category, for example, the consumption feature vector may be a purchase amount of the user under each commodity category, and exemplary consumption feature vectors may be (a, b, c, d, e), where a value at a-e may represent a number of purchases of a certain commodity category, a may represent a purchase amount of a drinking water category during the time period, b may represent a purchase amount of a convenience food category during the time period, c may represent a purchase amount of a can category during the time period, d may represent a purchase amount of a daily commodity category during the time period, e may represent a purchase amount of a wine and tobacco during the time period, and the like. For example, if the consumption feature vector of a certain user may be (3, 5, 5, 1), then 3 drinking water commodities, 5 convenience food commodities, 5 can commodities, and 1 tobacco and wine commodities are purchased on behalf of the user. In some embodiments, the elements in the consumption feature vector may also include other commodity types, which may be determined according to actual conditions.
In some embodiments, the consumption feature vector of the user can be determined by the consumption condition of the user in the first time to represent the consumption habit of the user in the first time. In some embodiments, the consumption feature vector of the user may be determined by the consumption condition of the user in the second time to represent the consumption habit of the user in the second time.
In some embodiments, the second time may also be a time segment group sampled multiple times, that is, multiple time segments with the same duration as the first time may be selected as the second time, and the consumption habit of the second time may be an average value of the consumption habits of the respective time segments.
In some embodiments, the consumption habit difference may be determined from a difference between the first time consumption feature vector and the second time consumption feature vector. For example, by calculating the difference between the two, calculating the vector distance between the two, etc.
The evaluation score may be used to describe the likelihood that the user is in an isolated state. In some embodiments, the evaluation score reflects a difference in consumption habits of the user at the first time and the second time. I.e. the assessment score is related to the difference in consumption habits. In some embodiments, the specific mapping relationship between the evaluation score and the consumption habit difference may be determined according to actual needs, for example, the evaluation score is positively correlated with the consumption habit difference, that is, when the consumption habit difference is large, the evaluation score is large, and thus the possibility that the user is in an isolated state is high. Otherwise, the evaluation score is smaller. In some embodiments, the consumption habit difference may be in a positive correlation non-linear relationship with the evaluation score.
In some embodiments, the consumption habit difference may be determined by a vector distance of the first time consumption feature vector and the second time consumption feature vector. The evaluation score can be determined according to the vector distance between the first time consumption feature vector and the second time consumption feature vector when calculating the evaluation score, wherein the vector distance can be determined according to the Euclidean distance, Manhattan distance, Chebyshev distance, Mahalanobis distance and other related algorithms. Determining the evaluation score based on the vector distance may be determined by a non-linear function, e.g., an exponential function, a power function, etc. In some embodiments, the evaluation score may range from 0-100.
In some embodiments, the user's evaluation score may be adjusted based on a weighted number of potentially isolated users in the user's vicinity.
In some embodiments, the weighted number of potential isolated users may be determined from the evaluation scores of the potential isolated users. The evaluation score of the potential isolation user can be determined according to the difference between the first time consumption feature vector of the potential isolation user and the second time consumption feature vector of the potential isolation user, and the difference can be determined by vector difference, inter-vector distance and the like. Only the differences in their own consumption habits over the first time period and the second time period are considered in determining the evaluation scores for the potential isolated users, regardless of the potential isolated users being present in the vicinity of the potential isolated users. For example, the user neighborhood includes 3 potential isolated users, whose evaluation scores are 90, 80, and 70, respectively, and the corresponding weighted numbers are determined to be 0.9, 0.8, and 0.7 according to the evaluation scores of the three potential isolated users, and it can be considered that (0.9+0.8+0.7) ═ 2.4 potential isolated users exist in the user neighborhood.
In some embodiments, adjusting the evaluation score of the user based on the weighted number of potential isolated users may include adjusting the evaluation score of the user by calculation based on the weighted number of potential isolated users. For example, an adjustment coefficient is determined according to the weighted number of potential isolated users and the total number of users in the user neighborhood, and the evaluation score is adjusted according to the adjustment coefficient, for example, if the evaluation score of the user a is 80 regardless of the potential isolated users in the neighborhood, the weighted number of potential isolated users in the neighborhood is 2.4, and the total number of users in the neighborhood is 20, the adjustment coefficient may be 2.4/20 — 0.12, and the adjusted evaluation score is 80 (1+1.12) or 89.6. For another example, an evaluation score balance value is determined according to the weighted number of potential isolated users, and the evaluation score is adjusted based on the balance value, for example, if the score of the potential isolated user in the neighboring area is 70, and the weighted number of potential isolated users in the neighboring area is 2.4, the adjusted evaluation score is 70+ 2.4-72.4.
In some embodiments, adjusting the evaluation score of a user based on the weighted number of potential isolated users may also be determined in other ways, as appropriate.
And step 420, determining whether the user is in the isolation state according to the evaluation score of the user. In some embodiments, step 420 may be performed by decision module 210.
In some embodiments, a user may be determined to be in the quarantine state when the user's assessment score is above a threshold, e.g., the assessment score may range from 0-100, and the threshold may be 60, i.e., a user with an assessment score above 60 is determined to be in the quarantine state. The threshold value can also be other fractional values, and can be determined according to actual conditions.
FIG. 5 provides a schematic diagram of a method of determining recommended merchandise, according to some embodiments of the present description. In some embodiments, flow 500 may be performed by processor 120. As shown in fig. 5, the process 500 includes the following steps:
step 510, determine the users in the isolated state and the corresponding user-related information.
For specific content of the user and the user-related information in the isolated state, reference may be made to the related descriptions in step 310 and step 320 in fig. 3, which are not described herein again.
And step 520, determining candidate recommended commodities based on the store position information and the store epidemic prevention evaluation information.
The store location information may include geographic location information of the store, such as latitude and longitude information of the store, or other information that may be used to determine the relative location of the store and the user. In some embodiments, the store in step 520 may be a store selected by the user for shopping, for example, the user may select a store for shopping at the sales service platform, and after selecting a store, a recommendation of goods may be made based on inventory goods of the store. In some embodiments, the store in step 520 may also be a nearby store of the user, wherein when the distance between the store and the user is less than the distance threshold, the store is determined to be a nearby store of the user.
In some embodiments, the store location information may also include information that affects the user's selection of stores, such as surrounding traffic conditions, size of stores, distance traveled between stores and the user, and the like. In some embodiments, the determination of whether the user will select a store may be assisted according to the traffic conditions between the store and the user, for example, when the distance between a certain store and the user is less than a distance threshold, but the store may not be determined as a neighboring store of the user due to a crossing of traffic conditions between the two (e.g., road construction, frequent road congestion, etc.). In some embodiments, the recommendation module 220 may determine the candidate recommended items in the user's nearby stores. For example, all or part of the articles in the adjacent store are taken as the candidate recommended articles.
The store epidemic prevention evaluation information can be the response condition of the store to the epidemic prevention policy and can be used for judging whether the store epidemic prevention measures are executed in place. For example, whether or not a store clerk performs nucleic acid detection regularly, whether or not a store is in a negative disinfection state, and the like. The store epidemic prevention evaluation information can be determined according to epidemic prevention rules in special periods, for example, the epidemic prevention rules stipulate that the nucleic acid detection needs to be carried out on the outer package of the frozen products in the stores every week, and the store epidemic prevention evaluation information can include the detection condition of the nucleic acid on the outer package of the frozen products in the stores. In some embodiments, the recommendation module 220 may determine the candidate recommended good in a store where epidemic prevention measures are in place.
In some embodiments, the recommendation module 220 may determine candidate goods from store location information and store epidemic prevention assessment information. The candidate recommended item may be an item that can be recommended to the user. For example, the recommendation module 220 determines at least one store near the user according to the store location information, determines epidemic prevention situations of the stores according to epidemic prevention evaluation information of the stores, determines merchants with executed epidemic prevention measures as candidate merchants, and determines candidate recommended goods among the candidate merchants. For another example, the recommendation module determines at least one store in which an epidemic prevention measure is performed according to the store epidemic prevention evaluation information, determines stores near the user as candidate merchants according to the location information of each store in the at least one store, and further determines candidate recommended commodities in the candidate merchants. The candidate recommended commodity can be a commodity meeting the epidemic prevention rule in store inventory commodities. In some embodiments, it may be determined whether the goods in the store are in compliance with the epidemic prevention rule according to the store epidemic prevention evaluation information, and the goods in compliance with the epidemic prevention rule may be used as candidate recommended goods, wherein the epidemic prevention rule may be determined according to local relevant regulations. For example, the epidemic prevention rule may include that the outer package of the frozen product in the store needs to be subjected to nucleic acid detection every week, and when the candidate recommended product is determined, it may be determined whether the outer package of the frozen product in the store is subjected to nucleic acid detection every week, and if so, the frozen product in the store may be used as the candidate recommended product. As another example, epidemic prevention rules may include that meat products must report a source, and that meat products cannot be a candidate recommended good when the source of meat products is uncertain in a store.
And step 530, determining recommended commodities through a recommendation model based on the user related information and the candidate recommended commodities.
In some embodiments, the recommended good may be determined from the candidate goods by a second recommendation model. In some embodiments, the second recommendation model may be a Convolutional Neural Network (CNN), Deep Neural Network (DNN), Long Short-Term Memory (LSTM) model, or the like. In some embodiments, the second recommendation model may also be other models, for example, machine learning algorithms or models such as support vector machine model, Logistic regression model, primitive bayesian classification model, gaussian distribution bayesian classification model, decision tree model, random forest model, and KNN classification model.
In some embodiments, the input of the second recommendation model may include the user-related information and the candidate recommended good in the quarantine state, and the output of the second recommendation model may be the recommended good.
In some embodiments, the second recommendation model may be trained based on historical data. The training sample with the label can be input into the initial second recommendation model, parameters of the initial second recommendation model are trained and updated through optimization methods such as random gradient descent, when the trained model meets preset conditions, the training is finished, and the trained recommendation model is obtained. The training sample can be historical user-related data in the historical data, and the label is a commodity which is purchased by the user in history.
For a detailed description of the user-related information and the historical user-related information, reference may be made to the related contents in step 320.
In some embodiments, the recommended goods may also be determined based on the user's preference value for the candidate recommended goods through a preset algorithm. The preference value may describe a user's propensity or probability of purchasing the item, e.g., the higher the preference value of the candidate recommended item, the more likely the user is to purchase the item.
In some embodiments, the recommendation module 220 may generate a user preference feature vector according to the user-related information, and generate a candidate recommended article feature vector according to the candidate recommended article. The user preference feature vector may be determined according to the types of the commodities purchased by the user, for example, the user preference feature vector is (L1, L2, L3), and L1, L2, and L3 respectively represent different types of the commodities purchased by the user; the candidate recommended merchandise feature vector may be determined according to the category of the candidate merchandise, for example, the candidate recommended merchandise feature vector is (R1, R2, R3), and R1, R2, and R3 respectively represent different categories of merchandise purchased by the user. In some embodiments, the preference value of each candidate recommended product may be determined by the user preference feature vector and the candidate recommended product feature vector, for example, the comparison user preference feature vector and the candidate recommended product feature vector may be subjected to vector inner product, and the inner product result may be used as the preference value of each candidate recommended product.
In some embodiments, a candidate recommended commodity having a preference value satisfying a preset rule may be selected as the recommended commodity. In some embodiments, the preset rule may include a preset threshold, that is, when the preference value of the candidate recommended product is greater than a threshold preset value, the candidate recommended product may be regarded as a recommended product and pushed to the user. In some embodiments, the preset rule may include a ranking rule, that is, N top-ranked commodities with preference values in the candidate recommended commodities may be selected as recommended commodities and pushed to the user.
FIG. 6 provides a flow chart of another method of determining recommended merchandise according to some embodiments of the present description. In some embodiments, flow 600 may be performed by processor 120. As shown in fig. 6, the process 600 includes the following steps:
step 610, determining the consumption condition of the potential isolation user in the user adjacent area, and determining the reference goods based on the consumption condition of the potential isolation user.
The neighboring area may refer to a peripheral area of the user in an isolated state, for example, the neighboring area may be an area within a preset distance (e.g., 1.2km, 2.5km, etc.) of the user.
In some embodiments, the proximity may be determined from a user's proximity store, e.g., the user's proximity store may be determined from the user's location information, the proximity being determined based on the radiation range of the proximity store. The radiation range of the adjacent stores can refer to an area where the stores can provide shopping services, and the radiation range can be related to the size, the geographic position and the traffic condition of the stores.
A potential isolated user may refer to a user that may be in an isolated state, for example, a potential isolated user may include a user having an evaluation score of no less than 60 points. For a specific determination method of the isolation status, reference may be made to step 310 and the related description in fig. 4, which is not repeated herein.
The reference item may be an item of merchandise purchased by the potential quarantine user, e.g., a purchase record of the ready-to-eat food is included in the consumption profile of potential quarantine user a, then the ready-to-eat food may be used as the reference item.
The reference commodity may be determined in some embodiments based on the purchase of the potential quarantine user. For example, the recommending module 220 may obtain historical purchase information of the potential isolated user from the server according to the determined potential isolated user, and determine the reference commodity according to commodity information in the historical purchase information. All commodities in the historical purchased commodities can be used as reference commodities, and isolated commodities can be used as reference commodities, and the specific situation can be determined according to actual conditions.
In step 620, the weight of the recommended item is adjusted based on the reference item.
In some embodiments, the recommendation module 220 may determine the recommended goods through a recommendation model and adjust the weight of the recommended goods according to the reference goods. In some embodiments, the recommending module 220 may adjust the weight of the recommended product according to the matching degree of the recommended product and the reference product, and the recommended product with a high weight may be preferentially displayed or marked, wherein the matching degree of the recommended product and the reference product may be determined by the distance between the two feature vectors. For example, the recommending module 220 determines that the goods a, B, C, and D are recommended goods, wherein the reference goods determined according to the consumption conditions of the potential isolation users are the goods B and D, the weights of the goods B and D may be increased when recommending the goods, for example, preferentially display the goods B and D, or mark the goods labels of the goods B and D with a word such as "other isolation/home users have also purchased the goods" to prompt the users.
For details of determining recommended merchandise, see fig. 3, 5 and related description.
It should be noted that the above description of the process 300-600 is only for illustration and description, and does not limit the applicable scope of the present specification. Various modifications and changes to the process 300-600 will be apparent to those skilled in the art in light of this disclosure. However, such modifications and variations are intended to be within the scope of the present description.
Some possible benefits of embodiments of the present disclosure include, but are not limited to: (1) and judging whether the user is in the isolation state or not, and providing accurate commodity recommendation for the user based on the user isolation state. And (2) by comparing the consumption habits of the user in different time periods, the accuracy of judging whether the user is in the isolation state is improved. (3) The commodity is recommended for the user based on the recommendation model, so that the recommended commodity meets the preference of the user, and the commodity recommendation accuracy is improved. (4) Considering the influence of potential isolation users of stores and surrounding areas, group purchase is easier to form when recommending commodities, and the stores are convenient to specify sales plans.
It should be noted that different embodiments may produce different advantages, and in different embodiments, the advantages that may be produced may be any one or combination of the above, or may be any other advantages that may be obtained.
Having thus described the basic concepts, it will be apparent to those skilled in the art that the foregoing detailed disclosure is to be considered merely illustrative and not restrictive of the broad description. Various modifications, improvements and adaptations to the present description may occur to those skilled in the art, although not specifically described herein. Such modifications, improvements and adaptations are proposed in the present specification and thus fall within the spirit and scope of the exemplary embodiments of the present specification.
Also, the description uses specific words to describe embodiments of the description. Reference throughout this specification to "one embodiment," "an embodiment," and/or "some embodiments" means that a particular feature, structure, or characteristic described in connection with at least one embodiment of the specification is included. Therefore, it is emphasized and should be appreciated that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, some features, structures, or characteristics of one or more embodiments of the specification may be combined as appropriate.
Additionally, the order in which elements and sequences of the process are recited in the specification, the use of alphanumeric characters, or the use of other designations, is not intended to limit the order of the processes and methods described in this specification, unless explicitly stated in the claims. While various presently contemplated embodiments of the invention have been discussed in the foregoing disclosure by way of example, it is to be understood that such detail is solely for that purpose and that the appended claims are not limited to the disclosed embodiments, but, on the contrary, are intended to cover all modifications and equivalent arrangements that are within the spirit and scope of the embodiments herein. For example, although the system components described above may be implemented by hardware devices, they may also be implemented by software-only solutions, such as installing the described system on an existing server or mobile device.
Similarly, it should be noted that in the preceding description of embodiments of the present specification, various features are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the embodiments. This method of disclosure, however, is not intended to imply that more features than are expressly recited in a claim. Indeed, the embodiments may be characterized as having less than all of the features of a single embodiment disclosed above.
Where numerals describing the number of components, attributes or the like are used in some embodiments, it is to be understood that such numerals used in the description of the embodiments are modified in some instances by the modifier "about", "approximately" or "substantially". Unless otherwise indicated, "about", "approximately" or "substantially" indicates that the number allows a variation of ± 20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximations that may vary depending upon the desired properties of the individual embodiments. In some embodiments, the numerical parameter should take into account the specified significant digits and employ a general digit-preserving approach. Notwithstanding that the numerical ranges and parameters setting forth the broad scope of the range are approximations, in the specific examples, such numerical values are set forth as precisely as possible within the scope of the application.
For each patent, patent application publication, and other material, such as articles, books, specifications, publications, documents, etc., cited in this specification, the entire contents of each are hereby incorporated by reference into this specification. Except where the application history document does not conform to or conflict with the contents of the present specification, it is to be understood that the application history document, as used herein in the present specification or appended claims, is intended to define the broadest scope of the present specification (whether presently or later in the specification) rather than the broadest scope of the present specification. It is to be understood that the descriptions, definitions and/or uses of terms in the accompanying materials of this specification shall control if they are inconsistent or contrary to the descriptions and/or uses of terms in this specification.
Finally, it should be understood that the embodiments described herein are merely illustrative of the principles of the embodiments of the present disclosure. Other variations are also possible within the scope of the present description. Thus, by way of example, and not limitation, alternative configurations of the embodiments of the specification can be considered consistent with the teachings of the specification. Accordingly, the embodiments of the present description are not limited to only those embodiments explicitly described and depicted herein.

Claims (10)

1. A commodity recommendation method based on machine learning comprises the following steps:
judging whether a user is in an isolated state or not based on the consumption condition and the position information of the user in the first time;
and determining recommended commodities pushed to the user through a recommendation model in response to the user being in the isolation state.
2. The commodity recommendation method of claim 1, wherein the determining whether the user is in an isolated state based on the consumption condition and the location information of the user within the first time comprises:
determining an evaluation score of the user based on a difference between the consumption habits of the user in the first time period and the consumption habits of the user in a second time period, wherein the second time period is a time period which is before the first time period and has the same length;
determining whether a user is in an isolated state according to the evaluation score of the user.
3. The merchandise recommendation method of claim 1, wherein determining, by a recommendation model, recommended merchandise to push to the user in response to the user being in an isolated state comprises:
determining candidate recommended commodities based on the store position information and the store epidemic prevention evaluation information;
and determining the recommended commodity from the candidate recommended commodities through the recommendation model.
4. The merchandise recommendation method of claim 1, wherein determining, by the recommendation model, recommended merchandise to push to the user in response to the user being in the quarantine state comprises:
determining potential isolated users within a proximity area based on the location information of the users;
determining the consumption condition of the potential isolation user, and determining a reference commodity based on the consumption condition of the potential isolation user;
determining the recommended commodity through a recommendation model based on the reference commodity.
5. A machine learning based merchandise recommendation system comprising:
the judging module is used for judging whether the user is in an isolated state or not based on the consumption condition and the position information of the user in the first time;
and the recommending module is used for responding to the condition that the user is in the isolation state and determining the commodities recommended to the user through a recommending model.
6. The item recommendation system of claim 5, the determination module further to:
determining an evaluation score of the user based on a difference between the purchasing habits of the user in the first time period and the purchasing habits of the user in a second time period, wherein the second time period is a time period which is before the first time period and has the same length;
determining whether a user is in an isolated state according to the evaluation score of the user.
7. The merchandise recommendation system of claim 5, the recommendation module further to:
determining candidate recommended commodities based on the store position information and the store epidemic prevention evaluation information;
and determining the recommended commodity from the candidate recommended commodities through a recommendation model.
8. The merchandise recommendation system of claim 5, the recommendation module further to:
determining potential isolated users within a vicinity based on the location information of the users;
determining the consumption condition of the potential isolation user, and determining a reference commodity based on the consumption condition of the potential isolation user;
determining the recommended commodity through a recommendation model based on the reference commodity.
9. A machine learning-based commodity recommendation device comprising a processor for executing the machine learning-based commodity recommendation method of any one of claims 1 to 4.
10. A computer-readable storage medium storing computer instructions, which when read by a computer, cause the computer to execute the method for recommending commodities based on machine learning according to any one of claims 1 to 4.
CN202210259968.0A 2022-03-16 2022-03-16 Commodity recommendation method based on machine learning Pending CN114971761A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210259968.0A CN114971761A (en) 2022-03-16 2022-03-16 Commodity recommendation method based on machine learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210259968.0A CN114971761A (en) 2022-03-16 2022-03-16 Commodity recommendation method based on machine learning

Publications (1)

Publication Number Publication Date
CN114971761A true CN114971761A (en) 2022-08-30

Family

ID=82976285

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210259968.0A Pending CN114971761A (en) 2022-03-16 2022-03-16 Commodity recommendation method based on machine learning

Country Status (1)

Country Link
CN (1) CN114971761A (en)

Similar Documents

Publication Publication Date Title
US11017422B2 (en) Dynamically generating digital product notifications over time based on product expiration data
US11341554B1 (en) Software platform to manage shoppers to fulfill orders for items received by an online concierge system
BR112019027187A2 (en) system for aggregating metadata of consumer behavior events in the store, data verification and artificial intelligence analysis for data interpretation and triggering associated action
KR20200108058A (en) Recommended device and method
US20230196389A1 (en) Training a model to predict likelihoods of users performing an action after being presented with a content item
CN112041874A (en) Computer-implemented method for generating suggestion list and system for generating order list
US20220179909A1 (en) Attribute node widgets in search results from an item graph
JP2021099718A (en) Information processing device and program
US20240311397A1 (en) Clustering data describing interactions performed after receipt of a query based on similarity between embeddings for different queries
US20240095673A1 (en) Creation and arrangement of items in an online concierge system-specific portion of a warehouse for order fulfillment
US11748339B2 (en) Anonymization of query information while retaining query structure and sizing information
JP2021513166A (en) Customer personalization control units, systems, and methods
WO2023154094A1 (en) Removing semantic duplicates from results based on similarity between embeddings for different results
CN115271884A (en) Multi-source data-based commodity selection method and device and electronic equipment
CN114971761A (en) Commodity recommendation method based on machine learning
US11978087B2 (en) Using a genetic algorithm to identify a balanced assignment of online system users to a control group and a test group for performing a test
US20240177226A1 (en) Dynamic augmented reality and gamification experience for in-store shopping
US20220318878A1 (en) Digital preferences based on physical store patterns
US20230385911A1 (en) Automatic selection of dynamic data entries for multiple dynamic databases
US20240070491A1 (en) Simulating an application of a treatment on a demand side and a supply side associated with an online system
US20240177211A1 (en) Identifying candidate replacement items with a source similarity score
US20230186367A1 (en) Transaction Recommendation and Purchasing Engine
US20240177212A1 (en) Determining search results for an online shopping concierge platform
US20240046313A1 (en) Pacing the delivery of content campaigns in an online concierge system using cross-retailer inventory stock levels
US20240070605A1 (en) Selectively providing machine learning model-based services

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