CN113516539A - Commodity recommendation method and device and computer-readable storage medium - Google Patents

Commodity recommendation method and device and computer-readable storage medium Download PDF

Info

Publication number
CN113516539A
CN113516539A CN202110868342.5A CN202110868342A CN113516539A CN 113516539 A CN113516539 A CN 113516539A CN 202110868342 A CN202110868342 A CN 202110868342A CN 113516539 A CN113516539 A CN 113516539A
Authority
CN
China
Prior art keywords
commodity
historical
behavior
service
user
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
CN202110868342.5A
Other languages
Chinese (zh)
Inventor
柯于皇
徐恒
刘锋
刘朋振
曹湘
卓亚丽
刘磊
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.)
China Mobile Communications Group Co Ltd
China Mobile Hangzhou Information Technology Co Ltd
Original Assignee
China Mobile Communications Group Co Ltd
China Mobile Hangzhou Information Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China Mobile Communications Group Co Ltd, China Mobile Hangzhou Information Technology Co Ltd filed Critical China Mobile Communications Group Co Ltd
Priority to CN202110868342.5A priority Critical patent/CN113516539A/en
Publication of CN113516539A publication Critical patent/CN113516539A/en
Pending legal-status Critical Current

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
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9536—Search customisation based on social or collaborative filtering

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Finance (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention discloses a commodity recommendation method, a commodity recommendation device and a computer readable storage medium, wherein the commodity recommendation method comprises the following steps: the method comprises the steps of obtaining historical behavior data of a user, wherein the historical behavior data comprises historical business behavior data and historical commodity behavior data, determining commodity behavior data corresponding to each historical business according to the historical commodity behavior data, determining historical business preferred by the user according to the historical business data and the commodity behavior data corresponding to each historical business, and determining recommended commodities corresponding to the user according to a commodity set preferred by the user and corresponding to the historical business. The method and the device can improve the accuracy of commodity recommendation under the condition that the behavior data of the commodity is insufficient.

Description

Commodity recommendation method and device and computer-readable storage medium
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a method and an apparatus for recommending a commodity, and a computer-readable storage medium.
Background
With the development of the smart home field, when an operator uses the method to recommend commodities to a user, especially to recommend hardware commodities of a smart home, the accuracy of the recommendation needs to be improved, when the recommendation is performed, the recommendation can be performed based on the interaction behavior of the user to the commodities, for example, based on the browsing behavior of detailed pages of the commodities, the purchasing behavior of the commodities, and the like, however, when the recommendation is performed by using the method, the consumption habits of the user can be accurately analyzed only by generating a large number of behaviors by the user, so that the accurate commodity recommendation can be obtained, the method is limited by that the use data of the hardware commodity store in the application of the operator is generally less, so that the behavior data of the hardware commodities is insufficient, so that the accuracy of the commodity recommendation is reduced, and therefore, a scheme for improving the accuracy of the commodity recommendation under the condition that the behavior data of the commodities is insufficient is required to be provided, based on this, the invention at least solves the following technical problems: how to improve the accuracy of commodity recommendation under the condition that the behavior data of the commodity is insufficient.
Disclosure of Invention
The invention mainly aims to provide a commodity recommendation method, a commodity recommendation device and a computer readable storage medium, and aims to solve the technical problem of improving the commodity recommendation accuracy rate under the condition of insufficient behavior data of commodities.
In order to achieve the above object, the present invention provides a commodity recommendation method, including:
acquiring historical behavior data of a user, wherein the historical behavior data comprises historical business behavior data and historical commodity behavior data;
determining commodity behavior data corresponding to each historical service according to the historical commodity behavior data;
determining the historical business preferred by the user according to the historical business data and the commodity behavior data corresponding to each historical business;
and determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user.
Optionally, the step of obtaining the historical behavior data of the user and the step of determining the recommended goods corresponding to the user according to the goods set corresponding to the historical service preferred by the user further include:
determining a candidate commodity set according to the historical commodity behavior data and preset commodity attributes, wherein the candidate commodity set comprises at least one of a brand preference commodity set, a category preference commodity set, a hot commodity set and a historical preference commodity set;
the step of determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user comprises the following steps:
and selecting the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user and the candidate commodity set.
Optionally, the step of determining a candidate commodity set according to the historical commodity behavior data and the preset commodity attribute includes:
determining a commodity behavior type corresponding to each historical commodity and a behavior attribute value corresponding to the commodity behavior type according to the historical commodity behavior data;
determining a preference value corresponding to each historical commodity according to the commodity behavior type corresponding to each historical commodity and the behavior attribute value corresponding to the commodity behavior type;
and determining the candidate commodity set according to the preference value corresponding to each historical commodity and the preset commodity attribute, wherein the preset commodity attribute comprises at least one of a commodity brand, a commodity category and a commodity heat degree.
Optionally, the step of determining, according to the behavior type corresponding to each of the commodities and the behavior attribute value corresponding to the behavior type, a preference value corresponding to each of the commodities includes:
the behavior attribute value corresponding to each commodity behavior type and the weight associated with the commodity behavior type are multiplied to obtain a behavior preference value corresponding to each commodity behavior type, the commodity behavior type comprises at least one of commodity link clicking, commodity detail page browsing, commodity video playing, customer service contact, shopping cart adding, order submitting, purchasing and evaluation, the behavior attribute value comprises at least one of behavior frequency, behavior aging coefficient and behavior duration, and the aging coefficient is smaller as the time point of the commodity behavior type is earlier;
and summing all the behavior preference values corresponding to the historical commodities respectively to obtain the preference value corresponding to each historical commodity.
Optionally, the step of selecting the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user and the candidate commodity set includes:
determining target commodity information according to a commodity set corresponding to the historical service preferred by the user and the candidate commodity set;
determining a feature vector corresponding to the user according to the target commodity information, the user attribute information of the user and the historical behavior data, wherein the user attribute information comprises at least one of age, gender, occupation, city, consumption capability and character preference of the user;
converting the feature vector into a target feature according to a preset gradient lifting tree model;
determining the click probability of the user on each target commodity according to a preset logistic regression model and the target characteristics;
and selecting the recommended commodity from all the target commodities according to the click probability of each target commodity.
Optionally, the step of determining target commodity information according to the commodity set corresponding to the historical service preferred by the user and the candidate commodity set includes:
determining historical commodities corresponding to all historical behaviors according to the historical behavior information;
filtering the historical commodities in a commodity set corresponding to the historical service preferred by the user and the candidate commodity set to obtain a target commodity set;
and acquiring the target commodity information corresponding to each target commodity in the target commodity set.
Optionally, the step of determining the historical service preferred by the user according to the historical service data and the commodity behavior data corresponding to each historical service includes:
determining a service behavior type corresponding to each historical service and a service behavior attribute value corresponding to the service behavior type according to the historical service data, wherein the service behavior type comprises at least one of browsing, clicking, handling, consulting customer service and complaints of the service, the service behavior attribute value comprises at least one of frequency, aging coefficient and duration of the service behavior, and the aging coefficient is lower as the time point corresponding to the service behavior type is earlier;
determining the commodity behavior type and the commodity behavior attribute value of the commodity corresponding to each historical service according to the commodity behavior data corresponding to each historical service;
determining a first preference value corresponding to each historical service according to the service behavior type corresponding to each historical service and the service behavior attribute value;
determining a second preference value of the historical commodity corresponding to each historical service according to the commodity behavior type of the historical commodity corresponding to each historical service and the commodity behavior attribute value;
determining a target preference value corresponding to each historical service according to a first preference value corresponding to each historical service and a second preference value of a historical commodity corresponding to each historical service;
and determining the historical services preferred by the user according to the target preference value corresponding to each historical service.
In addition, in order to achieve the above object, the present invention further provides a product recommendation device, including an obtaining module and a determining module, wherein:
the acquisition module is used for acquiring historical behavior data of a user, wherein the historical behavior data comprises historical business behavior data and historical commodity behavior data;
the determining module is used for determining commodity behavior data corresponding to each historical business according to the historical commodity behavior data; determining the historical business preferred by the user according to the historical business data and the commodity behavior data corresponding to each historical business; and determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user.
In addition, in order to achieve the above object, the present invention further provides a product recommendation device, including a memory, a processor, and a product recommendation program stored in the memory and executable on the processor, wherein the product recommendation program, when executed by the processor, implements the steps of the product recommendation method according to any one of the above aspects.
In addition, to achieve the above object, the present invention further provides a computer-readable storage medium having an article recommendation program stored thereon, where the article recommendation program, when executed by a processor, implements the steps of the article recommendation method according to any one of the above.
The commodity recommending method, the commodity recommending device and the computer-readable storage medium provided by the embodiment of the invention have the advantages that by acquiring historical behavior data of a user, wherein the historical behavior data comprises historical business behavior data and historical commodity behavior data, determining commodity behavior data corresponding to each historical business according to the historical commodity behavior data, determining historical business preferred by the user according to the historical business data and the commodity behavior data corresponding to each historical business, and determining recommended commodities corresponding to the user according to a commodity set preferred by the user and corresponding to the historical business, when commodity recommendation is carried out, the preferred business of the user is determined by combining the behaviors of the user on the historical business and the behaviors of the commodities corresponding to the historical business, and the recommended commodities corresponding to the user are determined by giving the commodity set preferred by the user and corresponding to the preferred business, even if the behavior data on the commodities are insufficient, the method can determine the favorite service of the user by combining the behavior of the historical service and the commodity behavior corresponding to the historical service, and recommend the favorite service based on the commodity corresponding to the service, and can more accurately determine the preference of the user because the commodity recommendation is performed based on the service behavior dimension and the commodity behavior corresponding to the service, so that the commodity recommending accuracy can be improved.
Drawings
FIG. 1 is a schematic diagram of the structure of an apparatus according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a first exemplary embodiment of a merchandise recommendation method according to the present invention;
FIG. 3 is a flowchart illustrating a second embodiment of a merchandise recommendation method according to the present invention;
FIG. 4 is a flowchart illustrating a merchandise recommendation method according to a third embodiment of the present invention;
FIG. 5 is a flowchart illustrating a fourth exemplary embodiment of a merchandise recommendation method according to the present invention;
fig. 6 is a schematic block diagram of a product recommendation device according to an embodiment of the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, fig. 1 is a schematic structural diagram of an apparatus according to an embodiment of the present invention.
The device related to the embodiment of the invention can be a server or other computer equipment.
As shown in fig. 1, the apparatus may include: a processor 1001, such as a CPU, a memory 1002, and a communication bus 1003. The communication bus 1003 is used to implement connection communication among these components. The memory 1002 may be a high-speed RAM memory or a non-volatile memory (e.g., a disk memory). The memory 1002 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the configuration of the device shown in fig. 1 is not intended to be limiting of the device and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, a memory 1002, which is a kind of computer storage medium, may include therein an article recommendation program.
In the apparatus shown in fig. 1, the processor 1001 may be configured to call the goods recommendation program stored in the memory 1002 and perform the following operations:
acquiring historical behavior data of a user, wherein the historical behavior data comprises historical business behavior data and historical commodity behavior data;
determining commodity behavior data corresponding to each historical service according to the historical commodity behavior data;
determining the historical business preferred by the user according to the historical business data and the commodity behavior data corresponding to each historical business;
and determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user.
Further, the processor 1001 may call the goods recommendation program stored in the memory 1002, and further perform the following operations:
determining a candidate commodity set according to the historical commodity behavior data and preset commodity attributes, wherein the candidate commodity set comprises at least one of a brand preference commodity set, a category preference commodity set, a hot commodity set and a historical preference commodity set;
the step of determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user comprises the following steps:
and selecting the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user and the candidate commodity set.
Further, the processor 1001 may call the goods recommendation program stored in the memory 1002, and further perform the following operations:
determining a behavior type corresponding to each commodity and a behavior attribute value corresponding to the behavior type according to the historical commodity behavior data;
determining a preference value corresponding to each commodity according to the behavior type corresponding to each commodity and the behavior attribute value corresponding to the behavior type;
and determining the candidate commodity set according to the preference value corresponding to each commodity and the preset commodity attribute, wherein the preset commodity attribute comprises at least one of a commodity brand, a commodity category and a commodity heat degree.
Further, the processor 1001 may call the goods recommendation program stored in the memory 1002, and further perform the following operations:
the behavior attribute value corresponding to each behavior type and the weight associated with the behavior type are multiplied to obtain a behavior preference value corresponding to each behavior type, the behavior type comprises at least one of commodity link clicking, commodity detail page browsing, commodity video playing, customer service contact, shopping cart adding, order submitting, purchasing and evaluation, the behavior attribute value comprises at least one of behavior frequency, behavior aging coefficient and behavior duration, and the aging coefficient is smaller as the time point of the behavior type is earlier;
and summing all the behavior preference values corresponding to the commodities respectively to obtain the preference value corresponding to each commodity.
Further, the processor 1001 may call the goods recommendation program stored in the memory 1002, and further perform the following operations:
determining target commodity information according to a commodity set corresponding to the historical service preferred by the user and the candidate commodity set;
determining a feature vector corresponding to the user according to the target commodity information, the user attribute information of the user and the historical behavior data, wherein the user attribute information comprises at least one of age, gender, occupation, city, consumption capability and character preference of the user;
converting the feature vector into a target feature according to a preset gradient lifting tree model;
determining the click probability of the user on each target commodity according to a preset logistic regression model and the target characteristics;
and selecting the recommended commodity from all the target commodities according to the click probability of each target commodity.
Further, the processor 1001 may call the goods recommendation program stored in the memory 1002, and further perform the following operations:
determining historical commodities corresponding to all historical behaviors according to the historical behavior information;
filtering the historical commodities in a commodity set corresponding to the historical service preferred by the user and the candidate commodity set to obtain a target commodity set;
and acquiring the target commodity information corresponding to each target commodity in the target commodity set.
Further, the processor 1001 may call the goods recommendation program stored in the memory 1002, and further perform the following operations:
determining a service behavior type corresponding to each historical service and a service behavior attribute value corresponding to the service behavior type according to the historical service data, wherein the service behavior type comprises at least one of browsing, clicking, handling, consulting customer service and complaints of the service, the service behavior attribute value comprises at least one of frequency, aging coefficient and duration of the service behavior, and the aging coefficient is lower as the time point corresponding to the service behavior type is earlier;
determining the commodity behavior type and the commodity behavior attribute value of the commodity corresponding to each historical service according to the commodity behavior data corresponding to each historical service;
determining a first preference value corresponding to each historical service according to the service behavior type corresponding to each historical service and the service behavior type;
determining a second preference value of the commodity corresponding to each historical service according to the commodity behavior type of the commodity corresponding to each historical service and the commodity behavior attribute value;
determining a target preference value corresponding to each historical service according to a first preference value corresponding to each historical service and a second preference value of a commodity corresponding to each historical service;
and determining the historical services preferred by the user according to the target preference value corresponding to each historical service.
Referring to fig. 2, a first embodiment of the present invention provides a product recommendation method, including:
step S10, obtaining historical behavior data of a user, wherein the historical behavior data comprises historical business behavior data and historical commodity behavior data;
in this embodiment, the execution subject is a commodity recommending apparatus, the commodity recommending apparatus may be a server, the historical behavior data is behavior data of a user in a historical time period, the historical behavior data includes historical service behavior data and historical commodity behavior data, the historical service behavior data is data of behaviors generated for historical services, the historical services are services that already exist before, the historical services are family services such as "mobile housekeeping" service, "family communication" service, "family meeting," and the like, the historical service behavior data is behavior data of clicking, handling, browsing, consulting customer service, complaints and the like for the historical services, the historical services may also be regarded as an existing service, the historical commodity behavior data is data of behaviors generated for historical commodities such as hardware commodities that exist before, hardware commodities such as a camera, a computer, and a computer-readable medium, The actions of the historical commodities include clicking commodity links of the historical commodities, browsing commodity detail pages, playing commodity introduction videos, contacting customer service, adding shopping carts, submitting orders, purchasing and evaluating.
The historical behavior data can be sent by the terminal device of the user at regular time, or actively obtained by the commodity recommending device at regular time, or the commodity recommending device can also obtain the historical behavior data of the user in other ways.
Step S20, determining commodity behavior data corresponding to each historical service according to the historical commodity behavior data;
the historical commodities and the historical services have corresponding relations, for example, each historical service corresponds to at least one historical commodity, each historical commodity corresponds to at least one historical service, and the commodity behavior data corresponding to each historical service can be determined according to the preset corresponding relation between the historical commodities and the historical services and the historical commodity behavior data, so that the historical services corresponding to the commodities can be determined.
For example, the historical service is a "mobile housekeeping" service, and the historical goods corresponding to the "mobile housekeeping" service include a full-color camera, an intelligent smoke detection alarm and an intelligent video doorbell, and for example, the historical service is a "home communication" service, and the historical goods corresponding to the "home communication" service include an intelligent sound box, an intelligent television and an intelligent video doorbell.
The historical commodity behavior data comprises information of historical commodities and information of behaviors of the historical commodities, and each historical commodity of behaviors generated by a user can be determined based on historical commodity behavior data, so that the commodity behavior data corresponding to each historical business can be determined by combining the historical commodity behavior data based on the corresponding relation between the historical commodities and the historical business, and based on different behaviors of the user in a historical time period, the commodity behavior data corresponding to part of the historical business is possible to be null.
Step S30, determining the historical business preferred by the user according to the historical business data and the commodity behavior data corresponding to each historical business;
after historical service data and commodity behavior data corresponding to each historical service are obtained, historical services preferred by a user can be further determined, wherein the historical service data comprises information of the historical services and information of behaviors of the historical services, for example, the historical services are respectively 'mobile housekeeper' and 'family communication', the behaviors of the 'mobile housekeeper' services comprise clicking, browsing and consulting customer service, the behaviors of the 'family communication' services comprise clicking and browsing, the preference of the historical services can be determined according to the behaviors of different historical services, for example, the number of the behaviors of the 'mobile housekeeper' services exceeds the number of the behaviors of the 'family communication' services, the preference degree of the 'mobile housekeeper' services can be determined to be higher, and if the priority of the 'consulting customer service' is higher, the services with the behavior of the 'consulting customer service' can be determined to be more likely to be the services preferred by the user, in order to improve the accuracy of determining the service preferred by the user, the embodiment further determines the historical service preferred by the user by combining the commodity behavior data corresponding to the historical service, wherein the behavior of the historical commodity can reflect the preference of the user to the service laterally, for example, the historical behavior data generated by the user in a certain historical time period includes the behavior data of two kinds of historical commodities, namely a full-color camera and an intelligent smoke detection alarm, and the historical services corresponding to the full-color camera and the intelligent smoke detection alarm are both the "mobile housekeeping" service, so that the preference or interest probability of the user to the "mobile housekeeping" service can be determined to be higher.
When determining the historical service preferred by the user according to the historical service data and the commodity behavior data corresponding to each historical service, determining a first service preferred by the user according to the historical service data, determining a second service preferred by the user according to the commodity behavior corresponding to each historical service, and determining the service preferred by the user in the first service and the second service, for example, determining the first service preferred by the user as a 'mobile housekeeping' service and a 'home communication' service according to the historical service data, determining the second service preferred by the user as a 'mobile housekeeping' service and an 'intelligent talkback' service according to the commodity behavior data corresponding to each historical service, determining the service preferred by the user as a 'mobile housekeeping' service, or determining the first service preferred by the user according to the historical service data, and further combining the commodity behavior data corresponding to each historical service, determining the goods preferred by the user under each service, determining the historical service preferred by the user according to the goods preferred by the user and the first service, for example, determining that the first service preferred by the user is a 'mobile housekeeper' service and a 'home communication' service according to historical service data, determining that the goods generating historical behaviors corresponding to the 'mobile housekeeper' service comprise a full-color camera and an intelligent smoke detection alarm according to the goods behavior data corresponding to each historical service, and determining that the goods generating historical behaviors corresponding to the 'home communication' service are empty, so that the preference degree of the user on the 'mobile housekeeper' service can be determined to be higher, and the 'mobile housekeeper' service can be used as the historical service preferred by the user.
In addition, historical business preferred by the user can be determined in other ways based on historical business data and commodity behavior data corresponding to each historical business.
The number of the history services preferred by the user can be one or more, wherein in order to improve the accuracy of determining the history services preferred by the user and further improve the accuracy of recommending commodities, the number of the history services preferred by the user can be set to be smaller than the number of all preset services, or the number of the history services preferred by the user can be set to be smaller than the number of all the history services, the preset services are all services already existing, the preset services include the history services, for example, the number of the preset services is 10, and the number of the history services determined according to the history behavior data is 5, so that only 2 history services can be used as the history services preferred by the user when the history services preferred by the user are determined, and the history services preferred by the user can be determined more accurately.
And step S40, determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user.
After determining the historical service preferred by the user, acquiring a commodity set corresponding to the historical service preferred by the user, and determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user, wherein all commodities in the commodity set corresponding to the historical service preferred by the user can be directly used as the recommended commodities corresponding to the user, or part of commodities can be acquired as the recommended commodities in the commodity set corresponding to the historical service preferred by the user, and in addition, the recommended commodities corresponding to the user can be further determined by further combining the commodity set corresponding to the historical service preferred by the user and other commodity sets.
After determining the recommended commodity, commodity information of the recommended commodity may also be output, such as sending the commodity information of the recommended commodity to a corresponding user.
In this embodiment, by obtaining historical behavior data of a user, the historical behavior data including historical service behavior data and historical commodity behavior data, determining commodity behavior data corresponding to each historical service according to the historical commodity behavior data, determining historical services preferred by the user according to the historical service data and the commodity behavior data corresponding to each historical service, and determining recommended commodities corresponding to the user according to commodity sets corresponding to the historical services preferred by the user, when commodity recommendation is performed, determining services preferred by the user in combination with behaviors of the user on the historical services and behaviors of the commodities corresponding to the historical services, and giving the commodity sets corresponding to the preferred services to the user, determining the recommended commodities corresponding to the user, even if the behavior data on the commodities is insufficient, determining favorite services of the user in combination with the behaviors of the historical services and the commodity behaviors corresponding to the historical services, the commodity recommendation method based on the business behavior dimension can determine the preference of the user more accurately and recommend the commodity based on the business behavior dimension and the commodity behavior dimension corresponding to the business, and can recommend the commodity based on the business behavior and the commodity behavior dimension corresponding to the business even if the commodity recommendation method based on the business behavior dimension and the commodity behavior dimension corresponding to the business is adopted on the premise that the commodity behavior data is insufficient, so that the problem of low accuracy caused by insufficient data quantity under the condition that the commodity behavior is directly recommended only is solved, and the commodity recommendation accuracy can be improved under the condition that the commodity behavior data is insufficient.
Referring to fig. 3, a second embodiment of the present invention provides a method for recommending merchandise, based on the first embodiment shown in fig. 2, between the step S10 and the step S40, the method further includes:
step S50, determining a candidate commodity set according to the historical commodity behavior data and preset commodity attributes, wherein the candidate commodity set comprises at least one of a brand preference commodity set, a category preference commodity set, a popularity commodity set and a historical preference commodity set;
in order to improve the accuracy of commodity recommendation, in this embodiment, after the historical line data of the user is acquired, a candidate commodity set is determined according to the historical commodity behavior data and the preset commodity attribute, and the recommended commodity corresponding to the user is determined according to the commodity set corresponding to the historical service preferred by the user and the candidate commodity set, so that the commodity set can be acquired from multiple aspects and commodity recommendation is performed, and the commodity is recommended to the user according to the historical behavior data of the user, so as to perform personalized recommendation, thereby improving the accuracy of commodity recommendation.
The preset commodity attribute is an attribute of a preset commodity, the preset commodity attribute comprises at least one of a commodity brand, a commodity category and a commodity heat, the brand preference commodity set is a set formed by commodities corresponding to the brand preferred by the user, the category preference commodity set is a set formed by commodities corresponding to the commodity category preferred by the user, the heat commodity set is a set formed by commodities with higher heat, the higher heat indicates that the sales volume of the commodity is higher, and the history preference commodity set is a commodity set obtained based on historical commodity behavior data and a preset collaborative filtering model.
When determining the candidate commodity set according to the historical commodity behavior data and the preset commodity attributes, the following method may be adopted: determining a commodity behavior type corresponding to each historical commodity and a behavior attribute value corresponding to the commodity behavior type according to the historical commodity behavior data; determining a preference value corresponding to each commodity according to the commodity behavior type corresponding to each historical commodity and the behavior attribute value corresponding to the commodity behavior type; determining a candidate commodity set according to the preference value corresponding to each commodity and the preset commodity attribute so as to improve the accuracy of commodity recommendation; alternatively, as can be understood by those skilled in the art, the candidate set may also be determined according to historical commodity behavior data, preset commodity attributes, and a preset machine learning model.
The commodity behavior type is a type into which the behavior of the user is divided, the commodity behavior type comprises at least one of clicking a commodity link, browsing a commodity detail page, playing a commodity video, contacting customer service, adding a shopping cart, submitting an order, purchasing and evaluating, and the behavior attribute value comprises at least one of behavior frequency, a time efficiency of the behavior and a duration of the behavior, wherein the time efficiency is smaller as the time point of the commodity behavior type is earlier.
When determining the preference value corresponding to each historical commodity according to the commodity behavior type corresponding to each historical commodity and the behavior attribute value corresponding to the commodity behavior type, the following method may be adopted: the behavior attribute value corresponding to each commodity behavior type and the weight associated with the commodity behavior type are multiplied to obtain a behavior preference value corresponding to each commodity behavior type, all behavior preference values corresponding to each historical commodity are summed respectively to obtain a preference value corresponding to each historical commodity, and therefore the commodity recommendation accuracy is improved; or, the product of the behavior attribute values corresponding to each commodity behavior type may be obtained to obtain a behavior preference value corresponding to each commodity behavior type, and the preference values corresponding to each historical commodity are summed respectively to obtain a preference value corresponding to each historical commodity.
The weight associated with the commodity behavior type is used for indicating the importance degree of the behavior type when determining the commodity preferred by the user, the weights of different behavior types may be different or the same, the frequency refers to the number of times of the behavior corresponding to the behavior type, the duration of the behavior refers to the duration of the behavior, the duration of the behavior can also be represented by a duration coefficient corresponding to the duration coefficient, the duration coefficient can be corresponding to the duration in advance, and the duration coefficient is used for replacing the duration when calculating the product of the behavior attribute value corresponding to each commodity behavior type and the weight associated with the commodity behavior type; the aging factor of the behavior can be calculated according to the time from the occurrence of the behavior to the present and according to a certain function or algorithm, for example, the aging factor is obtained by designing an exponential decay function, and the aging factor can be between 0 and 1.
When the category preference commodity set is determined, the commodity category preferred by the user can be determined according to the preference value corresponding to each historical commodity and the commodity category corresponding to each historical commodity, so that the category preference commodity set corresponding to the commodity category preferred by the user is obtained, for example, the historical commodities are respectively a camera, a doorbell and a sound box, the commodity categories corresponding to the camera and the doorbell are safety protection, the categories corresponding to the sound box are family entertainment, the preference value corresponding to the camera and the preference value corresponding to the doorbell are both greater than the preference value corresponding to the sound box, the commodity category preferred by the user is determined to be safety protection at the moment, the commodities corresponding to the safety protection category form the category preference commodity set, for example, the camera, the doorbell and the alarm form the category preference commodity set.
When the brand preference commodity set is determined, the commodity brand preferred by the user can be determined according to the preference value corresponding to each historical commodity and the commodity brand corresponding to each historical commodity so as to obtain the brand preference commodity set corresponding to the commodity brand preferred by the user, wherein the preference value corresponding to each historical commodity can be used as the preference value of the brand corresponding to the historical commodity, and therefore the brand with a high preference value is obtained and used as the brand preferred by the user.
The historical preferred commodity set can be obtained by the following method: and forming a preference value quantization matrix according to the preference value corresponding to each historical commodity, inputting the preference value quantization matrix into a preset collaborative filtering model, and obtaining a historical preference commodity set according to the output of the preset collaborative filtering model, wherein the preset collaborative filtering model can be a collaborative filtering model based on a user, a collaborative filtering model based on a commodity or a collaborative filtering model based on matrix decomposition.
The popularity commodity set can be obtained by adopting the following modes: the commodity recommending device obtains historical behavior data of all users, determines a preference value corresponding to each preset commodity according to the historical behavior data of all the users, determines the sales volume of each preset commodity in a preset time period, performs normalization processing on the preference value corresponding to each preset commodity and the sales volume corresponding to each preset commodity respectively, and then performs weighted summation to obtain the heat value of each preset commodity, selects the heat commodity according to the heat value, and obtains a heat commodity set.
The step S40 includes:
and step S41, selecting the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user and the candidate commodity set.
After the commodity set and the candidate commodity set corresponding to the historical service preferred by the user are obtained, recommended commodities corresponding to the user are selected from the commodity set and the candidate commodity set corresponding to the historical service preferred by the user, and in order to improve the commodity recommendation accuracy, part of commodities can be selected as recommended commodities.
In this embodiment, a candidate commodity set is determined according to historical commodity behavior data and preset commodity attributes, the candidate commodity set includes at least one of a brand preference commodity set, a category preference commodity set, a popularity preference commodity set and a historical preference commodity set, recommended commodities corresponding to a user are selected according to the commodity set and the candidate commodity set corresponding to the historical service preferred by the user, commodity recommendation is performed from multiple dimensions, and the commodity recommendation accuracy is further improved.
Referring to fig. 4, a third embodiment of the present invention provides a method for recommending a commodity, based on the second embodiment shown in fig. 3, where the step S41 includes:
step S411, determining target commodity information according to a commodity set corresponding to the historical service preferred by the user and the candidate commodity set;
the target commodity information is commodity information of a target commodity, and all or part of the target commodity information can be selected from a commodity set and a candidate commodity set corresponding to a historical service preferred by a user to serve as the target commodity.
In order to improve the accuracy of commodity recommendation, in this embodiment, the click probability of the user on the target commodity is predicted based on the preset gradient lifting tree model and the preset logistic regression model, and the recommended commodity is selected according to the click probability, so that a commodity which is more likely to be clicked by the user is obtained.
When the target commodity is determined, in order to avoid repeatedly recommending the commodities which are already viewed by the user to the user, the recommendation accuracy is improved, historical commodities corresponding to all historical behaviors can be determined according to the historical behavior information, historical commodities are filtered from a commodity set corresponding to the historical service preferred by the user and a candidate commodity set, a target commodity set is obtained, and target commodity information corresponding to each target commodity in the target commodity set is obtained.
Step S412, determining a feature vector corresponding to the user according to the target commodity information, the user attribute information of the user and the historical behavior data, wherein the user attribute information comprises at least one of age, gender, occupation, city, consumption capability and personality preference of the user;
target commodity information such as a name, price, brand, model, color, rating, and commodity associated business situation of the target commodity, and user attribute information including at least one of user age, gender, occupation, city, consumption ability, and character preference.
According to the historical behavior data, extracting user behavior characteristics and characteristics of the user related to commodity attributes, wherein the user behavior characteristics comprise the time from the latest browsing of historical commodities to the present, the times of browsing the historical commodities, whether historical commodities are collected or not and whether the historical commodities are commented or not, the characteristics of the user related to the commodity attributes comprise preference values of the user to categories to which the historical commodities belong and preference values of the user to brands to which the commodities belong, and feature vectors of user attribute information, target commodity information, user behavior characteristics and the characteristics of the user related to the commodity attributes are constructed to serve as feature vectors corresponding to the user.
Step S413, converting the feature vector into a target feature according to a preset gradient lifting tree model;
and inputting the feature vector into a preset gradient lifting tree model, and determining target features according to the output of the preset gradient lifting tree model.
Step S414, determining the click probability of the user to each target commodity according to a preset logistic regression model and the target characteristics;
inputting the target characteristics into a preset logistic regression model, determining the click probability of each target commodity for a user according to the output of the preset logistic regression model, and training the training model according to preset training data by the preset logistic regression model.
Step S415, selecting the recommended commodity from all the target commodities according to the click probability of each target commodity.
The target commodities with the highest click probability can be selected as recommended commodities, or the target commodities with the click probability larger than a preset threshold value are selected as recommended commodities, or the target commodities are sorted according to the click probability of the target commodities, and the target commodities sorted in the front are selected as recommended commodities.
Based on experimental determination, when commodity recommendation is performed based on the embodiment, the purchase conversion rate of the user is improved, so that the accuracy of commodity recommendation can be improved.
In this embodiment, target commodity information is determined through a commodity set and a candidate commodity set corresponding to a historical service preferred by a user, and a feature vector corresponding to the user is determined according to the target commodity information, user attribute information of the user and historical behavior data, wherein the user attribute information includes at least one of user age, gender, occupation, city, consumption capability and character preference; converting the feature vectors into target features according to a preset gradient lifting tree model, determining the click probability of each target commodity of a user according to a preset logistic regression model and the target features, selecting recommended commodities from all the target commodities according to the click probability of each target commodity, predicting the click probability of the target commodities by combining the preset gradient lifting tree model and the preset logistic regression model, and selecting the recommended commodities according to the click probability, so that the commodity recommendation accuracy is improved.
Referring to fig. 5, a fourth embodiment of the present invention provides a method for recommending a commodity, based on the first embodiment shown in fig. 2, where the step S30 includes:
step S31, according to the historical service data, determining a service behavior type corresponding to each historical service and a service behavior attribute value corresponding to the service behavior type, wherein the service behavior type comprises at least one of browsing, clicking, handling, consulting customer service and complaining about the service, the service behavior attribute value comprises at least one of frequency, aging factor and duration of the service behavior, and the aging factor is lower as the time point corresponding to the service behavior type is earlier;
step S32, determining a commodity behavior type and a commodity behavior attribute value of a commodity corresponding to each historical service according to the commodity behavior data corresponding to each historical service;
the execution sequence of step S31 and step S32 is not limited.
Step S33, determining a first preference value corresponding to each historical service according to the service behavior type corresponding to each historical service and the service behavior attribute value;
the first preference value is a preference value of the user for the historical service, and based on a preference degree of the user for the historical service behavior indication for the service, when the first preference value corresponding to each historical service is determined, a product of the service behavior attribute values corresponding to each service behavior type can be calculated to serve as the preference value of the service behavior type, and the preference values of all the service behavior types corresponding to each historical service are summed to obtain the first preference value corresponding to each historical service, or the product of the service behavior attribute values corresponding to each service behavior type and the associated weights can be calculated to serve as the preference value of the service behavior type, and the preference values of all the service behavior types corresponding to each historical service are summed to obtain the first preference value corresponding to each historical service.
Step S34, determining a second preference value of the commodity corresponding to each historical service according to the commodity behavior type of the commodity corresponding to each historical service and the commodity behavior attribute value;
the second preference value is a preference value of a historical commodity corresponding to the historical service, the preference degree of the historical service is indicated based on the behavior of the user on the historical commodity, the preference degree of the user on the historical service is evaluated by the first preference value and the second preference value from different dimensions, wherein the first preference value is obtained directly based on the behavior of the user on the historical service, the second preference value is obtained based on the behavior of the user on the historical commodity and the relation between the historical commodity and the historical service, and the second preference value is mainly based on the principle that the preference of the user on the commodity under a certain historical service is increased, although the user does not directly generate the behavior on the historical service, the preference of the user on the historical service can still be indirectly indicated.
The behavior preference value corresponding to each commodity behavior type can be obtained by multiplying the behavior attribute value corresponding to each commodity behavior type by the weight associated with the commodity behavior type, and the second preference value corresponding to each historical commodity is obtained by summing all the behavior preference values corresponding to each historical commodity.
Step S35, determining a target preference value corresponding to each historical service according to a first preference value corresponding to each historical service and a second preference value of a commodity corresponding to each historical service;
the sum of the first preference value and the second preference value may be used as the target preference value, or the target preference value may be obtained by weighted sum according to the first preference value and its associated weight, the second preference value and its associated weight.
Step S36, determining the historical service preferred by the user according to the target preference value corresponding to each historical service.
The historical services with the target preference values larger than the preset preference values can be selected as the historical services preferred by the user, or the historical services with the preset number with higher target preference values can be selected as the historical services preferred by the user, or the target preference values are sorted from big to small, and the historical services sorted before are selected as the historical services preferred by the user.
In this embodiment, by determining a service behavior type corresponding to each historical service and a service behavior attribute value corresponding to the service behavior type according to the historical service data, where the service behavior type includes at least one of browsing, clicking, handling, consulting customer service and complaints of a service, the service behavior attribute value includes at least one of frequency, aging factor and duration of a service behavior, the earlier the time point corresponding to the service behavior type is, the lower the aging factor is, determining a commodity behavior type and a commodity behavior attribute value of a commodity corresponding to each historical service according to the commodity behavior data corresponding to each historical service, and determining a first preference value corresponding to each historical service according to the service behavior type and the service behavior attribute value corresponding to each historical service, determining a second preference value of the historical commodity corresponding to each historical service according to the commodity behavior type of the historical commodity corresponding to each historical service and the commodity behavior attribute value, determining a target preference value corresponding to each historical service according to the first preference value corresponding to each historical service and the second preference value of the historical commodity corresponding to each historical service, determining the historical service preferred by the user according to the target preference value corresponding to each historical service, determining the historical service preferred by the user according to the first preference value of the historical service and the second preference value of the historical commodity corresponding to the historical service, and recommending the commodity based on the historical service preferred by the user, so that the commodity recommendation accuracy is improved.
Referring to fig. 6, fig. 6 is a schematic block diagram of a product recommendation device according to an embodiment of the present invention, where the product recommendation device includes an obtaining module 10 and a determining module 20, where:
the acquiring module 10 is configured to acquire historical behavior data of a user, where the historical behavior data includes historical business behavior data and historical commodity behavior data;
a determining module 20, configured to determine, according to the historical commodity behavior data, commodity behavior data corresponding to each historical service; determining the historical business preferred by the user according to the historical business data and the commodity behavior data corresponding to each historical business; and determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user.
In an embodiment, the article recommendation device further includes a selection module, wherein:
the determining module 20 is further configured to determine a candidate commodity set according to the historical commodity behavior data and preset commodity attributes, where the candidate commodity set includes at least one of a brand preference commodity set, a category preference commodity set, a hot commodity set, and a historical preference commodity set;
the selection module is used for selecting the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user and the candidate commodity set.
In an embodiment, the determining module 20 is further configured to:
determining a commodity behavior type corresponding to each historical commodity and a behavior attribute value corresponding to the commodity behavior type according to the historical commodity behavior data;
determining a preference value corresponding to each historical commodity according to the commodity behavior type corresponding to each historical commodity and the behavior attribute value corresponding to the commodity behavior type;
and determining the candidate commodity set according to the preference value corresponding to each historical commodity and the preset commodity attribute, wherein the preset commodity attribute comprises at least one of a commodity brand, a commodity category and a commodity heat degree.
In an embodiment, the determining module 20 is further configured to:
the behavior attribute value corresponding to each commodity behavior type and the weight associated with the commodity behavior type are multiplied to obtain a behavior preference value corresponding to each commodity behavior type, the commodity behavior type comprises at least one of commodity link clicking, commodity detail page browsing, commodity video playing, customer service contact, shopping cart adding, order submitting, purchasing and evaluation, the behavior attribute value comprises at least one of behavior frequency, behavior aging coefficient and behavior duration, and the aging coefficient is smaller as the time point of the commodity behavior type is earlier;
and summing all the behavior preference values corresponding to the historical commodities respectively to obtain the preference value corresponding to each historical commodity.
In one implementation, the selecting module is further configured to:
determining target commodity information according to a commodity set corresponding to the historical service preferred by the user and the candidate commodity set;
determining a feature vector corresponding to the user according to the target commodity information, the user attribute information of the user and the historical behavior data, wherein the user attribute information comprises at least one of age, gender, occupation, city, consumption capability and character preference of the user;
converting the feature vector into a target feature according to a preset gradient lifting tree model;
determining the click probability of the user on each target commodity according to a preset logistic regression model and the target characteristics;
and selecting the recommended commodity from all the target commodities according to the click probability of each target commodity.
In an embodiment, the determining module 20 is further configured to:
determining historical commodities corresponding to all historical behaviors according to the historical behavior information;
filtering the historical commodities in a commodity set corresponding to the historical service preferred by the user and the candidate commodity set to obtain a target commodity set;
and acquiring the target commodity information corresponding to each target commodity in the target commodity set.
In an embodiment, the determining module 20 is further configured to:
determining a service behavior type corresponding to each historical service and a service behavior attribute value corresponding to the service behavior type according to the historical service data, wherein the service behavior type comprises at least one of browsing, clicking, handling, consulting customer service and complaints of the service, the service behavior attribute value comprises at least one of frequency, aging coefficient and duration of the service behavior, and the aging coefficient is lower as the time point corresponding to the service behavior type is earlier;
determining the commodity behavior type and the commodity behavior attribute value of the commodity corresponding to each historical service according to the commodity behavior data corresponding to each historical service;
determining a first preference value corresponding to each historical service according to the service behavior type corresponding to each historical service and the service behavior attribute value;
determining a second preference value of the historical commodity corresponding to each historical service according to the commodity behavior type of the historical commodity corresponding to each historical service and the commodity behavior attribute value;
determining a target preference value corresponding to each historical service according to a first preference value corresponding to each historical service and a second preference value of a historical commodity corresponding to each historical service;
and determining the historical services preferred by the user according to the target preference value corresponding to each historical service.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for causing a commodity recommending apparatus (which may be a server) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.

Claims (10)

1. A commodity recommendation method, characterized by comprising:
acquiring historical behavior data of a user, wherein the historical behavior data comprises historical business behavior data and historical commodity behavior data;
determining commodity behavior data corresponding to each historical service according to the historical commodity behavior data;
determining the historical business preferred by the user according to the historical business data and the commodity behavior data corresponding to each historical business;
and determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user.
2. The commodity recommendation method according to claim 1, wherein the step of obtaining the historical behavior data of the user and the step of determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user further comprise:
determining a candidate commodity set according to the historical commodity behavior data and preset commodity attributes, wherein the candidate commodity set comprises at least one of a brand preference commodity set, a category preference commodity set, a hot commodity set and a historical preference commodity set;
the step of determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user comprises the following steps:
and selecting the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user and the candidate commodity set.
3. The commodity recommendation method according to claim 2, wherein the step of determining the candidate commodity set according to the historical commodity behavior data and the preset commodity attributes comprises:
determining a commodity behavior type corresponding to each historical commodity and a behavior attribute value corresponding to the commodity behavior type according to the historical commodity behavior data;
determining a preference value corresponding to each historical commodity according to the commodity behavior type corresponding to each historical commodity and the behavior attribute value corresponding to the commodity behavior type;
and determining the candidate commodity set according to the preference value corresponding to each historical commodity and the preset commodity attribute, wherein the preset commodity attribute comprises at least one of a commodity brand, a commodity category and a commodity heat degree.
4. The item recommendation method according to claim 3, wherein the step of determining the preference value corresponding to each item according to the item behavior type corresponding to each item and the behavior attribute value corresponding to the item behavior type comprises:
the behavior attribute value corresponding to each commodity behavior type and the weight associated with the commodity behavior type are multiplied to obtain a behavior preference value corresponding to each commodity behavior type, the commodity behavior type comprises at least one of commodity link clicking, commodity detail page browsing, commodity video playing, customer service contact, shopping cart adding, order submitting, purchasing and evaluation, the behavior attribute value comprises at least one of behavior frequency, behavior aging coefficient and behavior duration, and the aging coefficient is smaller as the time point of the commodity behavior type is earlier;
and summing all the behavior preference values corresponding to the historical commodities respectively to obtain the preference value corresponding to each historical commodity.
5. The commodity recommendation method according to claim 2, wherein the step of selecting the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user and the candidate commodity set comprises:
determining target commodity information according to a commodity set corresponding to the historical service preferred by the user and the candidate commodity set;
determining a feature vector corresponding to the user according to the target commodity information, the user attribute information of the user and the historical behavior data, wherein the user attribute information comprises at least one of age, gender, occupation, city, consumption capability and character preference of the user;
converting the feature vector into a target feature according to a preset gradient lifting tree model;
determining the click probability of the user on each target commodity according to a preset logistic regression model and the target characteristics;
and selecting the recommended commodity from all the target commodities according to the click probability of each target commodity.
6. The commodity recommendation method according to claim 5, wherein the step of determining target commodity information based on the commodity set corresponding to the historical service preferred by the user and the candidate commodity set comprises:
determining historical commodities corresponding to all historical behaviors according to the historical behavior information;
filtering the historical commodities in a commodity set corresponding to the historical service preferred by the user and the candidate commodity set to obtain a target commodity set;
and acquiring the target commodity information corresponding to each target commodity in the target commodity set.
7. The commodity recommendation method according to claim 1, wherein the step of determining the historical service preferred by the user based on the historical service data and the commodity behavior data corresponding to each historical service comprises:
determining a service behavior type corresponding to each historical service and a service behavior attribute value corresponding to the service behavior type according to the historical service data, wherein the service behavior type comprises at least one of browsing, clicking, handling, consulting customer service and complaints of the service, the service behavior attribute value comprises at least one of frequency, aging coefficient and duration of the service behavior, and the aging coefficient is lower as the time point corresponding to the service behavior type is earlier;
determining the commodity behavior type and the commodity behavior attribute value of the commodity corresponding to each historical service according to the commodity behavior data corresponding to each historical service;
determining a first preference value corresponding to each historical service according to the service behavior type corresponding to each historical service and the service behavior attribute value;
determining a second preference value of the historical commodity corresponding to each historical service according to the commodity behavior type of the historical commodity corresponding to each historical service and the commodity behavior attribute value;
determining a target preference value corresponding to each historical service according to a first preference value corresponding to each historical service and a second preference value of a historical commodity corresponding to each historical service;
and determining the historical services preferred by the user according to the target preference value corresponding to each historical service.
8. The commodity recommending device is characterized by comprising an obtaining module and a determining module, wherein:
the acquisition module is used for acquiring historical behavior data of a user, wherein the historical behavior data comprises historical business behavior data and historical commodity behavior data;
the determining module is used for determining commodity behavior data corresponding to each historical business according to the historical commodity behavior data; determining the historical business preferred by the user according to the historical business data and the commodity behavior data corresponding to each historical business; and determining the recommended commodity corresponding to the user according to the commodity set corresponding to the historical service preferred by the user.
9. An article recommendation device comprising a memory, a processor, and an article recommendation program stored on the memory and executable on the processor, the article recommendation program when executed by the processor implementing the steps of the article recommendation method as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that an article recommendation program is stored thereon, which when executed by a processor implements the steps of the article recommendation method according to any one of claims 1 to 7.
CN202110868342.5A 2021-07-29 2021-07-29 Commodity recommendation method and device and computer-readable storage medium Pending CN113516539A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110868342.5A CN113516539A (en) 2021-07-29 2021-07-29 Commodity recommendation method and device and computer-readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110868342.5A CN113516539A (en) 2021-07-29 2021-07-29 Commodity recommendation method and device and computer-readable storage medium

Publications (1)

Publication Number Publication Date
CN113516539A true CN113516539A (en) 2021-10-19

Family

ID=78067841

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110868342.5A Pending CN113516539A (en) 2021-07-29 2021-07-29 Commodity recommendation method and device and computer-readable storage medium

Country Status (1)

Country Link
CN (1) CN113516539A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115408589A (en) * 2022-08-31 2022-11-29 智城动力(深圳)科技有限公司 Client type matching method and system
CN116089712A (en) * 2022-12-29 2023-05-09 无锡东方健康科技有限公司 Hot conference recommending method and system based on data mining and analysis

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011209979A (en) * 2010-03-30 2011-10-20 Brother Industries Ltd Merchandise recommendation method and merchandise recommendation system
CN105528374A (en) * 2014-10-21 2016-04-27 苏宁云商集团股份有限公司 A commodity recommendation method in electronic commerce and a system using the same
CN106485562A (en) * 2015-09-01 2017-03-08 苏宁云商集团股份有限公司 A kind of commodity information recommendation method based on user's history behavior and system
CN109741146A (en) * 2019-01-04 2019-05-10 平安科技（深圳）有限公司 Products Show method, apparatus, equipment and storage medium based on user behavior
CN110348930A (en) * 2018-04-08 2019-10-18 阿里巴巴集团控股有限公司 Business object data processing method, the recommended method of business object information and device
CN110348920A (en) * 2018-04-02 2019-10-18 中移（杭州）信息技术有限公司 A kind of method and device of recommended products
CN112598467A (en) * 2020-12-23 2021-04-02 北京三快在线科技有限公司 Training method of commodity recommendation model, commodity recommendation method and device
US20210150378A1 (en) * 2018-04-12 2021-05-20 Boe Technology Group Co., Ltd. Recommendation method, recommendation apparatus, recommendation device, recommendation system and storage medium
US20210209491A1 (en) * 2020-09-01 2021-07-08 Beijing Baidu Netcom Science And Technology Co., Ltd. Method and apparatus for recommending content, device, and medium

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011209979A (en) * 2010-03-30 2011-10-20 Brother Industries Ltd Merchandise recommendation method and merchandise recommendation system
CN105528374A (en) * 2014-10-21 2016-04-27 苏宁云商集团股份有限公司 A commodity recommendation method in electronic commerce and a system using the same
CN106485562A (en) * 2015-09-01 2017-03-08 苏宁云商集团股份有限公司 A kind of commodity information recommendation method based on user's history behavior and system
CN110348920A (en) * 2018-04-02 2019-10-18 中移（杭州）信息技术有限公司 A kind of method and device of recommended products
CN110348930A (en) * 2018-04-08 2019-10-18 阿里巴巴集团控股有限公司 Business object data processing method, the recommended method of business object information and device
US20210150378A1 (en) * 2018-04-12 2021-05-20 Boe Technology Group Co., Ltd. Recommendation method, recommendation apparatus, recommendation device, recommendation system and storage medium
CN109741146A (en) * 2019-01-04 2019-05-10 平安科技（深圳）有限公司 Products Show method, apparatus, equipment and storage medium based on user behavior
US20210209491A1 (en) * 2020-09-01 2021-07-08 Beijing Baidu Netcom Science And Technology Co., Ltd. Method and apparatus for recommending content, device, and medium
CN112598467A (en) * 2020-12-23 2021-04-02 北京三快在线科技有限公司 Training method of commodity recommendation model, commodity recommendation method and device

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
CHUNHUA JU ET AL: "The commodity recommendation method for online shopping based on data mining", MULTIMEDIA TOOLS AND APPLICATIONS, vol. 78, pages 30097 - 30110, XP037054274, DOI: 10.1007/s11042-018-6980-7 *
郭丽 等: "基于情感分析的商品推荐***的设计与实现", 中原工学院学报, vol. 25, no. 3, pages 71 - 74 *
麻风梅 等: "基于用户社区的商品推荐方法", 计算机与数字工程, vol. 41, no. 8, pages 1354 - 1356 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115408589A (en) * 2022-08-31 2022-11-29 智城动力(深圳)科技有限公司 Client type matching method and system
CN116089712A (en) * 2022-12-29 2023-05-09 无锡东方健康科技有限公司 Hot conference recommending method and system based on data mining and analysis
CN116089712B (en) * 2022-12-29 2024-03-29 无锡东方健康科技有限公司 Hot conference recommending method and system based on data mining and analysis

Similar Documents

Publication Publication Date Title
CN112182412B (en) Method, computing device, and computer storage medium for recommending physical examination items
CN111626832B (en) Product recommendation method and device and computer equipment
CN110473040B (en) Product recommendation method and device and electronic equipment
CN111523976A (en) Commodity recommendation method and device, electronic equipment and storage medium
CN109903103B (en) Method and device for recommending articles
WO2021139524A1 (en) Method and apparatus for processing interaction data by using lstm neural network model
CN110598847A (en) Method and device for processing interactive sequence data
CN103118111A (en) Information push method based on data from a plurality of data interaction centers
CN113516539A (en) Commodity recommendation method and device and computer-readable storage medium
CN111008335B (en) Information processing method, device, equipment and storage medium
CN112015986B (en) Data pushing method, device, electronic equipment and computer readable storage medium
CN111914188A (en) Method, system, device and storage medium for selecting recommendation target user
CN111461827B (en) Push method and device for product evaluation information
CN113780479A (en) Periodic prediction model training method and device, and periodic prediction method and equipment
CN115564531A (en) Commodity combination recommendation method and device, equipment, medium and product thereof
CN112036987B (en) Method and device for determining recommended commodity
CN113761352A (en) Information pushing method, device, equipment and storage medium
CN113763119A (en) Payment method, payment device, electronic equipment and computer-readable storage medium
CN115587875B (en) Textile e-commerce recommendation method and device based on balanced perception attention network
CN115719244A (en) User behavior prediction method and device
CN113158032B (en) Information pushing method and device
CN114841781A (en) Commodity information recommendation method, device, equipment and medium
CN115392943A (en) Method, device, equipment and storage medium for sending display information
CN112307319A (en) Page generation method and device
CN116911913B (en) Method and device for predicting interaction result

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
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20211019

WD01 Invention patent application deemed withdrawn after publication