WO2020020137A1 - 商品推荐方法、装置、系统及计算机可读存储介质 - Google Patents

商品推荐方法、装置、系统及计算机可读存储介质 Download PDF

Info

Publication number
WO2020020137A1
WO2020020137A1 PCT/CN2019/097244 CN2019097244W WO2020020137A1 WO 2020020137 A1 WO2020020137 A1 WO 2020020137A1 CN 2019097244 W CN2019097244 W CN 2019097244W WO 2020020137 A1 WO2020020137 A1 WO 2020020137A1
Authority
WO
WIPO (PCT)
Prior art keywords
product
recommended
products
recommendation
specific offline
Prior art date
Application number
PCT/CN2019/097244
Other languages
English (en)
French (fr)
Inventor
石海林
梅涛
周伯文
赵何
龚书
Original Assignee
北京京东尚科信息技术有限公司
北京京东世纪贸易有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京京东尚科信息技术有限公司, 北京京东世纪贸易有限公司 filed Critical 北京京东尚科信息技术有限公司
Priority to EP19840851.0A priority Critical patent/EP3806021A4/en
Priority to US17/258,349 priority patent/US20210295414A1/en
Publication of WO2020020137A1 publication Critical patent/WO2020020137A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0251—Targeted advertisements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0251—Targeted advertisements
    • G06Q30/0269—Targeted advertisements based on user profile or attribute
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L13/00—Speech synthesis; Text to speech systems
    • G10L13/02—Methods for producing synthetic speech; Speech synthesisers

Definitions

  • the present disclosure relates to the field of computer technology, and in particular, to a method, device, system, and computer-readable storage medium for product recommendation.
  • shoppers mainly recommend products to customers by shopping guides, or circulate advertisements on billboards to attract customers.
  • One technical problem to be solved by the embodiments of the present disclosure is to improve the accuracy of recommending products to specific offline customers.
  • a product recommendation method including: analyzing an image of a specific offline customer currently entering a store to obtain attributes of the specific offline customer; and according to the specific offline customer Determine the online users that match the specific offline customer; construct the popular product set of the store based on the products corresponding to the residence location of the offline customers who previously entered the store; and according to the online The user's historical shopping information and the popular product collection recommend products to the specific offline customer.
  • recommending products to the specific offline customer according to the historical shopping information of the online user and the popular product set includes: determining a plurality of to be recommended according to the historical shopping information of the online user. Products; and determining the recommendation degree of each to-be-recommended product according to the product attention degree and recommendation weight of each to-be-recommended product, wherein the recommendation right of the to-be-recommended product belonging to the popular product set is greater than that not belonging to the popular product set The recommendation weight of the product to be recommended; recommending the plurality of products to be recommended to the specific offline customer in order of the degree of recommendation.
  • constructing the popular product set includes: generating a heat map of the offline customers based on the resident locations of the offline customers who previously entered the store; and according to the heat map of the offline customers and The arrangement position corresponding to the products, to obtain the popular product set.
  • the method further includes: obtaining at least one of a preference item of the specific offline customer and historical shopping information of the specific offline customer in the store; according to the history of the online user
  • the shopping information and the popular product set, recommending the product to the specific offline customer include: according to at least one of the specific offline customer's preference product and the specific offline customer's historical shopping information in the shop, The historical shopping information of the online user and the set of popular products recommend products to the specific offline customer.
  • Recommending products to the specific offline customer includes: determining a plurality of first to-be-recommended products according to the historical shopping information of the online user; and according to the preference products of the specific offline customer and the specific offline customer's Determining at least one of the historical shopping information of the shop, a plurality of second to-be-recommended products; and constructing a set of to-be-recommended products based on the plurality of first to-be-recommended products and the plurality of second to-be-recommended products, the The set of products to be recommended includes a plurality of products to be recommended, and the plurality of products to be recommended includes at least one of the plurality of first products to be recommended and at least one of the plurality of second products to be recommended; The product attention degree and recommendation weight of the product to be recommended, and the recommendation degree of each product
  • the preference product of the specific offline customer is determined according to the following manner: according to the previous resident location of the specific offline customer in the shop, generating a heat map of the specific offline customer; The heat map of the specific offline customer and the corresponding placement position of the product, calculating the residence time of the specific offline customer at the placement position of different products; and determining the product corresponding to the placement position where the residence time is greater than a preset time It is a preference product of the specific offline customer.
  • recommending the plurality of products to be recommended to the specific offline customer includes: generating a match between the names of the specific offline customers and the names of the plurality of products to be recommended through natural language generation technology.
  • a grammatical rule sentence ; and converting the sentence into a voice and sending it to a shopping guide of the shop so that the shopping guide recommends the plurality of products to be recommended to the specific offline customer according to the voice.
  • recommending the plurality of products to be recommended to the specific offline customer includes: obtaining pictures corresponding to each product to be recommended; and generating images of each product to be recommended according to the images corresponding to each product to be recommended. Descriptive information; and in accordance with the order of recommendation from large to small, pictures and description information of each to-be-recommended product are sequentially output to a display screen of the shop for display.
  • the product attention degree of each to-be-recommended product is based on at least one of the number of visits to the to-be-recommended product, feedback information of online users, the degree of habit matching with the specific offline customer, and the cost performance of the product. The higher the number of visits, the better the feedback from online users, the higher the degree of matching with the habits of the particular offline customer, or the higher the cost-effectiveness of the product, the higher the product attention of the product to be recommended.
  • a product recommendation device including: an analysis module configured to analyze images of a specific offline customer currently entering a store to obtain attributes of the specific offline customer; A matching module configured to determine an online user matched with the specific offline customer according to the attributes of the specific offline customer; a building module configured to be based on a residence location of the offline customer who previously entered the shop The corresponding product constructs a popular product set of the shop; and a recommendation module is configured to recommend the product to the specific offline customer according to the historical shopping information of the online user and the popular product set.
  • the recommendation module is configured to: determine a plurality of products to be recommended according to the historical shopping information of the online user; and determine each of the products to be recommended according to the product attention and recommendation weight of each product to be recommended.
  • the degree of recommendation of the recommended products wherein the recommendation weight of the products to be recommended belonging to the popular product set is greater than the recommendation weight of the products to be recommended not belonging to the popular product set; and
  • the specific offline customer recommends the plurality of products to be recommended.
  • the building module is configured to: generate a heat map of the offline customers based on the resident location of the offline customers who previously entered the shop; and according to the heat map of the offline customers and The arrangement position corresponding to the products, to obtain the popular product set.
  • the device further includes: an obtaining module configured to obtain at least one of a preference product of the specific offline customer and historical shopping information of the specific offline customer in the store; the The recommendation module is configured to be based on at least one of the specific offline customer's preferred products and the specific offline customer's historical shopping information in the store, the online user's historical shopping information, and the popular product set. To recommend a product to the specific offline customer.
  • the recommendation module is configured to: determine a plurality of first products to be recommended according to the historical shopping information of the online user; and according to the preference products of the specific offline customer and the specific offline The customer determines at least one of the historical shopping information of the store a plurality of second to-be-recommended products; and constructs a set of to-be-recommended products based on the plurality of first to-be-recommended products and the plurality of second to-be-recommended products,
  • the set of products to be recommended includes a plurality of products to be recommended, the plurality of products to be recommended includes at least one of the plurality of first products to be recommended and at least one of the plurality of second products to be recommended;
  • Product attention degree and recommendation weight of each product to be recommended determine the recommendation degree of each product to be recommended, wherein the recommendation right of the product to be recommended belonging to the popular product set is greater than the product to be recommended not belonging to the popular product set
  • the preference product of the specific offline customer is determined according to the following manner: according to the previous resident location of the specific offline customer in the shop, generating a heat map of the specific offline customer; The heat map of the specific offline customer and the corresponding placement position of the product, calculating the residence time of the specific offline customer at the placement position of different products; and determining the product corresponding to the placement position where the residence time is greater than a preset time It is a preference product of the specific offline customer.
  • the recommendation module is configured to generate, by using natural language generation technology, the names of the specific offline customers and the names of the plurality of products to be recommended to comply with grammatical rules; and The sentence is converted into voice and sent to the shopping guide of the shop, so that the shopping guide recommends the plurality of products to be recommended to the specific offline customer according to the voice.
  • the recommendation module is configured to: obtain pictures corresponding to each product to be recommended; generate description information of each product to be recommended according to the image corresponding to each product to be recommended; In small order, pictures and description information of each product to be recommended are sequentially output to a display screen of the shop for display.
  • the product attention degree of each to-be-recommended product is based on at least one of the number of visits to the to-be-recommended product, feedback information of online users, the degree of habit matching with the specific offline customer, and the cost-effectiveness of the product. The higher the number of visits, the better the feedback from online users, the higher the degree of matching with the habits of the particular offline customer, or the higher the cost-effectiveness of the product, the higher the product attention of the product to be recommended.
  • a product recommendation device including: a memory; and a processor coupled to the memory, the processor being configured to execute any of the above based on instructions stored in the memory. The method described in one embodiment.
  • a computer-readable storage medium having computer program instructions stored thereon, which are executed by a processor to implement a method according to any one of the above embodiments.
  • a product recommendation system including: the product recommendation device described in any one of the above embodiments; a camera configured to collect images of offline customers who are currently entering a store, and the collected images Input to the product recommendation device.
  • FIG. 1 is a schematic flowchart illustrating a method for recommending products according to some embodiments of the present disclosure
  • FIG. 2 is a schematic flowchart illustrating a method for recommending products according to other embodiments of the present disclosure
  • FIG. 3 is a schematic flowchart illustrating a process of recommending a product to a specific offline customer according to some implementations of the present disclosure
  • FIG. 4 is a schematic structural diagram illustrating a product recommendation device according to some embodiments of the present disclosure.
  • FIG. 5 is a schematic structural diagram illustrating a product recommendation device according to other embodiments of the present disclosure.
  • FIG. 6 is a schematic structural diagram illustrating a product recommendation device according to some embodiments of the present disclosure.
  • FIG. 7 is a schematic structural diagram illustrating a product recommendation system according to some embodiments of the present disclosure.
  • a specific component when it is described that a specific component is located between the first component and the second component, there may or may not be an intermediate component between the specific component and the first component or the second component.
  • the specific component When it is described that a specific component is connected to other components, the specific component may be directly connected to the other components without an intervening component, or may be directly connected to the other components without an intervening component.
  • FIG. 1 is a schematic flowchart of a method for recommending products according to some embodiments of the present disclosure.
  • step 102 an image of a specific offline customer currently entering a set area (eg, a store) is analyzed to obtain attributes of the specific offline customer.
  • a set area eg, a store
  • a camera can be installed in a store to capture images of offline customers entering the store. Face detection and recognition technology is used to process images of offline customers, which can identify specific offline customers, such as VIP customers.
  • the face attribute analysis technology can be used to analyze the images of the specific offline customers to obtain the attributes of the specific offline customers.
  • the information of the detected face area can be input into a deep learning model to obtain the attributes of a specific offline customer.
  • the attributes of a specific offline customer may include, but are not limited to, at least one of the following: gender, age, expression, race, face value, whether to wear glasses, whether to wear sunglasses, whether to have a beard, and the like.
  • the attributes of a particular offline customer can be represented as an attribute vector, and each element in the attribute vector represents a certain attribute.
  • the vector (1,2,0,1,1,1) indicates that the age of a particular offline customer is 20-30, Caucasian, female, without masks, glasses, and sunglasses.
  • step 104 an online user matching the specific offline customer is determined according to the attributes of the specific offline customer.
  • an online user may be any user who consumes through an Internet platform.
  • the attributes of an online user can be obtained from an online database, and then the attributes of a specific offline customer are matched with the attributes of an online user (such as a comparison of attribute vectors) to determine that the attributes of the specific offline customer are the same or similar Of online users.
  • An online user who has the same or similar attributes as a specific offline customer can be regarded as an online user that matches a specific offline customer.
  • the attributes of the online user may be determined using a method similar to the method for determining the attributes of the specific offline customer described above.
  • some online users can be selected for subsequent product recommendation according to the ownership of specific offline customers. This can reduce the amount of data processing and improve the efficiency of product recommendation.
  • the home location of the online user may be determined according to the registered address of the online user on the Internet platform, a commonly used shipping address, or a network IP address.
  • a popular product set of the shop is constructed according to an object (for example, a product) corresponding to the residence location of the offline customer who has previously entered the shop.
  • a heat map for offline customers may be generated based on where the offline customers entered the store before. Then, according to the heat map of the offline customers and the corresponding arrangement position of the products, a set of popular products is obtained.
  • the heat map of offline customers can reflect the distribution of offline customers' locations in the shops in the past period. According to the distribution of offline customers' resident locations in the shop, it can be obtained which offline customers are more interested in, and then a set of popular products can be obtained.
  • step 108 according to the historical shopping information of the online user and the popular product set, a product is recommended to a specific offline customer.
  • the following describes an implementation method for recommending products to a specific offline customer.
  • a plurality of products to be recommended are determined based on historical shopping information of online users matched with a specific offline customer.
  • the historical shopping information may include products purchased by online users, favorite products, and the like.
  • the recommendation degree of each product to be recommended is determined.
  • the recommendation weight of the product to be recommended that belongs to the popular product set is greater than the recommendation weight of the product to be recommended that does not belong to the popular product set.
  • the product of the product attention degree and the recommendation weight of each product to be recommended may be calculated to obtain the recommendation degree of each product to be recommended.
  • the sum S of the product of the product attention degree and the recommendation weight of each product to be recommended may be calculated, and then each The ratio of the products s and S is used as the recommendation degree of each product to be recommended.
  • the product attention degree of each product to be recommended may be determined according to at least one of the following: the number of visits of the product to be recommended, feedback information from online users, the degree of matching with the habits of a specific offline customer, and the cost-effectiveness of the product.
  • the initial value of the recommendation weight of each product to be recommended may be the same. If a certain product to be recommended belongs to a popular product set, the recommendation weight of the product to be recommended is increased.
  • a plurality of products to be recommended to a specific offline customer are recommended, that is, the products to be recommended with a greater recommendation are preferentially recommended.
  • only the products to be recommended with a recommendation degree greater than the preset recommendation degree among the plurality of products to be recommended may be recommended.
  • multiple offline products can be recommended to specific offline customers according to the following methods: First, by using natural language generation technology, the names of specific offline customers and the names of multiple products to be recommended are generated in accordance with grammatical rules Sentence; then, the sentence is converted into voice and sent to the shop's shopping guide, so that the shopping guide recommends a number of products to be recommended to specific offline customers according to the voice, so as to provide customized services for specific offline customers. For example, Zhang San enters the store and recommends to him gray sweaters, white hats, blue jeans, and so on.
  • multiple offline products can be recommended to offline customers in the following manner: First, a picture corresponding to each of the recommended products is obtained. Then, according to a picture corresponding to each product to be recommended, description information of each product to be recommended is generated. For example, a picture corresponding to a product to be recommended may be input to an image caption model, and the model may output information describing the input picture, so as to obtain description information of each product to be recommended. After that, in accordance with the recommendation degree, the pictures and description information of each product to be recommended are sequentially output to the store's display screen for display, so as to attract the attention of specific offline customers. For example, a product with a higher recommendation level is preferentially displayed on a display screen. For example, gray sweaters, white hats, blue jeans, etc. are displayed on the display screen in turn, with corresponding text descriptions.
  • the historical shopping information and popular product set of the online user matching the specific offline customer are comprehensively considered.
  • Such a recommendation method is targeted to specific offline customers, and can improve the accuracy of recommending products to specific offline customers.
  • specific offline customers can understand the current trend and provide a more comfortable shopping experience for specific offline customers.
  • FIG. 2 is a schematic flowchart illustrating a method for recommending products according to other embodiments of the present disclosure.
  • the differences between FIG. 2 and FIG. 1 are mainly introduced below. For other related points, reference may be made to the above description, and details are not described herein again.
  • step 202 the image of the specific offline customer currently entering the store is analyzed to obtain the attributes of the specific offline customer.
  • step 204 an online user matching the specific offline customer is determined according to the attributes of the specific offline customer.
  • step 206 at least one of a preference item of the specific offline customer and historical shopping information of the specific offline customer in the store is acquired.
  • the preference product of a particular offline customer can be determined as follows.
  • a heat map for a specific offline customer is generated based on the previous resident location of the specific offline customer in the store.
  • the heat map of the specific offline customer can reflect the distribution of the previous offline customers' resident locations in the store. According to the distribution of the resident locations of the specific offline customers in the stores, it can be obtained which products in the stores are more interested by the specific offline customers.
  • the residence time of the specific offline customer at the placement position of the different product is calculated.
  • the product corresponding to the placement position where the dwell time is longer than the preset time is determined as the preferred product of the specific offline customer.
  • a popular product set of the shop is constructed according to the products corresponding to the resident locations of the offline customers who previously entered the shop.
  • a product is recommended to a specific offline customer according to at least one of a specific offline customer's preference product and a specific offline customer's historical shopping information in the store, the online user's historical shopping information, and a popular product set.
  • the preference product of a specific offline customer may also be the purchase history of the specific offline customer in the store, that is, the product that has been purchased.
  • At least one of the preference product of the specific offline customer and the historical shopping information of the specific offline customer in the store are also considered.
  • Such a recommendation method can recommend more interesting products to the particular offline customer, further improving the accuracy of the recommendation.
  • FIG. 3 is a schematic diagram illustrating a process of recommending a product to a specific offline customer according to some implementations of the present disclosure.
  • step 302 a plurality of first to-be-recommended products are determined according to the historical shopping information of an online user matched with a specific offline customer.
  • a product that an online user once purchased or favorited may be determined as the first product to be recommended.
  • step 304 a plurality of second to-be-recommended products are determined according to at least one of the preference products of the specific offline customer and the historical shopping information of the specific offline customer in the store.
  • a specific offline customer's preferred product or a product similar to the preferred product may be used as the second to-be-recommended product, or a specific offline customer may purchase a product in the store ’s history or a product similar to the historical purchase product as the second product to be recommended
  • the to-be-recommended product, or a second-to-be-recommended product may be a preferred product of a specific offline customer, a product similar to the preferred product, a historical offline purchase product of the specific offline customer and a product similar to the historical purchase product.
  • a set of products to be recommended is constructed according to a plurality of first products to be recommended and a plurality of second products to be recommended, and the set of products to be recommended includes a plurality of products to be recommended.
  • the plurality of products to be recommended in the constructed set of products to be recommended may include some or all of the first products to be recommended determined in step 302 and some or all of the second products to be recommended determined in step 304.
  • the same first product to be recommended and the second product to be recommended may be combined into one product to be recommended.
  • the recommendation degree of each product to be recommended is determined according to the product attention degree and recommendation weight of each product to be recommended.
  • the recommendation weight of the product to be recommended that belongs to the popular product set is greater than the recommendation weight of the product to be recommended that does not belong to the popular product set.
  • step 310 a plurality of products to be recommended are recommended to a specific offline customer in the order of the degree of recommendation.
  • a plurality of products to be recommended can be recommended to a specific offline customer in the manner given above.
  • FIG. 4 is a schematic structural diagram illustrating a product recommendation device according to some embodiments of the present disclosure.
  • the product recommendation device 400 of this embodiment includes an analysis module 401, a matching module 402, a construction module 403, and a recommendation module 404.
  • the analysis module 401 is configured to analyze an image of a specific offline customer currently entering a store to obtain attributes of the specific offline customer.
  • the matching module 402 is configured to determine an online user matching a specific offline customer according to the attributes of the specific offline customer.
  • the building module 403 is configured to build a popular product set of a store according to the products corresponding to the resident locations of offline customers who have previously entered the store.
  • the building module 403 is configured to generate a heat map of the offline customers based on the resident locations of the offline customers who have previously entered the store; based on the heat map of the offline customers and the corresponding arrangement positions of the products, obtain popular products set.
  • the recommendation module 404 is configured to recommend products to a specific offline customer based on the historical shopping information of the online user and a set of popular products.
  • the historical shopping information and popular product set of the online user matching the specific offline customer are comprehensively considered.
  • Such a recommendation method is targeted to specific offline customers, and can improve the accuracy of recommending products to specific offline customers.
  • specific offline customers can understand the current trend and provide a more comfortable shopping experience for specific offline customers.
  • the recommendation module 404 is configured to recommend products to specific offline customers in the following manner: determining a plurality of products to be recommended based on the historical shopping information of the online user; according to the product attention degree and The recommendation weight determines the recommendation degree of each product to be recommended. Among them, the recommendation weight of a product to be recommended that belongs to a popular product set is greater than the recommendation weight of a product to be recommended that does not belong to a popular product set. Recommend multiple products to be recommended to specific offline customers.
  • the product attention degree of each product to be recommended may be determined according to at least one of the number of visits to the product to be recommended, feedback information from online users, the degree of habit matching with a specific offline customer, and the cost performance of the product. Here, the higher the number of visits, the better the feedback from online users, the higher the degree of matching with the habits of a particular offline customer, or the higher the cost-effectiveness of the product, the higher the product attention of the product to be recommended.
  • FIG. 5 is a schematic structural diagram illustrating a product recommendation device according to other embodiments of the present disclosure.
  • the product recommendation device 500 of this embodiment further includes an obtaining module 501 compared with the product recommendation device 400 shown in FIG. 4.
  • the acquisition module 501 is configured to acquire at least one of a preference item of a specific offline customer and historical shopping information of a specific offline customer in a store.
  • the recommendation module 404 in this embodiment is configured to be based on at least one of a specific offline customer's preference product and a specific offline customer's historical shopping information in the store, the online shopping history shopping information, and the popular product set, Recommend products to specific offline customers.
  • the recommendation module 404 is configured to recommend products to a specific offline customer according to the following manner: determining a plurality of first to-be-recommended products based on the historical shopping information of the online user; according to the preference products of the specific offline customer and the specific offline customer Determine at least one of the historical shopping information of the store to determine a plurality of second to-be-recommended products; and construct a set of to-be-recommended products based on the plurality of first to-be-recommended products and the plurality of second to-be-recommended products Products to be recommended, the plurality of products to be recommended includes at least one of the plurality of first products to be recommended and at least one of the plurality of second products to be recommended; The recommendation degree of each product to be recommended. Among them, the recommendation weight of the product to be recommended belonging to the popular product set is greater than the recommendation weight of the product to be recommended not belonging to the popular product set. Recommend multiple products to be recommended.
  • the specific offline customer's preferred products can be determined according to the following methods: generating a specific offline customer's heat map according to the previous offline location of the specific offline customer in the store; according to the specific offline customer's heat map The placement position corresponding to the product, calculates the residence time of the specific offline customer at the placement position of the different products; determines the product corresponding to the placement position where the residence time is longer than a preset time as the preference product of the specific offline customer.
  • the preference products of the specific offline customers may be stored in the preference product database. For example, different logos can be used to distinguish different offline customers.
  • the recommendation module 404 is configured to use natural language generation technology to generate grammatical rules for the names of specific offline customers and the names of multiple products to be recommended; convert the sentences into speech and send it to the store Shopping guide, so that the shopping guide can recommend multiple products to be recommended to specific offline customers based on their voice.
  • the recommendation module 404 is configured to obtain pictures corresponding to each product to be recommended; generate description information of each product to be recommended according to the image corresponding to each product to be recommended; Order, the pictures and description information of each product to be recommended are sequentially output to the store's display screen for display.
  • FIG. 6 is a schematic structural diagram illustrating a product recommendation device according to some embodiments of the present disclosure.
  • the product recommendation device 600 of this embodiment includes a memory 601 and a processor 602 coupled to the memory 601.
  • the processor 602 is configured to execute any one of the foregoing embodiments based on instructions stored in the memory 601.
  • the memory 601 may include, for example, a system memory, a fixed non-volatile storage medium, and the like.
  • the system memory may store, for example, an operating system, an application program, a boot loader (Boot Loader), and other programs.
  • the product recommendation device 600 may further include an input-output interface 603, a network interface 604, a storage interface 605, and the like. These interfaces 603, 604, and 605, and the memory 601 and the processor 602 may be connected through a bus 606, for example.
  • the input / output interface 603 provides a connection interface for input / output devices such as a display, a mouse, a keyboard, and a touch screen.
  • the network interface 604 provides a connection interface for various networked devices.
  • the storage interface 605 provides a connection interface for external storage devices such as an SD card and a U disk.
  • FIG. 7 is a schematic structural diagram illustrating a product recommendation system according to some embodiments of the present disclosure. As shown in FIG. 7, the product recommendation system of this embodiment includes the product recommendation device 400/500/600 and the camera 701 of any one of the foregoing embodiments.
  • the camera 701 is configured to capture images of offline customers currently entering the store, and input the captured images to the product recommendation device 400/500/600.
  • the product recommendation devices 400/500/600 can process images of offline customers in the manner given above, and then recommend products to specific offline customers.
  • the product recommendation system may further include an online database 702.
  • the online database 702 is configured to store attributes and historical shopping information of online users.
  • the product recommendation system may further include a display screen 703.
  • the display screen 703 is configured to display product information recommended to a specific offline customer, such as a picture and description information of a product to be recommended.
  • An embodiment of the present disclosure further provides a computer-readable storage medium having computer program instructions stored thereon, which are executed by a processor to implement the method of any one of the foregoing embodiments.
  • the embodiments of the present disclosure may be provided as a method, a system, or a computer program product. Therefore, the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present disclosure may take the form of a computer program product implemented on one or more computer-usable non-transitory storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code therein .
  • computer-usable non-transitory storage media including but not limited to disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to work in a particular manner such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, the instructions
  • the device implements the functions specified in one or more flowcharts and / or one or more blocks of the block diagram.
  • These computer program instructions can also be loaded on a computer or other programmable data processing device, so that a series of steps can be performed on the computer or other programmable device to produce a computer-implemented process, which can be executed on the computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more flowcharts and / or one or more blocks of the block diagrams.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Acoustics & Sound (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

一种商品推荐方法、装置、系统及计算机可读存储介质，涉及计算机技术领域，所述方法包括：对当前进入商铺的特定线下顾客的图像进行分析，以得到所述特定线下顾客的属性（102）；根据所述特定线下顾客的属性，确定与所述特定线下顾客匹配的线上用户（104）；根据以前进入所述商铺的线下顾客的驻留位置对应的商品，构建所述商铺的热门商品集合（106）；和根据所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品（108）。

Description

商品推荐方法、装置、系统及计算机可读存储介质
相关申请的交叉引用
本申请是以CN申请号为201810822977.X，申请日为2018年7月25日的申请为基础，并主张其优先权，该CN申请的公开内容在此作为整体引入本申请中。
技术领域
本公开涉及计算机技术领域，尤其是一种商品推荐方法、装置、系统及计算机可读存储介质。
背景技术
随着互联网技术的不断发展，线上消费行为越来越普及。但是，线上消费始终无法完全取代线下消费，尤其是一些奢侈品消费，人们更倾向在线下的商铺中进行。
目前，传统的商品推荐方式中，商铺主要由导购员为顾客推荐商品，或者通过广告牌循环播放广告来吸引顾客。
发明内容
发明人发现：传统的商品推荐方式中，为顾客推荐什么样的商品是由人工决定的，这无法向顾客准确地推荐商品。
本公开实施例所要解决的一个技术问题是：提高向特定线下顾客推荐商品的准确性。
根据本公开实施例的一方面，提供一种商品推荐方法，包括：对当前进入商铺的特定线下顾客的图像进行分析，以得到所述特定线下顾客的属性；根据所述特定线下顾客的属性，确定与所述特定线下顾客匹配的线上用户；根据以前进入所述商铺的线下顾客的驻留位置对应的商品，构建所述商铺的热门商品集合；和根据所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品。
在一些实施例中，根据所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品包括：根据所述线上用户的历史购物信息，确定多个待推荐商品；和根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于所述热门商品集合的待推荐商品的推荐权重大于不属于所述热门商品 集合的待推荐商品的推荐权重；按照推荐度从大到小的顺序，向所述特定线下顾客推荐所述多个待推荐商品。
在一些实施例中，构建所述热门商品集合包括：根据以前进入所述商铺的线下顾客的驻留位置，生成所述线下顾客的热力图；和根据所述线下顾客的热力图和商品对应的布置位置，得到所述热门商品集合。
在一些实施例中，所述方法还包括：获取所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个；根据所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品包括：根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个、所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品。
在一些实施例中，根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息的至少一个、所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品包括：根据所述线上用户的历史购物信息，确定多个第一待推荐商品；根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个，确定多个第二待推荐商品；根据所述多个第一待推荐商品和所述多个第二待推荐商品，构建待推荐商品集合，所述待推荐商品集合包括多个待推荐商品，所述多个待推荐商品包括所述多个第一待推荐商品中的至少一个和所述多个第二待推荐商品中的至少一个；根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于所述热门商品集合的待推荐商品的推荐权重大于不属于所述热门商品集合的待推荐商品的推荐权重；和按照推荐度从大到小的顺序，向所述特定线下顾客推荐所述多个待推荐商品。
在一些实施例中，所述特定线下顾客的偏好商品根据如下方式来确定：根据所述特定线下顾客以前在所述商铺的驻留位置，生成所述特定线下顾客的热力图；根据所述特定线下顾客的热力图和商品对应的布置位置，计算所述特定线下顾客在不同商品的布置位置的驻留时间；和将驻留时间大于预设时间的布置位置对应的商品确定为所述特定线下顾客的偏好商品。
在一些实施例中，向所述特定线下顾客推荐所述多个待推荐商品包括：通过自然语言生成技术，将所述特定线下顾客的姓名和所述多个待推荐商品的名称生成符合语法规则的语句；和将所述语句转换成语音，并发送给所述商铺的导购员，以便所述导购员根据所述语音向所述特定线下顾客推荐所述多个待推荐商品。
在一些实施例中，向所述特定线下顾客推荐所述多个待推荐商品包括：获取每个待推荐商品对应的图片；根据每个待推荐商品对应的图片，生成每个待推荐商品的描述信息；和按照推荐度从大到小的顺序，将每个待推荐商品的图片和描述信息依次输出到所述商铺的显示屏进行显示。
在一些实施例中，每个待推荐商品的商品关注度根据该待推荐商品的访问次数、线上用户的反馈信息、与所述特定线下顾客的习惯匹配度、商品性价比中的至少一项来确定；其中，访问次数越高、线上用户的反馈越好、与所述特定线下顾客的习惯匹配度越高或商品性价比越高，该待推荐商品的商品关注度越高。
根据本公开实施例的另一方面，提供一种商品推荐装置，包括：分析模块，被配置为对当前进入商铺的特定线下顾客的图像进行分析，以得到所述特定线下顾客的属性；匹配模块，被配置为根据所述特定线下顾客的属性，确定与所述特定线下顾客匹配的线上用户；构建模块，被配置为根据以前进入所述商铺的线下顾客的驻留位置对应的商品，构建所述商铺的热门商品集合；和推荐模块，被配置为根据所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品。
在一些实施例中，所述推荐模块被配置为：根据所述线上用户的历史购物信息，确定多个待推荐商品；根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于所述热门商品集合的待推荐商品的推荐权重大于不属于所述热门商品集合的待推荐商品的推荐权重；和按照推荐度从大到小的顺序，向所述特定线下顾客推荐所述多个待推荐商品。
在一些实施例中，所述构建模块被配置为：根据以前进入所述商铺的线下顾客的驻留位置，生成所述线下顾客的热力图；和根据所述线下顾客的热力图和商品对应的布置位置，得到所述热门商品集合。
在一些实施例中，所述装置还包括：获取模块，被配置为获取所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个；所述推荐模块被配置为根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个、所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品。
在一些实施例中，所述推荐模块被配置为：根据所述线上用户的历史购物信息，确定多个第一待推荐商品；根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个，确定多个第二待推荐商品；根据所述多个第 一待推荐商品和所述多个第二待推荐商品，构建待推荐商品集合，所述待推荐商品集合包括多个待推荐商品，所述多个待推荐商品包括所述多个第一待推荐商品中的至少一个和所述多个第二待推荐商品中的至少一个；根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于所述热门商品集合的待推荐商品的推荐权重大于不属于所述热门商品集合的待推荐商品的推荐权重；和按照推荐度从大到小的顺序，向所述特定线下顾客推荐所述多个待推荐商品。
在一些实施例中，所述特定线下顾客的偏好商品根据如下方式来确定：根据所述特定线下顾客以前在所述商铺的驻留位置，生成所述特定线下顾客的热力图；根据所述特定线下顾客的热力图和商品对应的布置位置，计算所述特定线下顾客在不同商品的布置位置的驻留时间；和将驻留时间大于预设时间的布置位置对应的商品确定为所述特定线下顾客的偏好商品。
在一些实施例中，所述推荐模块被配置为：通过自然语言生成技术，将所述特定线下顾客的姓名和所述多个待推荐商品的名称生成符合语法规则的语句；和将所述语句转换成语音，并发送给所述商铺的导购员，以便所述导购员根据所述语音向所述特定线下顾客推荐所述多个待推荐商品。
在一些实施例中，所述推荐模块被配置为：获取每个待推荐商品对应的图片；根据每个待推荐商品对应的图片，生成每个待推荐商品的描述信息；和按照推荐度从大到小的顺序，将每个待推荐商品的图片和描述信息依次输出到所述商铺的显示屏进行显示。
在一些实施例中，每个待推荐商品的商品关注度根据该待推荐商品的访问次数、线上用户的反馈信息、与所述特定线下顾客的习惯匹配度、商品性价比中的至少一项来确定；其中，访问次数越高、线上用户的反馈越好、与所述特定线下顾客的习惯匹配度越高或商品性价比越高，该待推荐商品的商品关注度越高。
根据本公开实施例的又一方面，提供一种商品推荐装置，包括：存储器；以及耦接至所述存储器的处理器，所述处理器被配置为基于存储在所述存储器的指令执行上述任意一个实施例所述的方法。
根据本公开实施例的再一方面，提供一种计算机可读存储介质，其上存储有计算机程序指令，该指令被处理器执行时实现上述任意一个实施例所述的方法。
根据本公开实施例的还一方面，提供商品推荐系统，包括：上述任意一个实施例所述的商品推荐装置；摄像头，被配置为采集当前进入商铺的线下顾客的图像，并将 采集的图像输入到所述商品推荐装置。
本公开实施例中，在向特定线下顾客推荐商品时，综合考虑了与特定线下顾客匹配的线上用户的历史购物信息和热门商品集合。这样的推荐方式对特定线下顾客具有针对性，可以提高向特定线下顾客推荐商品的准确性。另外，可以让特定线下顾客了解时下潮流，为特定线下顾客提供更舒适的购物体验。
下面通过附图和实施例，对本公开的技术方案做进一步的详细描述。
附图说明
构成说明书的一部分的附图描述了本公开的实施例，并且连同说明书一起用于解释本公开的原理。
参照附图，根据下面的详细描述，可以更加清楚地理解本公开，其中：
图1是示出根据本公开一些实施例的商品推荐方法的流程示意图；
图2是示出根据本公开另一些实施例的商品推荐方法的流程示意图；
图3是示出根据本公开一些实现方式的向特定线下顾客推荐商品的流程示意图；
图4是示出根据本公开一些实施例的商品推荐装置的结构示意图；
图5是示出根据本公开另一些实施例的商品推荐装置的结构示意图；
图6是示出根据本公开又一些实施例的商品推荐装置的结构示意图；
图7是示出根据本公开一些实施例的商品推荐系统的结构示意图。
应当明白，附图中所示出的各个部分的尺寸并不必然是按照实际的比例关系绘制的。此外，相同或类似的参考标号表示相同或类似的构件。
具体实施方式
现在将参照附图来详细描述本公开的各种示例性实施例。对示例性实施例的描述仅仅是说明性的，决不作为对本公开及其应用或使用的任何限制。本公开可以以许多不同的形式实现，不限于这里所述的实施例。提供这些实施例是为了使本公开透彻且完整，并且向本领域技术人员充分表达本公开的范围。应注意到：除非另外具体说明，否则在这些实施例中阐述的部件和步骤的相对布置、材料的组分、数字表达式和数值应被解释为仅仅是示例性的，而不是作为限制。
本公开中使用的“第一”、“第二”以及类似的词语并不表示任何顺序、数量或者重要性，而只是用来区分不同的部分。“包括”或者“包含”等类似的词语意指在 该词前的要素涵盖在该词后列举的要素，并不排除也涵盖其他要素的可能。“上”、“下”等仅用于表示相对位置关系，当被描述对象的绝对位置改变后，则该相对位置关系也可能相应地改变。
在本公开中，当描述到特定部件位于第一部件和第二部件之间时，在该特定部件与第一部件或第二部件之间可以存在居间部件，也可以不存在居间部件。当描述到特定部件连接其它部件时，该特定部件可以与所述其它部件直接连接而不具有居间部件，也可以不与所述其它部件直接连接而具有居间部件。
本公开使用的所有术语(包括技术术语或者科学术语)与本公开所属领域的普通技术人员理解的含义相同，除非另外特别定义。还应当理解，在诸如通用字典中定义的术语应当被解释为具有与它们在相关技术的上下文中的含义相一致的含义，而不应用理想化或极度形式化的意义来解释，除非这里明确地这样定义。
对于相关领域普通技术人员已知的技术、方法和设备可能不作详细讨论，但在适当情况下，所述技术、方法和设备应当被视为说明书的一部分。
图1示出是根据本公开一些实施例的商品推荐方法的流程示意图。
在步骤102，对当前进入设定区域(例如商铺)的特定线下顾客的图像进行分析，以得到特定线下顾客的属性。
例如，可以在商铺中安装摄像头来采集进入商铺的线下顾客的图像。采用人脸检测识别技术对线下顾客的图像进行处理，可以识别出特定线下顾客，例如VIP顾客。
进而，可以采用人脸属性分析技术，对特定线下顾客的图像进行分析，以得到特定线下顾客的属性。例如，可以将检测到的人脸区域的信息输入到深度学习模型中，以得到特定线下顾客的属性。
特定线下顾客的属性例如可以包括但不限于下列中的至少一项：性别、年龄、表情、种族、颜值、是否戴眼镜、是否戴墨镜、是否留胡子等。例如，特定线下顾客的属性可以表示为属性向量，该属性向量中的每个元素表示某种属性。例如，向量(1,2,0,1,1,1)表示特定线下顾客的年龄为20-30，白种人，女性，不戴口罩，戴眼镜，不戴墨镜。
在步骤104，根据特定线下顾客的属性，确定与特定线下顾客匹配的线上用户。
应理解，线上用户可以是任何通过互联网平台进行消费的用户。例如，可以从线上数据库获取线上用户的属性，然后将特定线下顾客的属性与线上用户的属性进行匹配(例如比对属性向量)，以确定与特定线下顾客的属性相同或相似的线上用户。与 特定线下顾客属性相同或相似的线上用户即可视为与特定线下顾客匹配的线上用户。
线上用户的属性可以采用与上述特定线下顾客的属性的确定方法类似的方法来确定。
为了提高向特定线下顾客推荐商品的准确性，可以根据特定线下顾客的归属地来选取部分线上用户用于后续商品推荐，如此可以减少数据处理量，提高商品推荐效率。另外，根据与特定线下顾客的归属地相同的线上用户的数据为特定线下顾客推荐商品，也更加准确。例如，在对北京某商铺内的特定线下顾客推荐商品时，可以选取归属地为北京的线上用户用于后续商品推荐。在一些实施例中，可以根据线上用户在互联网平台的注册地址、常用的送货地址或网络IP地址来确定线上用户的归属地。
在步骤106，根据以前进入商铺的线下顾客的驻留位置对应的对象(例如商品)，构建商铺的热门商品集合。
例如，可以根据以前进入商铺的线下顾客的驻留位置，生成线下顾客的热力图。然后根据线下顾客的热力图和商品对应的布置位置，得到热门商品集合。
线下顾客的热力图可以反映过去一段时间内的线下顾客在商铺的驻留位置的分布情况。根据线下顾客在商铺的驻留位置的分布情况可以得到线下顾客对哪些商品更感兴趣，进而可以得到热门商品集合。
在步骤108，根据线上用户的历史购物信息和热门商品集合，向特定线下顾客推荐商品。
下面介绍一种向特定线下顾客推荐商品的实现方式。
首先，根据与特定线下顾客匹配的线上用户的历史购物信息，确定多个待推荐商品。
历史购物信息可以包括线上用户购买过的商品、收藏过的商品等。
然后，根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度。这里，属于热门商品集合的待推荐商品的推荐权重大于不属于热门商品集合的待推荐商品的推荐权重。
在一些实施例中，可以计算每个待推荐商品的商品关注度和推荐权重的乘积，以得到每个待推荐商品的推荐度。在另一些实施例中，在计算每个待推荐商品的商品关注度和推荐权重的乘积s后，还可以计算每个待推荐商品的商品关注度和推荐权重的乘积之和S，然后计算每个乘积s与S的比值，将该比值作为每个待推荐商品的推荐度。
例如，每个待推荐商品的商品关注度可以根据下列中的至少一项来确定：该待推荐商品的访问次数、线上用户的反馈信息、与特定线下顾客的习惯匹配度、商品性价比。访问次数越高、线上用户的反馈越好、与特定线下顾客的习惯匹配度越高或商品性价比越高，该待推荐商品的商品关注度越高。
例如，每个待推荐商品的推荐权重的初始值可以相同，如果某个待推荐商品属于热门商品集合，则增大该待推荐商品的推荐权重。
之后，按照推荐度从大到小的顺序，向特定线下顾客推荐多个待推荐商品，即，优先推荐推荐度更大的待推荐商品。在某些实施例中，可以仅推荐多个待推荐商品中推荐度大于预设推荐度的待推荐商品。
在一些实施例中，可以根据如下方式向特定线下顾客推荐多个待推荐商品：首先，通过自然语言生成技术，将特定线下顾客的姓名和多个待推荐商品的名称生成符合语法规则的语句；然后，将语句转换成语音，并发送给商铺的导购员，以便导购员根据语音向特定线下顾客推荐多个待推荐商品，从而为特定线下顾客提供定制化的服务。例如，张三进入店铺，依次向他推荐灰色卫衣、白色帽子、蓝色牛仔裤等。
在另一些实施例中，可以根据如下方式向线下顾客推荐多个待推荐商品：首先，获取每个待推荐商品对应的图片。然后，根据每个待推荐商品对应的图片，生成每个待推荐商品的描述信息。例如，可以将待推荐商品对应的图片输入到图像标注(Image Caption)模型，该模型可以输出对输入的图片进行描述的信息，如此即可得到每个待推荐商品的描述信息。之后，按照推荐度从大到小的顺序，将每个待推荐商品的图片和描述信息依次输出到商铺的显示屏进行显示，以引起特定线下顾客的注意。例如，优先在显示屏显示推荐度更大的待推荐商品。例如，依次在显示屏显示灰色卫衣、白色帽子、蓝色牛仔裤等，并配上相应的文字描述。
上述实施例中，在向特定线下顾客推荐商品时，综合考虑了与特定线下顾客匹配的线上用户的历史购物信息和热门商品集合。这样的推荐方式对特定线下顾客具有针对性，可以提高向特定线下顾客推荐商品的准确性。另外，可以让特定线下顾客了解时下潮流，为特定线下顾客提供更舒适的购物体验。
图2是示出根据本公开另一些实施例的商品推荐方法的流程示意图。下面重点介绍图2与图1的不同之处，其他相关之处可以参照上面的描述，在此不再赘述。
在步骤202，对当前进入商铺的特定线下顾客的图像进行分析，以得到特定线下顾客的属性。
在步骤204，根据特定线下顾客的属性，确定与特定线下顾客匹配的线上用户。
在步骤206，获取特定线下顾客的偏好商品和特定线下顾客在商铺的历史购物信息中的至少一个。
例如，可以根据如下方式来确定特定线下顾客的偏好商品。
首先，根据特定线下顾客以前在商铺的驻留位置，生成特定线下顾客的热力图。该特定线下顾客的热力图可以反映该特定线下顾客以前在商铺的驻留位置的分布。根据该特定线下顾客在商铺的驻留位置的分布情况可以得到该特定线下顾客对商铺内的哪些商品更感兴趣。
然后，根据特定线下顾客的热力图和商品对应的布置位置，计算特定线下顾客在不同商品的布置位置的驻留时间。
之后，将驻留时间大于预设时间的布置位置对应的商品确定为特定线下顾客的偏好商品。
在步骤208，根据以前进入商铺的线下顾客的驻留位置对应的商品，构建商铺的热门商品集合。
在步骤210，根据特定线下顾客的偏好商品和特定线下顾客在商铺的历史购物信息中的至少一个、线上用户的历史购物信息和热门商品集合，向特定线下顾客推荐商品。
这里，需要指出的是，特定线下顾客的偏好商品也可以是该特定线下顾客在商铺的历史购买商品，即曾经购买过的商品。
上述实施例中，在向特定线下顾客推荐商品时，还考虑了该特定线下顾客的偏好商品和该特定线下顾客在商铺的历史购物信息中的至少一个。这样的推荐方式可以向该特定线下顾客推荐其更感兴趣的商品，进一步提高了推荐的准确性。
图3是示出根据本公开一些实现方式的向特定线下顾客推荐商品的流程示意图。
在步骤302，根据与特定线下顾客匹配的线上用户的历史购物信息，确定多个第一待推荐商品。
例如，可以将线上用户曾经购买或收藏的商品确定为第一待推荐商品。
在步骤304，根据特定线下顾客的偏好商品和特定线下顾客在商铺的历史购物信息中的至少一个，确定多个第二待推荐商品。
例如，可以以特定线下顾客的偏好商品或与偏好商品类似的商品作为第二待推荐商品，或者，可以以特定线下顾客在商铺的历史购买商品或与历史购买商品类似的商 品作为第二待推荐商品，或者，可以以特定线下顾客的偏好商品、与偏好商品类似的商品、特定线下顾客在商铺的历史购买商品和与历史购买商品类似的商品均作为第二待推荐商品。
在步骤306，根据多个第一待推荐商品和多个第二待推荐商品，构建待推荐商品集合，待推荐商品集合包括多个待推荐商品。
这里，构建的待推荐商品集合中的多个待推荐商品可以包括步骤302确定的部分或全部第一待推荐商品、以及步骤304确定的部分或全部第二待推荐商品。另外，可以将相同的第一待推荐商品和第二待推荐商品合并为一个待推荐商品。
在步骤308，根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度。这里，属于热门商品集合的待推荐商品的推荐权重大于不属于热门商品集合的待推荐商品的推荐权重。
在步骤310，按照推荐度从大到小的顺序，向特定线下顾客推荐多个待推荐商品。
这里，可以按照上面给出的方式向特定线下顾客推荐多个待推荐商品。
本说明书中各个实施例均采用递进的方式描述，每个实施例重点说明的都是与其它实施例的不同之处，各个实施例之间相同或相似的部分相互参见即可。对于装置实施例而言，由于其与方法实施例基本对应，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
图4是示出根据本公开一些实施例的商品推荐装置的结构示意图。如图4所示，该实施例的商品推荐装置400包括分析模块401、匹配模块402、构建模块403和推荐模块404。
分析模块401被配置为对当前进入商铺的特定线下顾客的图像进行分析，以得到特定线下顾客的属性。
匹配模块402被配置为根据特定线下顾客的属性，确定与特定线下顾客匹配的线上用户。
构建模块403被配置为根据以前进入商铺的线下顾客的驻留位置对应的商品，构建商铺的热门商品集合。在一些实施例中，构建模块403被配置为根据以前进入商铺的线下顾客的驻留位置，生成线下顾客的热力图；根据线下顾客的热力图和商品对应的布置位置，得到热门商品集合。
推荐模块404被配置为根据线上用户的历史购物信息和热门商品集合，向特定线下顾客推荐商品。
上述实施例中，在向特定线下顾客推荐商品时，综合考虑了与特定线下顾客匹配的线上用户的历史购物信息和热门商品集合。这样的推荐方式对特定线下顾客具有针对性，可以提高向特定线下顾客推荐商品的准确性。另外，可以让特定线下顾客了解时下潮流，为特定线下顾客提供更舒适的购物体验。
在一些实施例中，推荐模块404被配置为按照如下方式向特定线下顾客推荐商品：根据线上用户的历史购物信息，确定多个待推荐商品；根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于热门商品集合的待推荐商品的推荐权重大于不属于热门商品集合的待推荐商品的推荐权重；按照推荐度从大到小的顺序，向特定线下顾客推荐多个待推荐商品。作为一些示例，每个待推荐商品的商品关注度可以根据该待推荐商品的访问次数、线上用户的反馈信息、与特定线下顾客的习惯匹配度、商品性价比中的至少一项来确定。这里，访问次数越高、线上用户的反馈越好、与特定线下顾客的习惯匹配度越高或商品性价比越高，该待推荐商品的商品关注度越高。
图5是示出根据本公开另一些实施例的商品推荐装置的结构示意图。如图5所示，该实施例的商品推荐装置500与图4所示的商品推荐装置400相比还包括获取模块501。获取模块501被配置为获取特定线下顾客的偏好商品和特定线下顾客在商铺的历史购物信息中的至少一个。
相应地，该实施例中的推荐模块404被配置为根据特定线下顾客的偏好商品和特定线下顾客在商铺的历史购物信息中的至少一个、线上用户的历史购物信息和热门商品集合，向特定线下顾客推荐商品。
例如，推荐模块404被配置为根据如下方式向特定线下顾客推荐商品：根据线上用户的历史购物信息，确定多个第一待推荐商品；根据特定线下顾客的偏好商品和特定线下顾客在商铺的历史购物信息中的至少一个，确定多个第二待推荐商品；根据多个第一待推荐商品和多个第二待推荐商品，构建待推荐商品集合，待推荐商品集合包括多个待推荐商品，多个待推荐商品包括多个第一待推荐商品中的至少一个和多个第二待推荐商品中的至少一个；根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于热门商品集合的待推荐商品的推荐权重大于不属于热门商品集合的待推荐商品的推荐权重；按照推荐度从大到小的顺序，向特定线下顾客推荐多个待推荐商品。
在一些实施例中，特定线下顾客的偏好商品可以根据如下方式来确定：根据特定 线下顾客以前在商铺的驻留位置，生成特定线下顾客的热力图；根据特定线下顾客的热力图和商品对应的布置位置，计算特定线下顾客在不同商品的布置位置的驻留时间；将驻留时间大于预设时间的布置位置对应的商品确定为特定线下顾客的偏好商品。在确定特定线下顾客的偏好商品后，可以将特定线下顾客的偏好商品存储在偏好商品数据库中。例如，可以不同的标识来区分不同的线下顾客。
在一些实施例中，推荐模块404被配置为通过自然语言生成技术，将特定线下顾客的姓名和多个待推荐商品的名称生成符合语法规则的语句；将语句转换成语音，并发送给商铺的导购员，以便导购员根据语音向特定线下顾客推荐多个待推荐商品。在另一些实施例中，推荐模块404被配置为获取每个待推荐商品对应的图片；根据每个待推荐商品对应的图片，生成每个待推荐商品的描述信息；按照推荐度从大到小的顺序，将每个待推荐商品的图片和描述信息依次输出到商铺的显示屏进行显示。
图6是示出根据本公开又一些实施例的商品推荐装置的结构示意图。如图6所示，该实施例的商品推荐装置600包括存储器601以及耦接至该存储器601的处理器602，处理器602被配置为基于存储在存储器601中的指令，执行前述任意一个实施例中的商品推荐方法。
存储器601例如可以包括系统存储器、固定非易失性存储介质等。系统存储器例如可以存储有操作系统、应用程序、引导装载程序(Boot Loader)以及其他程序等。
商品推荐装置600还可以包括输入输出接口603、网络接口604、存储接口605等。这些接口603、604、605之间、以及存储器601与处理器602之间例如可以通过总线606连接。输入输出接口603为显示器、鼠标、键盘、触摸屏等输入输出设备提供连接接口。网络接口604为各种联网设备提供连接接口。存储接口605为SD卡、U盘等外置存储设备提供连接接口。
图7是示出根据本公开一些实施例的商品推荐系统的结构示意图。如图7所示，该实施例的商品推荐系统包括上述任意一个实施例的商品推荐装置400/500/600和摄像头701。
摄像头701被配置为采集当前进入商铺的线下顾客的图像，并将采集的图像输入到商品推荐装置400/500/600。商品推荐装置400/500/600可以按照上面给出的方式对线下顾客的图像进行处理，进而向特定线下顾客推荐商品。
在一些实施例中，商品推荐系统还可以包括线上数据库702。线上数据库702被配置为存储线上用户的属性和历史购物信息。在一些实施例中，商品推荐系统还可以 包括显示屏703。显示屏703被配置为显示向特定线下顾客推荐的商品信息，例如待推荐商品的图片和描述信息等。
本公开实施例还提供了一种计算机可读存储介质，其上存储有计算机程序指令，该指令被处理器执行时实现上述任意一个实施例的方法。
至此，已经详细描述了本公开的各实施例。为了避免遮蔽本公开的构思，没有描述本领域所公知的一些细节。本领域技术人员根据上面的描述，完全可以明白如何实施这里公开的技术方案。
本领域内的技术人员应当明白，本公开的实施例可提供为方法、系统、或计算机程序产品。因此，本公开可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本公开可采用在一个或多个其中包含有计算机可用程序代码的计算机可用非瞬时性存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本公开是参照根据本公开实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解，可由计算机程序指令实现流程图中一个流程或多个流程和/或方框图中一个方框或多个方框中指定的功能。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
虽然已经通过示例对本公开的一些特定实施例进行了详细说明，但是本领域的技术人员应该理解，以上示例仅是为了进行说明，而不是为了限制本公开的范围。本领域的技术人员应该理解，可在不脱离本公开的范围和精神的情况下，对以上实施例进 行修改或者对部分技术特征进行等同替换。本公开的范围由所附权利要求来限定。

Claims (21)

  1. 一种商品推荐方法，包括：
    对当前进入商铺的特定线下顾客的图像进行分析，以得到所述特定线下顾客的属性；
    根据所述特定线下顾客的属性，确定与所述特定线下顾客匹配的线上用户；
    根据以前进入所述商铺的线下顾客的驻留位置对应的商品，构建所述商铺的热门商品集合；和
    根据所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品。
  2. 根据权利要求1所述的商品推荐方法，其中，根据所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品包括：
    根据所述线上用户的历史购物信息，确定多个待推荐商品；
    根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于所述热门商品集合的待推荐商品的推荐权重大于不属于所述热门商品集合的待推荐商品的推荐权重；和
    按照推荐度从大到小的顺序，向所述特定线下顾客推荐所述多个待推荐商品。
  3. 根据权利要求1或2所述的商品推荐方法，其中，构建所述热门商品集合包括：
    根据以前进入所述商铺的线下顾客的驻留位置，生成所述线下顾客的热力图；和
    根据所述线下顾客的热力图和商品对应的布置位置，得到所述热门商品集合。
  4. 根据权利要求1所述的商品推荐方法，还包括：
    获取所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个；
    根据所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品包括：
    根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物 信息中的至少一个、所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品。
  5. 根据权利要求4所述的商品推荐方法，其中，根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息的至少一个、所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品包括：
    根据所述线上用户的历史购物信息，确定多个第一待推荐商品；
    根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个，确定多个第二待推荐商品；
    根据所述多个第一待推荐商品和所述多个第二待推荐商品，构建待推荐商品集合，所述待推荐商品集合包括多个待推荐商品，所述多个待推荐商品包括所述多个第一待推荐商品中的至少一个和所述多个第二待推荐商品中的至少一个；
    根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于所述热门商品集合的待推荐商品的推荐权重大于不属于所述热门商品集合的待推荐商品的推荐权重；和
    按照推荐度从大到小的顺序，向所述特定线下顾客推荐所述多个待推荐商品。
  6. 根据权利要求4所述的商品推荐方法，其中，所述特定线下顾客的偏好商品根据如下方式来确定：
    根据所述特定线下顾客以前在所述商铺的驻留位置，生成所述特定线下顾客的热力图；
    根据所述特定线下顾客的热力图和商品对应的布置位置，计算所述特定线下顾客在不同商品的布置位置的驻留时间；和
    将驻留时间大于预设时间的布置位置对应的商品确定为所述特定线下顾客的偏好商品。
  7. 根据权利要求2或5所述的商品推荐方法，其中，向所述特定线下顾客推荐所述多个待推荐商品包括：
    通过自然语言生成技术，将所述特定线下顾客的姓名和所述多个待推荐商品的名称生成符合语法规则的语句；和
    将所述语句转换成语音，并发送给所述商铺的导购员，以便所述导购员根据所述语音向所述特定线下顾客推荐所述多个待推荐商品。
  8. 根据权利要求2或5所述的商品推荐方法，其中，向所述特定线下顾客推荐所述多个待推荐商品包括：
    获取每个待推荐商品对应的图片；
    根据每个待推荐商品对应的图片，生成每个待推荐商品的描述信息；和
    按照推荐度从大到小的顺序，将每个待推荐商品的图片和描述信息依次输出到所述商铺的显示屏进行显示。
  9. 根据权利要求2所述的商品推荐方法，其中，每个待推荐商品的商品关注度根据该待推荐商品的访问次数、线上用户的反馈信息、与所述特定线下顾客的习惯匹配度、商品性价比中的至少一项来确定；
    其中，访问次数越高、线上用户的反馈越好、与所述特定线下顾客的习惯匹配度越高或商品性价比越高，该待推荐商品的商品关注度越高。
  10. 一种商品推荐装置，包括：
    分析模块，被配置为对当前进入商铺的特定线下顾客的图像进行分析，以得到所述特定线下顾客的属性；
    匹配模块，被配置为根据所述特定线下顾客的属性，确定与所述特定线下顾客匹配的线上用户；
    构建模块，被配置为根据以前进入所述商铺的线下顾客的驻留位置对应的商品，构建所述商铺的热门商品集合；和
    推荐模块，被配置为根据所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品。
  11. 根据权利要求10所述的商品推荐装置，其中，所述推荐模块被配置为：
    根据所述线上用户的历史购物信息，确定多个待推荐商品；
    根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于所述热门商品集合的待推荐商品的推荐权重大于不属于所述热门商品集合 的待推荐商品的推荐权重；和
    按照推荐度从大到小的顺序，向所述特定线下顾客推荐所述多个待推荐商品。
  12. 根据权利要求10或11所述的商品推荐装置，其中，所述构建模块被配置为：
    根据以前进入所述商铺的线下顾客的驻留位置，生成所述线下顾客的热力图；和
    根据所述线下顾客的热力图和商品对应的布置位置，得到所述热门商品集合。
  13. 根据权利要求10所述的商品推荐装置，还包括：
    获取模块，用于获取所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个；
    所述推荐模块被配置为根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个、所述线上用户的历史购物信息和所述热门商品集合，向所述特定线下顾客推荐商品。
  14. 根据权利要求13所述的商品推荐装置，其中，所述推荐模块被配置为：
    根据所述线上用户的历史购物信息，确定多个第一待推荐商品；
    根据所述特定线下顾客的偏好商品和所述特定线下顾客在所述商铺的历史购物信息中的至少一个，确定多个第二待推荐商品；
    根据所述多个第一待推荐商品和所述多个第二待推荐商品，构建待推荐商品集合，所述待推荐商品集合包括多个待推荐商品，所述多个待推荐商品包括所述多个第一待推荐商品中的至少一个和所述多个第二待推荐商品中的至少一个；
    根据每个待推荐商品的商品关注度和推荐权重，确定每个待推荐商品的推荐度，其中，属于所述热门商品集合的待推荐商品的推荐权重大于不属于所述热门商品集合的待推荐商品的推荐权重；和
    按照推荐度从大到小的顺序，向所述特定线下顾客推荐所述多个待推荐商品。
  15. 根据权利要求13所述的商品推荐装置，其中，所述特定线下顾客的偏好商品根据如下方式来确定：
    根据所述特定线下顾客以前在所述商铺的驻留位置，生成所述特定线下顾客的热力图；
    根据所述特定线下顾客的热力图和商品对应的布置位置，计算所述特定线下顾客在不同商品的布置位置的驻留时间；和
    将驻留时间大于预设时间的布置位置对应的商品确定为所述特定线下顾客的偏好商品。
  16. 根据权利要求11或14所述的商品推荐装置，其中，所述推荐模块被配置为：
    通过自然语言生成技术，将所述特定线下顾客的姓名和所述多个待推荐商品的名称生成符合语法规则的语句；和
    将所述语句转换成语音，并发送给所述商铺的导购员，以便所述导购员根据所述语音向所述特定线下顾客推荐所述多个待推荐商品。
  17. 根据权利要求11或14所述的商品推荐装置，其中，所述推荐模块被配置为：
    获取每个待推荐商品对应的图片；
    根据每个待推荐商品对应的图片，生成每个待推荐商品的描述信息；和
    按照推荐度从大到小的顺序，将每个待推荐商品的图片和描述信息依次输出到所述商铺的显示屏进行显示。
  18. 根据权利要求11所述的商品推荐装置，其中，每个待推荐商品的商品关注度根据该待推荐商品的访问次数、线上用户的反馈信息、与所述特定线下顾客的习惯匹配度、商品性价比中的至少一项来确定；
    其中，访问次数越高、线上用户的反馈越好、与所述特定线下顾客的习惯匹配度越高或商品性价比越高，该待推荐商品的商品关注度越高。
  19. 一种商品推荐装置，包括：
    存储器；以及
    耦接至所述存储器的处理器，所述处理器被配置为基于存储在所述存储器中的指令，执行权利要求1-9中任意一项所述的方法。
  20. 一种计算机可读存储介质，其上存储有计算机程序指令，其中，该指令被处理器执行时实现权利要求1-9中任意一项所述的方法。
  21. 一种商品推荐系统，包括：
    权利要求10-19任意一项所述的商品推荐装置；和
    摄像头，被配置为采集当前进入商铺的线下顾客的图像，并将采集的图像输入到所述商品推荐装置。
PCT/CN2019/097244 2018-07-25 2019-07-23 商品推荐方法、装置、系统及计算机可读存储介质 WO2020020137A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP19840851.0A EP3806021A4 (en) 2018-07-25 2019-07-23 COMMODITY RECOMMENDATION METHOD, DEVICE AND SYSTEM AND COMPUTER READABLE STORAGE MEDIA
US17/258,349 US20210295414A1 (en) 2018-07-25 2019-07-23 Commodity recommendation method, apparatus, system and computer readable storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810822977.X 2018-07-25
CN201810822977.XA CN109034973B (zh) 2018-07-25 2018-07-25 商品推荐方法、装置、系统及计算机可读存储介质

Publications (1)

Publication Number Publication Date
WO2020020137A1 true WO2020020137A1 (zh) 2020-01-30

Family

ID=64645790

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/097244 WO2020020137A1 (zh) 2018-07-25 2019-07-23 商品推荐方法、装置、系统及计算机可读存储介质

Country Status (4)

Country Link
US (1) US20210295414A1 (zh)
EP (1) EP3806021A4 (zh)
CN (1) CN109034973B (zh)
WO (1) WO2020020137A1 (zh)

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108876526B (zh) * 2018-06-06 2021-11-02 北京京东尚科信息技术有限公司 商品推荐方法、装置和计算机可读存储介质
CN109034973B (zh) * 2018-07-25 2021-03-30 北京京东尚科信息技术有限公司 商品推荐方法、装置、系统及计算机可读存储介质
CN109816441B (zh) * 2018-12-29 2021-05-11 江苏云天励飞技术有限公司 策略推送方法、系统及相关装置
CN110322315A (zh) * 2019-05-30 2019-10-11 苏宁云计算有限公司 一种商品自动分配方法、装置、计算机设备及存储介质
CN110288400A (zh) * 2019-06-25 2019-09-27 联想(北京)有限公司 信息处理方法、信息处理装置以及信息处理系统
CN110751501B (zh) * 2019-09-06 2023-08-22 平安科技（深圳）有限公司 商品导购方法、装置、设备及存储介质
CN112907312A (zh) * 2019-12-04 2021-06-04 深圳云天励飞技术有限公司 一种针对线下用户的物品推荐方法及系统、存储介质
CN111598607A (zh) * 2020-04-07 2020-08-28 北京旷视科技有限公司 商铺的部署方法、装置、计算机设备及可读存储介质
CN111750891B (zh) * 2020-08-04 2022-07-12 上海擎感智能科技有限公司 用于信息处理的方法、计算设备和计算机存储介质
CN112215657A (zh) * 2020-10-14 2021-01-12 珠海格力电器股份有限公司 一种推荐商品确定方法、装置、电子设备及存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106296242A (zh) * 2015-05-22 2017-01-04 苏宁云商集团股份有限公司 一种用于电子商务中商品推荐列表的生成方法及生成系统
CN107247759A (zh) * 2017-05-31 2017-10-13 深圳正品创想科技有限公司 一种商品推荐方法及装置
CN107507017A (zh) * 2017-07-07 2017-12-22 阿里巴巴集团控股有限公司 一种线下导购方法和装置
CN108022152A (zh) * 2017-11-30 2018-05-11 北京长城华冠汽车技术开发有限公司 基于图像识别的用户商品自动推荐系统及推荐方法
CN108198051A (zh) * 2018-03-01 2018-06-22 口碑(上海)信息技术有限公司 跨商品类别的商品推荐方法及装置
CN109034973A (zh) * 2018-07-25 2018-12-18 北京京东尚科信息技术有限公司 商品推荐方法、装置、系统及计算机可读存储介质

Family Cites Families (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8438175B2 (en) * 2010-03-17 2013-05-07 Lighthaus Logic Inc. Systems, methods and articles for video analysis reporting
US10977701B2 (en) * 2012-12-04 2021-04-13 Crutchfield Corporation Techniques for providing retail customers a seamless, individualized discovery and shopping experience between online and brick and mortar retail locations
KR20140132844A (ko) * 2013-05-08 2014-11-19 강신태 오프라인과 온라인 결합방식의 상거래정보 처리방법
CN105469263A (zh) * 2014-09-24 2016-04-06 阿里巴巴集团控股有限公司 一种商品推荐方法及装置
CN104392370A (zh) * 2014-11-06 2015-03-04 上海斐讯数据通信技术有限公司 一种自动获取顾客信息的电子商务系统和方法
US10275522B1 (en) * 2015-06-11 2019-04-30 State Farm Mutual Automobile Insurance Company Speech recognition for providing assistance during customer interaction
US20180096440A1 (en) * 2015-06-30 2018-04-05 Amazon Technologies, Inc. Restaurant Specific Personalization
US10387828B2 (en) * 2015-11-12 2019-08-20 Mobile Price Card Electronic product information display and method thereof
US11393007B2 (en) * 2016-03-31 2022-07-19 Under Armour, Inc. Methods and apparatus for enhanced product recommendations
CN206178982U (zh) * 2016-09-30 2017-05-17 重庆智道云科技有限公司 一种基于物联网的实体店顾客、商品、场景管理设备
CN107025589A (zh) * 2017-03-13 2017-08-08 联想(北京)有限公司 一种信息处理方法及电子设备
CN107527268A (zh) * 2017-08-23 2017-12-29 苏州浩哥文化传播有限公司 一种用于购物导航的智能公共服务亭及其工作方法
CN107609945A (zh) * 2017-09-15 2018-01-19 泾县麦蓝网络技术服务有限公司 一种应用于实体商店的商品推荐方法和系统
CN107767168A (zh) * 2017-09-19 2018-03-06 神策网络科技（北京）有限公司 用户行为数据处理方法及装置、电子设备和存储介质
CN107909443B (zh) * 2017-11-27 2020-04-03 北京旷视科技有限公司 信息推送方法、装置及系统
CN108876526B (zh) * 2018-06-06 2021-11-02 北京京东尚科信息技术有限公司 商品推荐方法、装置和计算机可读存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106296242A (zh) * 2015-05-22 2017-01-04 苏宁云商集团股份有限公司 一种用于电子商务中商品推荐列表的生成方法及生成系统
CN107247759A (zh) * 2017-05-31 2017-10-13 深圳正品创想科技有限公司 一种商品推荐方法及装置
CN107507017A (zh) * 2017-07-07 2017-12-22 阿里巴巴集团控股有限公司 一种线下导购方法和装置
CN108022152A (zh) * 2017-11-30 2018-05-11 北京长城华冠汽车技术开发有限公司 基于图像识别的用户商品自动推荐系统及推荐方法
CN108198051A (zh) * 2018-03-01 2018-06-22 口碑(上海)信息技术有限公司 跨商品类别的商品推荐方法及装置
CN109034973A (zh) * 2018-07-25 2018-12-18 北京京东尚科信息技术有限公司 商品推荐方法、装置、系统及计算机可读存储介质

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3806021A4 *

Also Published As

Publication number Publication date
EP3806021A4 (en) 2022-03-23
EP3806021A1 (en) 2021-04-14
US20210295414A1 (en) 2021-09-23
CN109034973A (zh) 2018-12-18
CN109034973B (zh) 2021-03-30

Similar Documents

Publication Publication Date Title
WO2020020137A1 (zh) 商品推荐方法、装置、系统及计算机可读存储介质
JP6543986B2 (ja) 情報処理装置、情報処理方法およびプログラム
KR102220273B1 (ko) 상품 추천 방법 및 이를 이용하는 서버
Lu et al. A video-based automated recommender (VAR) system for garments
US10043109B1 (en) Attribute similarity-based search
US10402917B2 (en) Color-related social networking recommendations using affiliated colors
US20190130285A1 (en) Artificial intelligence system for real-time visual feedback-based refinement of query results
US9607010B1 (en) Techniques for shape-based search of content
US11127074B2 (en) Recommendations based on object detected in an image
CN103038769B (zh) 用于将内容导引到社交网络引擎用户的系统和方法
US20200342320A1 (en) Non-binary gender filter
US10733658B2 (en) Methods and systems of discovery of products in E-commerce
US8655737B1 (en) Brand name synonymy
US20150302515A1 (en) Method, apparatus, and system for simulating objects
US20140379484A1 (en) Visually Generated Consumer Product Presentation
US10970296B2 (en) System and method for data mining and similarity estimation
US11461630B1 (en) Machine learning systems and methods for extracting user body shape from behavioral data
JP6527275B1 (ja) イメージ内の複数の客体の調和に基づく調和検索方法、コンピュータ装置およびコンピュータプログラム
WO2019099913A1 (en) Aspect pre-selection using machine learning
CN113946754A (zh) 基于用户画像的权益推荐方法、装置、设备及存储介质
US20180276731A1 (en) System and Method for Automated Product Recommendations
JP6037540B1 (ja) 検索システム、検索方法およびプログラム
CN116842478A (zh) 一种基于推特内容的用户属性预测方法
KR20190079582A (ko) 선물 추천 서비스 방법 및 이를 실행하기 위한 선물 추천 서버
US10776860B2 (en) Retail website user interface, systems, and methods for displaying trending looks

Legal Events

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

Ref document number: 19840851

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2019840851

Country of ref document: EP

Effective date: 20210107