WO2020056973A1 - 跨平台产品推荐方法、装置、服务器和存储介质 - Google Patents

跨平台产品推荐方法、装置、服务器和存储介质 Download PDF

Info

Publication number
WO2020056973A1
WO2020056973A1 PCT/CN2018/123342 CN2018123342W WO2020056973A1 WO 2020056973 A1 WO2020056973 A1 WO 2020056973A1 CN 2018123342 W CN2018123342 W CN 2018123342W WO 2020056973 A1 WO2020056973 A1 WO 2020056973A1
Authority
WO
WIPO (PCT)
Prior art keywords
platform
product
identifier
recommendation
user
Prior art date
Application number
PCT/CN2018/123342
Other languages
English (en)
French (fr)
Inventor
鲁宁
Original Assignee
平安科技(深圳)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 平安科技(深圳)有限公司 filed Critical 平安科技(深圳)有限公司
Publication of WO2020056973A1 publication Critical patent/WO2020056973A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce
    • G06Q30/06Buying, selling or leasing transactions
    • G06Q30/0601Electronic shopping [e-shopping]
    • G06Q30/0631Item recommendations

Definitions

  • the present application relates to a cross-platform product recommendation method, device, server, and storage medium.
  • Cross-platform products can be a cross-platform product recommendation method that fits user needs well.
  • a cross-platform product recommendation method is provided.
  • a cross-platform product recommendation method includes:
  • the correlation degree is calculated according to the product purchase behavior of the co-purchaser of the first platform and the second platform.
  • a cross-platform product recommendation device includes:
  • a recommendation request receiving module configured to receive a product recommendation request, where the product recommendation request carries a user identifier
  • a first platform product identifier search module configured to find a first platform product identifier corresponding to the user identifier
  • a second platform product identification search module configured to find all second platform product identifications associated with the first platform product identification
  • a target product identification determining module configured to obtain an association degree between the first platform product identifier and the searched second platform product identifier, and determine the target second platform product identifier according to the association degree;
  • the correlation between the first platform product identifier and the second platform product identifier is calculated based on the product purchase behavior of the co-purchaser of the first platform and the second platform;
  • a cross-platform product push module is configured to push the target second platform product identifier to a product recommendation requesting terminal.
  • a server includes a memory and one or more processors.
  • the memory stores computer-readable instructions.
  • the steps of the cross-platform product recommendation method provided in any embodiment of the present application are implemented.
  • One or more non-volatile readable storage media storing computer-readable instructions, and when the computer-readable instructions are executed by one or more processors, the one or more processors implement the provided in any one of the embodiments of the present application Steps of a cross-platform product recommendation method.
  • FIG. 1 is an application scenario diagram of a cross-platform product recommendation method according to one or more embodiments.
  • FIG. 2 is a schematic flowchart of a step of calculating a correlation degree between products on different platforms according to one or more embodiments.
  • FIG. 3 is a schematic flowchart of a cross-platform product recommendation method according to one or more embodiments.
  • FIG. 4 is a schematic flowchart of a step of determining a cross-platform product to be recommended according to a correlation degree according to one or more embodiments.
  • FIG. 5 is a block diagram of a cross-platform product recommendation apparatus according to one or more embodiments.
  • FIG. 6 is a block diagram of a server according to one or more embodiments.
  • the cross-platform product recommendation method provided in this application can be applied to the application environment shown in FIG. 1.
  • the user terminal 102, the first platform 104 and the second platform 106, the user terminal 102 and the first platform 104 and the second platform 106 can all communicate through a network, and the first platform 104 and the second platform 106 can communicate through a network.
  • the user terminal 102 may be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices.
  • the first platform 104 and the second platform 106 may use independent servers or a server cluster composed of multiple servers. to fulfill.
  • Step 202 Determine a co-purchaser of the first platform and the second platform, a first platform user identifier corresponding to the co-purchaser is associated with the first platform product identifier, and a second platform user identifier corresponding to the co-purchaser is associated with the second platform product identifier.
  • the registered user is the purchaser of the platform.
  • the same person purchases the first platform product on the first platform and the second platform product on the second platform. This person is the co-purchaser.
  • the method for determining the co-purchaser of the first platform and the second platform may be as follows: the first platform obtains all the first platform user identifiers associated with at least one first platform product identifier in the platform.
  • the second platform obtains all second platform user IDs associated with at least one second platform product ID within the platform.
  • the second platform may send all the obtained second platform user identifiers associated with at least one second platform product identifier and user information corresponding to the second platform user identifier to the first platform. If the first platform user ID and the second platform user ID correspond to the same user information, the first platform user ID or the second platform user ID is a co-purchaser.
  • Little A is the user ID of the first platform, and Little A purchases the product a of the first platform; then Little A is associated with product a.
  • the first platform obtains the following data from the second platform: Little A 'is the user identifier of the second platform, and Little A' has purchased the product b of the second platform; it is determined that the small A and the small A 'are the same through the user information associated with the user identifier.
  • Person, small A (small A ') is the co-purchaser.
  • the first platform may mark identified co-purchasers.
  • Step 204 Determine all cross-platform product combinations of the first platform and the second platform and the number of purchases corresponding to each product combination according to the first platform product identifier and the second platform product identifier associated with all co-purchasers.
  • the co-purchaser corresponds to a first platform user identity and a first second platform user identity, and obtains the first platform product identity (that is, the co-purchaser purchased on the first platform) associated with the co-purchaser's first platform user identity.
  • First platform product to obtain a second platform product ID associated with the second user ID of the co-purchaser, and determine multiple sets of cross-platform product combinations based on the first platform product ID and the second platform product ID associated with the co-purchaser, where
  • Each cross-platform product combination includes a first platform product purchased by a co-buyer and a second platform product purchased by a co-buyer.
  • a product-to-product combination determined by a product associated with a co-purchaser small A is (a, b), where a is a first platform product and b is a second platform product. If the co-buyer purchases the first platform products a1, a2 and the second platform product b2, the generated cross-platform product combinations are (a1, b2) and (a2, b2).
  • the cross-platform purchase records of all co-purchasers determine the number of purchases corresponding to all cross-platform product combinations of the first platform and the second platform and each cross-platform product combination. Among them, the number of co-purchasers corresponding to cross-platform product combinations is the number of purchases. If the cross-platform product combination corresponding to small A is (a, b), and the combination corresponding to small B, the cross-platform product combination is (a, b). The number of purchases is 2, and so on.
  • Step 206 Calculate an association degree between the first platform product identifier and the second platform product identifier in each product combination according to the number of purchases corresponding to the product combination.
  • the two products in the cross-platform product portfolio have an association relationship, and the specific degree of correlation between the two platform products with an association is calculated by the number of purchases corresponding to each product portfolio combination.
  • the correlation between a1 and b2 is a set value, such as tending to 1. .
  • the correlation between cross-platform products can be calculated by the following formula:
  • n (a1, b2) is the number of purchases corresponding to the (a1, b2) combination
  • ⁇ n (a1, bi) is the sum of the number of purchases of all product combinations including product a1;
  • ⁇ ni is the sum of the number of purchases of all product combinations between the first platform and the second platform
  • ⁇ n (ai, b2) contains the sum of the number of purchases corresponding to all product combinations of product b2.
  • association rules between the cross-sold products are mined, and the association relationship between the products of the two platforms is calculated according to the association rules.
  • a cross-platform product recommendation method is provided, which is applied to the first platform.
  • the platform is taken as an example to explain, including the following steps:
  • Step 302 Receive a product recommendation request, and the product recommendation request carries a user identification.
  • the user terminal logs in to the first platform, and the application terminal on the first platform sends a product recommendation request to the first platform, and the user identifier carried in the product recommendation request is the first platform user identifier in the first platform.
  • Step 304 Obtain the first platform product identifier corresponding to the user identifier.
  • the first platform searches for the first platform product identifier purchased by the user in the platform according to the user identifier, that is, searches for the first platform product identifier associated with the user identifier in the platform.
  • Step 306 Find all the second platform product identifiers associated with the first platform product identifier.
  • the two cross-platform products that constitute a product-to-product combination have an associated relationship.
  • the first platform stores the association relationship between the first platform product and the second platform product. After determining the first platform product identifier corresponding to the user requesting product recommendation, the first platform searches for a second platform product identifier having an association relationship with the first platform product identifier.
  • the products on the first platform are: a1, a2, and the products on the second platform are: b1, b2, b3.
  • the product combination between the first platform and the second platform includes (a1, b1), (a2 , B2) (a1, b2) and (a2, b3).
  • the first platform product identifier corresponding to the user requesting the recommended product is a1, and the second platform product identifier associated with a1 includes b1 and b2.
  • Step 308 Obtain the correlation degree between the first platform product identifier and the searched second platform product identifier, and determine the target second platform product identifier according to the correlation degree.
  • the correlation between the first platform product identifier and the second platform product identifier is based on The product purchase behavior of the co-purchaser of the first platform and the second platform is calculated.
  • a correlation degree between the associated first platform product identification and the second platform product identification is calculated in advance.
  • the first platform obtains the degree of association between the product identifier of the first platform and each of the product identifiers of the second platform found. For example, a1 and b1 and the correlation degree are obtained, and a1 and b2 and the correlation degree are obtained.
  • the second platform product identifier corresponding to the maximum correlation degree may be used as the target second platform product identifier. If a1 and b1 and the correlation degree are 0.7, and a1 and b2 and the correlation degree are 0.9, b2 is the target second platform product identification. It is also possible to use the second platform product identification whose correlation is greater than a set threshold as the target second platform product identification. If the set threshold is 0.7, b1 and b2 are the target second platform product identification.
  • Step 310 Push the target second platform product identifier to the requesting terminal.
  • the second platform product corresponding to the determined target second platform product identifier is used as a recommended product, and is sent by the first platform to a user terminal requesting product recommendation.
  • the product of the other platform is pushed to the users of one platform through the pre-calculated association relationship between the products of the two platforms, and the product is pushed across platforms.
  • the first platform stores the correlation between the first platform product and the second platform product, that is, ⁇ (ai, bi) .
  • the first platform ⁇ (ai, bi) recommends the products of the second platform to the users of the first platform, as shown in the embodiment in FIG. 3.
  • the correlation degree ⁇ (bi, ai) ( ⁇ (bi, ai) ⁇ ⁇ (ai, bi) ) between the second platform product and the first platform product is stored in the second platform, and the second platform user
  • the second platform recommends the product of the first platform to the users of the second platform according to ⁇ (bi, ai) .
  • n (b2, a1) is equal to n (a1, b2); ⁇ n (b2, ai) is equal to ⁇ n (ai, b2); ⁇ n (bi, a1) is equal to ⁇ n (a1, bi).
  • a third platform may also be included. Based on the product sales records of the third platform and the product sales records of the first platform, the co-purchaser of the first platform and the third platform is determined, and the first purchase is calculated based on the purchase behavior of the co-purchaser.
  • the degree of association between the platform product and the third platform product based on the degree of association, pushes the associated third platform product to the first platform user, or pushes the associated first platform product to the third platform user.
  • step 308 determining the target second platform product identifier according to the correlation, including:
  • Step 402 Calculate the full degree of correlation corresponding to each associated second platform product identifier.
  • the full degree of correlation is the sum of the degrees of association between each first platform product identifier corresponding to the user identifier and the second platform product identifier.
  • the second platform product identification associated with the first platform product identification is all second platform product identifications associated with any one of the first platform product identifications.
  • the second platform product identifier associated with a1 includes b1 and b2, and the second platform product identifier associated with a2 is b2 and b3.
  • the second platform product is identified as b1, b2, and b3.
  • the first platform obtains the association degree of the user who requests the recommended product corresponding to all combinations existing between the first platform product identifier and all second platform product identifiers, including (a1, b1), (a2, b2) (a1, b2) and (a2, b3).
  • the full correlation degree of b1 is the correlation degree with a1 and b1 plus the correlation degree of a2 and b1, because a2 and b1 There is no correlation, and the correlation is 0, so the total correlation of b1 is the correlation of a1 and b1.
  • the same b2 total degree of correlation is the degree of correlation with a1 and b2 plus the degree of correlation with a2 and b2.
  • the total degree of correlation of b3 is the degree of correlation with a1 and b3 plus the degree of correlation with a2 and b3, which is the degree of correlation with a2 and b3.
  • Step 404 Determine the target second platform product identifier from all the associated second platform product identifiers according to the full degree of association.
  • the target second platform product identity can be determined by setting a threshold for the full amount of relevance. Or it is recommended that the product identity of the second platform corresponding to the maximum full correlation degree.
  • product recommendation is performed according to the relevance of all products purchased by the user to other platform products, thereby achieving more accurate product recommendation.
  • a product recommendation request sent by a user of the first platform is received, and the first platform product identifier corresponding to the user is searched. If the user has not purchased any products in the platform, the first corresponding product identifier is not found
  • the cluster to which the user ID belongs is found, where user IDs belonging to the same cluster have similar attributes in at least one dimension. Analyze the products preferred by the cluster according to the purchase records of each user in the cluster. The preferred products can be products of this platform or products of other platforms with associated relationships. The product belonging to the cluster preference is recommended to the requesting user as a recommended product.
  • the clustering method may be to group all users in the platform according to the attribute characteristics of the users, and divide users with similar attributes into one cluster. For example, cluster users based on their industry, asset size, and customer attributes. Count the purchase records of users in each cluster, and analyze the product preferences of users in the cluster based on the purchase records of all users in the cluster. If most of the clustered users purchase the A product of the first platform and the B product of the second platform, they recommend the A product of the first platform and the B product of the second platform to the requesting user.
  • the cross-platform product recommendation method may further include the following steps:
  • step a a product recommendation request is received, and the product recommendation request carries a user identification and a product recommendation mode, wherein the product recommendation mode includes intra-platform recommendation, cross-platform recommendation, intra-platform and cross-platform recommendation.
  • Step b if the product recommendation mode is in-platform recommendation, obtain the product identification in the platform to be recommended, and request the product recommendation request terminal to recommend the obtained product identification in the platform.
  • the product recommendation in the platform can adopt the following method. Cluster all users in the platform in advance. When product recommendation in the platform, first find the cluster where the target user identifier (the user requesting product recommendation) is located. Then, the product preference similarity between the target user identifier and each clustered user identifier is calculated according to the product purchase record, the similar users are determined according to the product preference similarity, and the products purchased by the similar users are pushed to the user terminal requesting recommendation.
  • the similarity of product preferences between users can be adopted as follows: a user-product dictionary is constructed in advance, and a product matrix corresponding to the user is generated according to the user-product dictionary. Each element in the matrix identifies the number of times the user purchases each product. Calculate the cosine similarity between the product matrix corresponding to the target user ID and the product matrix corresponding to each clustered user ID.
  • the cosine similarity is the similarity of product preference between the target user and the clustered user.
  • the K clustered users most similar to the target user are determined. Get the products purchased by the K cluster users, and recommend similar cluster user purchase products to the target user. The recommended products do not include products that the target user has already purchased.
  • step c if the product recommendation mode is cross-platform recommendation, the cross-platform product recommendation method provided in FIG. 2, FIG. 3, and FIG. 4 is used to recommend the cross-platform product to the target user.
  • Step d if the product recommendation mode is in-platform and cross-platform recommendation, obtain the in-platform product to be recommended in step b, and obtain the in-platform product to be recommended in step c, and push the obtained in-platform product and cross-platform product To the target user.
  • multiple recommendation modes are set to organically integrate product recommendations within the platform and product recommendations across platforms, making product recommendation more flexible and diversified.
  • steps in the flowcharts of FIGS. 2-4 are sequentially displayed in accordance with the directions of the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless explicitly stated in this document, the execution of these steps is not strictly limited, and these steps can be performed in other orders. Moreover, at least a part of the steps in Figure 2-4 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily performed at the same time, but may be performed at different times. These sub-steps or stages The execution order of is not necessarily performed sequentially, but may be performed in turn or alternately with at least a part of another step or a sub-step or stage of another step.
  • a cross-platform product recommendation device including: a recommendation request receiving module 502, a first platform product identification search module 504, a second platform product identification search module 506, and a target product
  • the identification determination module 508 and the cross-platform product push module 510 are provided, including: a recommendation request receiving module 502, a first platform product identification search module 504, a second platform product identification search module 506, and a target product.
  • the recommendation request receiving module 502 is configured to receive a product recommendation request, where the product recommendation request carries a user identifier.
  • the first platform product identifier search module 504 is configured to obtain a first platform product identifier corresponding to the user identifier.
  • the second platform product identifier search module 506 is configured to search for all second platform product identifiers associated with the first platform product identifier.
  • the target product identification determining module 508 is configured to obtain a correlation between the first platform product identification and the searched second platform product identification, and determine a target second platform product identification according to the correlation;
  • the correlation between the platform product identifier and the second platform product identifier is calculated according to the product purchase behavior of the co-purchaser of the first platform and the second platform.
  • the cross-platform product push module 510 is configured to push the target second platform product identifier to a product recommendation requesting terminal.
  • the cross-platform product recommendation device further includes:
  • the co-purchaser determination module is configured to determine a co-purchaser of the first platform and the second platform, and a first platform user identifier corresponding to the co-purchaser is associated with a first platform product identifier, and the co-purchaser corresponds to The second platform user ID is associated with the second platform product ID.
  • the inter-product combination module is configured to determine all cross-platform product combinations of the first platform and the second platform according to the first platform product identifier and the second platform product identifier associated with all the co-purchasers. , And the number of purchases for each combination of products.
  • An association degree calculation module is configured to calculate an association degree between the first platform product identifier and the second platform product identifier in each of the inter-product combinations according to the number of purchases corresponding to the inter-product combinations.
  • the user identifier corresponds to multiple first platform product identifiers; the target product identifier determining module 508 is further configured to calculate a full degree of association corresponding to each associated second platform product identifier, so The full amount of correlation is the sum of the correlations between each of the first platform product identifier and the second platform product identifier corresponding to the user identification; and according to the full amount of correlation, all the second The target product identifier of the second platform is determined in the platform product identifier.
  • the cross-platform product recommendation device further includes:
  • a cluster recommendation module configured to find the cluster to which the user ID belongs if the first platform product ID corresponding to the user ID is not found; obtain the product ID preferred by the cluster, and push the product ID To a product recommendation requesting terminal, wherein the product identifier preferred by the cluster includes a first platform product identifier and / or a second platform product identifier.
  • the cross-platform product recommendation device further includes:
  • the sub-pattern recommendation module is configured to receive a product recommendation request, wherein the product recommendation request carries a user identification and a product recommendation mode, wherein the product recommendation mode includes intra-platform recommendation, cross-platform recommendation, intra-platform and cross-platform recommendation; if If the product recommendation mode is in-platform recommendation, obtain the product identification in the platform to be recommended; if the product recommendation mode is cross-platform recommendation, obtain the product identification of the associated platform to be recommended; if the product recommendation mode is platform Internal and cross-platform recommendation, then obtain the product identification in the platform to be recommended and the product identification of the associated platform to be recommended.
  • Each module in the aforementioned cross-platform product recommendation device may be implemented in whole or in part by software, hardware, and a combination thereof.
  • Each of the above modules may be embedded in the processor in the form of hardware or independent of the processor in the server, or may be stored in the memory of the server in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
  • a server is provided, and its internal structure diagram can be as shown in FIG. 6.
  • the server includes a processor, memory, network interface, and database connected via a system bus.
  • the server's processor is used to provide computing and control capabilities.
  • the memory of the server includes a non-volatile readable storage medium and an internal memory.
  • the non-volatile readable storage medium stores an operating system, computer-readable instructions, and a database.
  • the internal memory provides an environment for operating the operating system and computer-readable instructions in a non-volatile readable storage medium.
  • the server's database is used to store the calculated correlation between cross-platform products.
  • the network interface of the server is used to communicate with external terminals through a network connection.
  • the computer-readable instructions are executed by a processor to implement a cross-platform product recommendation method.
  • FIG. 6 is only a block diagram of a part of the structure related to the solution of the application, and does not constitute a limitation on the server to which the solution of the application is applied.
  • the specific server may include More or fewer components are shown in the figure, or some components are combined, or have different component arrangements.
  • a server includes a memory and one or more processors.
  • the memory stores computer-readable instructions.
  • the steps of the cross-platform product recommendation method provided in any embodiment of the present application are implemented.
  • One or more non-volatile readable storage media storing computer-readable instructions, and when the computer-readable instructions are executed by one or more processors, the one or more processors implement the provided in any one of the embodiments of the present application Steps of a cross-platform product recommendation method.
  • Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory.
  • Volatile memory can include random access memory (RAM) or external cache memory.
  • RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Synchlink DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
  • SRAM static RAM
  • DRAM dynamic RAM
  • SDRAM synchronous DRAM
  • DDRSDRAM dual data rate SDRAM
  • ESDRAM enhanced SDRAM
  • SLDRAM synchronous chain Synchlink DRAM
  • Rambus direct RAM
  • DRAM direct memory bus dynamic RAM
  • RDRAM memory bus dynamic RAM

Landscapes

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

Abstract

一种跨平台产品推荐方法,包括:接收产品推荐请求,所述产品推荐请求中携带用户标识;获取所述用户标识对应的第一平台产品标识;查找与所述第一平台产品标识关联的所有第二平台产品标识;获取所述第一平台产品标识与查找的所述第二平台产品标识的关联度,根据所述关联度确定目标第二平台产品标识;其中,所述关联度是根据第一平台和第二平台的共同购买人的产品购买行为计算得到的;及将所述目标第二平台产品标识推送至请求终端。

Description

跨平台产品推荐方法、装置、服务器和存储介质
相关申请的交叉引用
本申请要求于2018年9月18日提交中国专利局,申请号为2018110892056,申请名称为“跨平台产品推荐方法、装置、计算机设备和存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及一种跨平台产品推荐方法、装置、服务器和存储介质。
背景技术
随着电商、P2P等网络销售平台的客户量的迅速增加,为了更好的服务在线客户,线上的产品推荐***几乎被所有的在线网络平台需求。在线的网络平台都在迫切地寻求一种更加贴合用户需求的产品推荐***。
然而,传统的产品推荐***都是基于自身平台数据的平台内产品推荐。当客户想要购买多个平台的产品时,需要进入到多个平台对应的多个应用程序去查看平台的推荐,因此,发明人意识到,迫切需要寻求一种能够跨平台产品推荐且推荐的跨平台产品能够很好的贴合用户需求的跨平台产品推荐方法。
发明内容
根据本申请公开的各种实施例,提供一种跨平台产品推荐方法、装置、服务器和存储介质。
一种跨平台产品推荐方法,包括:
接收产品推荐请求,所述产品推荐请求中携带用户标识;
获取所述用户标识对应的第一平台产品标识;
查找与所述第一平台产品标识关联的所有第二平台产品标识;
获取所述第一平台产品标识与查找的所述第二平台产品标识的关联度,根据所述关联度确定目标第二平台产品标识;
其中,所述关联度是根据第一平台和第二平台的共同购买人的产品购买行为计算得到的;及
将所述目标第二平台产品标识推送至请求终端。
一种跨平台产品推荐装置包括:
推荐请求接收模块,用于接收产品推荐请求,所述产品推荐请求中携带用户标识;
第一平台产品标识查找模块,用于查找所述用户标识对应的第一平台产品标识;
第二平台产品标识查找模块,用于查找与所述第一平台产品标识关联的所有第二平台产品标识;
目标产品标识确定模块,用于获取所述第一平台产品标识与查找的所述第二平台产品标识之间的关联度,根据所述关联度确定目标第二平台产品标识;
其中,所述第一平台产品标识与所述第二平台产品标识的关联度是根据第一平台和第二平台的共同购买人的产品购买行为计算得到的;及
跨平台产品推送模块,用于将所述目标第二平台产品标识推送至产品推荐请求终端。
一种服务器,包括存储器和一个或多个处理器,存储器中存储有计算机可读指令,计算机可读指令被处理器执行时实现本申请任意一个实施例中提供的跨平台产品推荐方法的步骤。
一个或多个存储有计算机可读指令的非易失性可读存储介质,计算机可读指令被一个或多个处理器执行时,使得一个或多个处理器实现本申请任意一个实施例中提供的跨平台产品推荐方法的步骤。
本申请的一个或多个实施例的细节在下面的附图和描述中提出。本申请的其它特征和优点将从说明书、附图以及权利要求书变得明显。
附图说明
为了更清楚地说明本申请实施例中的技术方案,下面将对实施例中所需 要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其它的附图。
图1为根据一个或多个实施例中跨平台产品推荐方法的应用场景图。
图2为根据一个或多个实施例中计算不同平台间产品的关联度步骤的流程示意图。
图3为根据一个或多个实施例中跨平台产品推荐方法的流程示意图。
图4为根据一个或多个实施例中根据关联度确定待推荐的跨平台产品步骤的流程示意图。
图5为根据一个或多个实施例中跨平台产品推荐装置的框图。
图6为根据一个或多个实施例中服务器的框图。
具体实施方式
为了使本申请的技术方案及优点更加清楚明白,以下结合附图及实施例,对本申请进行进一步详细说明。应当理解,此处描述的具体实施例仅仅用以解释本申请,并不用于限定本申请。
本申请提供的跨平台产品推荐方法,可以应用于如图1所示的应用环境中。用户终端102、第一平台104和第二平台106,用户终端102与第一平台104和第二平台106均可通过网络进行通信,第一平台104和第二平台106可通过网络进行通信。用户终端102可以但不限于是各种个人计算机、笔记本电脑、智能手机、平板电脑和便携式可穿戴设备,第一平台104和第二平台106可以用独立的服务器或者是多个服务器组成的服务器集群来实现。
在一些实施例中,如图2所示,在实施跨平台产品推荐方法之前,首先计算第一平台与第二平台中产品间的关联关系,计算第一平台与第二平台产品间的关联度具体包括如下步骤:
步骤202,确定第一平台和第二平台的共同购买人,共同购买人对应的第一平台用户标识关联第一平台产品标识,共同购买人对应的第二平台用户标识关联第二平台产品标识。
在第一平台和第二平台中,注册用户购买了本平台内的产品,则该注册 用户即为该平台的购买人。同一个人在第一平台购买了第一平台产品,又在第二平台中购买了第二平台产品,此人即为共同购买人。
当用户在平台内购买了产品,则用户对应的用户标识与购买的产品标识间具有关联关系。确定第一平台和第二平台的共同购买人可采用如下方法:第一平台获取本平台内关联至少一个第一平台产品标识的所有第一平台用户标识。第二平台获取本平台内关联至少一个第二平台产品标识的所有第二平台用户标识。第二平台可将获取的关联至少一个第二平台产品标识的所有第二平台用户标识以及第二平台用户标识对应的用户信息发送至第一平台。若第一平台用户标识与第二平台用户标识对应同一用户信息,则第一平台用户标识或者第二平台用户标识为共同购买人。
举例来说,小A是第一平台用户标识,小A购买了第一平台的产品a;则小A关联产品a。第一平台从第二平台获取到如下数据:小A’是第二平台用户标识,小A’购买了第二平台的产品b;通过用户标识关联的用户信息判断出小A和小A’同一人,则小A(小A’)是共同购买人。第一平台可对确定的共同购买人进行标记。
步骤204,根据所有共同购买人关联的第一平台产品标识和第二平台产品标识确定第一平台和第二平台的所有跨平台产品间组合,以及每个产品间组合对应的购买次数。
共同购买人对应一个第一平台用户标识和第一个第二平台用户标识,获取共同购买人的第一平台用户标识关联的第一平台产品标识(也就是共同购买人在第一平台内购买的第一平台产品),获取共同购买人的第二用户标识关联的第二平台产品标识,根据共同购买人关联的第一平台产品标识和第二平台产品标识确定多组跨平台产品间组合,其中,每组跨平台产品间组合包括共同购买人购买的一个第一平台产品和共同购买人购买的第二平台产品。
继续步骤202中的举例,由共同购买人小A(小A’)关联的产品确定的一个产品间组合为(a,b),其中,a为第一平台产品,b为第二平台产品。若共同购买人购买了第一平台产品a1、a2,第二平台产品b2,则生成的跨平台产品间组合为(a1、b2)和(a2、b2)。
根据所有共同购买人的跨平台购买记录,确定第一平台和第二平台的所有跨平台产品间组合以及每个跨平台产品间组合对应的购买次数。其中,跨平台产品间组合对应的共同购买人的数量即为购买次数。如小A对应的跨平台产品间组合为(a,b),小B也对应的该组合,则跨平台产品间组合为(a,b)的购买次数为2,依此类推。
步骤206,根据产品间组合对应的购买次数计算每个产品间组合中的第一平台产品标识与第二平台产品标识之间的关联度。
跨平台产品间组合中的两个产品具有关联关系,通过每个产品间组合对应的购买次数计算具有关联的两个平台产品之间的具体的关联度。
在一些实施例中,产品间组合对应的购买次数越多,且产品间组合中的两个产品与其他产品构成的其他产品组合越少,且构成的其他产品间组合对应的购买次数越少,该产品间组合中两个产品之间的关联度越大。
举例来说,一个产品间组合(a1、b2),a1与其他bi均不构成组合,b2与其他ai均不构成组合,则a1、b2间的关联度为设定大值,如趋向于1。
在一些实施例中,可通过如下公式计算跨平台产品间关联度:
Figure PCTCN2018123342-appb-000001
n(a1、b2)是(a1、b2)组合对应的购买次数;
Σn(a1,bi)是包含产品a1的所有产品间组合的购买次数的加和;
Σni是第一平台和第二平台对应的所有产品间组合的购买次数的加和;
Σn(ai、b2)包含产品b2的所有产品间组合对应的购买次数的加和。
本实施例中,基于两个平台间的交叉销售记录,挖掘交叉销售的产品间的关联规则,根据关联规则计算两个平台的产品间的关联关系。
在一些实施例中,基于上一个实施例中计算的第一平台和第二平台产品间的关联度,如图3所示,提供了一种跨平台产品推荐方法,该方法以应用到第一平台为例进行说明,具体包括如下步骤:
步骤302,接收产品推荐请求,产品推荐请求中携带用户标识。
用户终端登录第一平台,在第一平台应用终端向第一平台发送产品推荐 请求,该产品推荐请求携带的用户标识为在第一平台中的第一平台用户标识。
步骤304,获取用户标识对应的第一平台产品标识。
第一平台根据用户标识查找该用户在本平台内购买的第一平台产品标识,即在本平台内查找与用户标识关联的第一平台产品标识。
步骤306,查找与第一平台产品标识关联的所有第二平台产品标识。
获取预先建立的第一平台和第二平台中所有跨平台产品间组合,以及每个产品间组合中两个跨平台产品的关联度。查找是否存在包括第一平台产品标识的产品间组合,若是,包括第一平台产品标识的产品间组合中的第二平台产品标识即为与第一平台产品标识关联的产品标识。
通俗地讲,构成产品间组合的两个跨平台产品具有关联关系。第一平台中存储第一平台产品与第二平台产品之间的关联关系。确定请求产品推荐的用户对应的第一平台产品标识后,第一平台查找与该第一平台产品标识具有关联关系的第二平台产品标识。
举例来说,第一平台中的产品有:a1、a2、第二平台中的产品有:b1、b2、b3,第一平台和第二平台的产品间组合包括(a1、b1)、(a2、b2)(a1、b2)和(a2、b3)。请求推荐产品的用户对应的第一平台产品标识为a1,则与a1关联的第二平台产品标识包括b1、b2。
步骤308,获取第一平台产品标识与查找的第二平台产品标识的关联度,根据关联度确定目标第二平台产品标识;其中,第一平台产品标识与第二平台产品标识的关联度是根据第一平台和第二平台的共同购买人的产品购买行为计算得到的。
预先计算关联的第一平台产品标识与第二平台产品标识之间的关联度。此处,第一平台获取第一平台产品标识与查找出的每个第二平台产品标识之间的关联度。如获取a1与b1和关联度,a1与b2和关联度。
可将对应最大关联度的第二平台产品标识作为目标第二平台产品标识。如a1与b1和关联度为0.7,a1与b2和关联度为0.9,则b2为目标第二平台产品标识。还可将关联度大于设定阈值的第二平台产品标识作为目标第二平台产品标识,如设定阈值是0.7,则将b1和b2为目标第二平台产品标识。
步骤310,将目标第二平台产品标识推送至请求终端。
将确定的目标第二平台产品标识对应的第二平台产品作为推荐产品,由第一平台发送至请求产品推荐的用户终端。
本实施例中,通过预先计算的两个平台产品间的关联关系,向一个平台的用户推送另一个平台的产品,实现了产品的跨平台推送。
需要说明的是,第一平台中存储的是第一平台产品与第二平台产品之间的关联度,即κ (ai、bi),第一平台用户请求推荐跨平台产品时,第一平台根据κ (ai、bi)向第一平台用户推荐第二平台的产品,如图3中的实施例。相应的,第二平台中存储的是第二平台产品与第一平台产品之间的关联度κ (bi、ai)(bi、ai)≠κ (ai、bi)),第二平台用户请求推荐跨平台产品时,第二平台根据κ (bi、ai)向第二平台用户推荐第一平台的产品,以下为第二平台计算关联度的公式:
Figure PCTCN2018123342-appb-000002
n(b2、a1)与n(a1、b2)相等;Σn(b2、ai)与Σn(ai、b2)相等;Σn(bi、a1)与Σn(a1,bi)相等。
进一步的,还可以包括第三平台,通过第三平台的产品销售记录和第一平台的产品销售记录,确定第一平台和第三平台的共同购买人,基于共同购买人的购买行为计算第一平台产品与第三平台产品之间的关联度,基于关联度向第一平台用户推送关联的第三平台产品,或者向第三平台用户推送关联的第一平台产品。以此类推,还可以存在第四平台、第五平台等。
在一些实施例中,若请求产品推荐的第一平台用户标识关联多个第一平台产品标识。如图4所示,步骤308,根据关联度确定目标第二平台产品标识,包括:
步骤402,计算关联的每个第二平台产品标识对应的全量关联度,全量关联度是用户标识对应的每个第一平台产品标识与第二平台产品标识的关联度的加和。
当请求产品推荐的用户标识对应多个第一平台产品标识时,与第一平台产品标识关联的第二平台产品标识为与任意一个第一平台产品标识关联的所 有第二平台产品标识。
如请求推荐产品的用户对应的第一平台产品标识为a1和a2,则与a1关联的第二平台产品标识包括b1和b2,与a2关联的第二平台产品标识为b2和b3,则关联的第二平台产品标识为b1、b2和b3。
此时,第一平台获取请求推荐产品的用户对应第一平台产品标识与所有第二平台产品标识之间存在的所有组合的关联度,包括(a1、b1)、(a2、b2)(a1、b2)和(a2、b3)。
计算关联的每一个第二平台产品标识对应的全量关联度,包括计算b1的全量关联度,b1的全量关联度为与a1与b1的关联度加上a2与b1的关联度,由于a2与b1没有关联关系,关联度为0,所以b1的全量关联度为a1与b1的关联度。同样的b2的全量关联度为与a1与b2的关联度加上a2与b2的关联度。b3的全量关联度为与a1与b3的关联度加上a2与b3的关联度,即为a2与b3的关联度。
步骤404,根据全量关联度从关联的所有第二平台产品标识中确定目标第二平台产品标识。
可通过设置全量关联度阈值确定目标第二平台产品标识。或者推荐最大全量关联度对应的第二平台产品标识。
参考表1,预向平安银行的客户“上海世方建筑工程有限公司”推荐平安产险产品。“上海世方建筑工程有限公司”在平安银行购买的平安银行产品有多个,包括“5L-05单位活期存款”、“5L-05其他国内对公结算业务”、“5L-05集团现金管理”以及“5L-05公司金卫士”。经查找平安产险平台中具有关联的产品为“建筑工程一切险及第三者责任险”,计算“建筑工程一切险及第三者责任险”对应的全量关联度为其与每个“上海世方建筑工程有限公司”在平台银行购买的产品的关联度的加和,即0.4+0.7+0.4+0.45=1.95。若预先设定的全量关联度的阈值为1,1.95大于1,则向“上海世方建筑工程有限公司”推荐平安产险产品“建筑工程一切险及第三者责任险”。
表1
Figure PCTCN2018123342-appb-000003
本实施例中,根据用户购买的所有产品与其他平台产品的关联度进行产品推荐,实现了更加精准的产品推荐。
在一些实施例中,接收第一平台用户发送的产品推荐请求,查找该用户对应的第一平台产品标识,若该用户尚未在本平台内购买过任何产品,即未查找到用户标识对应的第一平台产品标识,则查找该用户标识所属聚类,其中,属于同一聚类的用户标识之间在至少一个维度具有相似的属性。根据聚类中每一个用户的购买记录分析该聚类所偏好的产品,偏好的产品可以是本平台的产品也可以是具有关联关系的其他平台的产品。将所属聚类偏好的产品作为推荐产品推荐给请求用户。
聚类方法可以是,根据用户的属性特征,将平台内的所有用户进行聚类分组,将具有相似属性的用户划分为一个聚类。如根据用户所属行业、资产规模、客户属性等对用户进行聚类。统计每一个聚类的用户的购买记录,根据聚类中所有用户的购买记录,分析该聚类用户的产品偏好。如该聚类用户中大部分用户购买第一平台的A产品,且购买了第二平台的B产品,则向请求用户推荐第一平台的A产品和第二平台的B产品。再如,如该聚类用户中大部分用户购买第一平台的A产品,且第一平台的A产品与第二平台的B产品为强关联关系,(关联度较大)则向请求用户推荐第一平台的A产品和第二平台的B产品。
在一些实施例中,跨平台产品推荐方法还可以包括如下步骤:
步骤a,接收产品推荐请求,产品推荐请求中携带用户标识和产品推荐模式,其中,产品推荐模式包括平台内推荐、跨平台推荐、平台内和跨平台推荐。
步骤b,若产品推荐模式为平台内推荐,则获取待推荐的平台内产品标识,向产品推荐请求终端推荐获取的平台内的产品标识。
平台内的产品推荐可采用如下方法,预先对平台内的所有的用户进行聚类,在平台内产品推荐时,首先查找目标用户标识(请求产品推荐用户)所在聚类。然后,根据产品购买记录计算目标用户标识与每一个聚类用户标识之间的产品偏好相似度,根据产品偏好相似度确定相似用户,向请求推荐的用户终端推送相似用户购买的产品。
用户间的产品偏好相似度可采用如下方法:预先构建用户-产品字典,根据用户-产品字典生成用户对应的产品矩阵,矩阵中的每一个元素标识用户对每个产品的购买次数。计算目标用户标识对应的产品矩阵与每一个聚类用户标识对应的产品矩阵的余弦相似度,余弦相似度即为目标用户与聚类用户之间的产品偏好相似度。根据余弦相似度确定与目标用户的最相似的K个聚类用户。获取这K个聚类用户购买的产品,向目标用户推荐相似的聚类用户购买的产品,推荐的产品不包括目标用户已经购买的产品。
步骤c,若产品推荐模式为跨平台推荐,则采用图2、图3和图4中提供的跨平台产品推荐方法向目标用户推荐跨平台产品。
步骤d,若产品推荐模式为平台内和跨平台推荐,则通过步骤b获取待推荐的平台内产品,通过步骤c获取待推荐的跨平台产品,将获取的平台内产品和跨平台产品均推送给目标用户。
本实施例中,通过设定多种推荐模式,将平台内产品推荐和跨平台产品推荐有机融合,使得产品推荐更加灵活且多元化。
应该理解的是,虽然图2-4的流程图中的各个步骤按照箭头的指示依次显示,但是这些步骤并不是必然按照箭头指示的顺序依次执行。除非本文中有明确的说明,这些步骤的执行并没有严格的顺序限制,这些步骤可以以其它的顺序执行。而且,图2-4中的至少一部分步骤可以包括多个子步骤或者多个阶段,这些子步骤或者阶段并不必然是在同一时刻执行完成,而是可以在不同的时刻执行,这些子步骤或者阶段的执行顺序也不必然是依次进行,而是可以与其它步骤或者其它步骤的子步骤或者阶段的至少一部分轮流或者交替地执行。
在一个实施例中,如图5所示,提供了一种跨平台产品推荐装置,包括: 推荐请求接收模块502、第一平台产品标识查找模块504、第二平台产品标识查找模块506和目标产品标识确定模块508以及跨平台产品推送模块510,
推荐请求接收模块502,用于接收产品推荐请求,所述产品推荐请求中携带用户标识。
第一平台产品标识查找模块504,用于获取所述用户标识对应的第一平台产品标识。
第二平台产品标识查找模块506,用于查找与所述第一平台产品标识关联的所有第二平台产品标识。
目标产品标识确定模块508,用于获取所述第一平台产品标识与查找的所述第二平台产品标识的关联度,根据所述关联度确定目标第二平台产品标识;其中,所述第一平台产品标识与所述第二平台产品标识的关联度是根据第一平台和第二平台的共同购买人的产品购买行为计算得到的。及
跨平台产品推送模块510,用于将所述目标第二平台产品标识推送至产品推荐请求终端。
在一些实施例中,跨平台产品推荐装置还包括:
共同购买人确定模块,用于确定所述第一平台和所述第二平台的共同购买人,所述共同购买人对应的第一平台用户标识关联第一平台产品标识,所述共同购买人对应的第二平台用户标识关联第二平台产品标识。
产品间组合模块,用于根据所有所述共同购买人关联的所述第一平台产品标识和所述第二平台产品标识确定所述第一平台和所述第二平台的所有跨平台产品间组合,以及每个产品间组合对应的购买次数。及
关联度计算模块,用于根据所述产品间组合对应的购买次数计算每个所述产品间组合中的所述第一平台产品标识与所述第二平台产品标识之间的关联度。
在一些实施例中,所述用户标识对应多个第一平台产品标识;所述目标产品标识确定模块508,还用于计算关联的每一个所述第二平台产品标识对应的全量关联度,所述全量关联度为所述用户标识对应的每一个所述第一平台产品标识与所述第二平台产品标识的关联度的加和;根据所述全量关联度,从关联的所有所述第二平台产品标识中确定目标第二平台产品标识。
在一些实施例中,跨平台产品推荐装置还包括:
聚类推荐模块,用于若未查找到所述用户标识对应的第一平台产品标识,则查找所述用户标识所属聚类;获取所述聚类所偏好的产品标识,将所述产品标识推送至产品推荐请求终端,其中,所述聚类所偏好的产品标识包括第一平台产品标识和/或第二平台产品标识。
在一些实施例中,跨平台产品推荐装置还包括:
分模式推荐模块,用于接收产品推荐请求,所述产品推荐请求中携带用户标识和产品推荐模式,其中,所述产品推荐模式包括平台内推荐、跨平台推荐、平台内和跨平台推荐;若所述产品推荐模式为平台内推荐,则获取待推荐的平台内产品标识;若所述产品推荐模式为跨平台推荐,则获取待推荐的关联平台的产品标识;若所述产品推荐模式为平台内和跨平台推荐,则获取待推荐的平台内产品标识和待推荐的关联平台的产品标识。
关于跨平台产品推荐装置的具体限定可以参见上文中对于跨平台产品推荐方法的限定,在此不再赘述。上述跨平台产品推荐装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于服务器中的处理器中,也可以以软件形式存储于服务器中的存储器中,以便于处理器调用执行以上各个模块对应的操作。
在一些实施例中,提供了一种服务器,其内部结构图可以如图6所示。该服务器包括通过***总线连接的处理器、存储器、网络接口和数据库。其中,该服务器的处理器用于提供计算和控制能力。该服务器的存储器包括非易失性可读存储介质、内存储器。该非易失性可读存储介质存储有操作***、计算机可读指令和数据库。该内存储器为非易失性可读存储介质中的操作***和计算机可读指令的运行提供环境。该服务器的数据库用于存储计算的跨平台产品间关联度。该服务器的网络接口用于与外部的终端通过网络连接通信。该计算机可读指令被处理器执行时以实现一种跨平台产品推荐方法。
本领域技术人员可以理解,图6中示出的结构,仅仅是与本申请方案相关的部分结构的框图,并不构成对本申请方案所应用于其上的服务器的限定,具体的服务器可以包括比图中所示更多或更少的部件,或者组合某些部件,或者具有不同的部件布置。
一种服务器,包括存储器和一个或多个处理器,存储器中存储有计算机可读指令,计算机可读指令被处理器执行时实现本申请任意一个实施例中提供的跨平台产品推荐方法的步骤。
一个或多个存储有计算机可读指令的非易失性可读存储介质,计算机可读指令被一个或多个处理器执行时,使得一个或多个处理器实现本申请任意一个实施例中提供的跨平台产品推荐方法的步骤。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机可读指令来指令相关的硬件来完成,所述的计算机可读指令可存储于一非易失性计算机可读取存储介质中,该计算机可读指令在执行时,可包括如上述各方法的实施例的流程。其中,本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用,均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限,RAM以多种形式可得,诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchlink)DRAM(SLDRAM)、存储器总线(Rambus)直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。
以上实施例的各技术特征可以进行任意的组合,为使描述简洁,未对上述实施例中的各个技术特征所有可能的组合都进行描述,然而,只要这些技术特征的组合不存在矛盾,都应当认为是本说明书记载的范围。
以上所述实施例仅表达了本申请的几种实施方式,其描述较为具体和详细,但并不能因此而理解为对发明专利范围的限制。应当指出的是,对于本领域的普通技术人员来说,在不脱离本申请构思的前提下,还可以做出若干变形和改进,这些都属于本申请的保护范围。因此,本申请专利的保护范围应以所附权利要求为准。

Claims (20)

  1. 一种跨平台产品推荐方法,包括:
    接收产品推荐请求,所述产品推荐请求中携带用户标识;
    获取所述用户标识对应的第一平台产品标识;
    查找与所述第一平台产品标识关联的所有第二平台产品标识;
    获取所述第一平台产品标识与查找的所述第二平台产品标识的关联度,根据所述关联度确定目标第二平台产品标识;
    其中,所述关联度是根据第一平台和第二平台的共同购买人的产品购买行为计算得到的;及
    将所述目标第二平台产品标识推送至请求终端。
  2. 根据权利要求1所述的方法,其特征在于,还包括:
    确定所述第一平台和所述第二平台的共同购买人,所述共同购买人对应的第一平台用户标识关联第一平台产品标识,所述共同购买人对应的第二平台用户标识关联第二平台产品标识;
    根据所有所述共同购买人关联的所述第一平台产品标识和所述第二平台产品标识确定所述第一平台和所述第二平台的所有跨平台产品间组合,以及每个产品间组合对应的购买次数;及
    根据所述产品间组合对应的购买次数计算每个所述产品间组合中的所述第一平台产品标识与所述第二平台产品标识之间的关联度。
  3. 根据权利要求1或2所述的方法,其特征在于,所述用户标识对应多个第一平台产品标识;所述根据关联度确定所述目标第二平台产品标识,包括:
    计算关联的每个所述第二平台产品标识对应的全量关联度,所述全量关联度是所述用户标识对应的每个所述第一平台产品标识与所述第二平台产品标识的所述关联度的加和;及
    根据所述全量关联度从关联的所有所述第二平台产品标识中确定目标第二平台产品标识。
  4. 根据权利要求1所述的方法,其特征在于,还包括:
    若未查找到所述用户标识对应的第一平台产品标识,则查找所述用户标 识所属聚类;及
    获取所述聚类所偏好的产品标识,将所述产品标识推送至产品推荐请求终端,其中,所述聚类所偏好的产品标识包括第一平台产品标识和/或第二平台产品标识。
  5. 根据权利要求1所述的方法,其特征在于,还包括:
    接收产品推荐请求,所述产品推荐请求中携带用户标识和产品推荐模式,其中,所述产品推荐模式包括平台内推荐、跨平台推荐、平台内和跨平台推荐;
    若所述产品推荐模式为平台内推荐,则获取待推荐的平台内产品标识;
    若所述产品推荐模式为跨平台推荐,则获取待推荐的关联平台的产品标识;
    若所述产品推荐模式为平台内和跨平台推荐,则获取待推荐的平台内产品标识和待推荐的关联平台的产品标识。
  6. 一种跨平台产品推荐装置,包括:
    推荐请求接收模块,用于接收产品推荐请求,所述产品推荐请求中携带用户标识;
    第一平台产品标识查找模块,用于获取所述用户标识对应的第一平台产品标识;
    第二平台产品标识查找模块,用于查找与所述第一平台产品标识关联的所有第二平台产品标识;
    目标产品标识确定模块,用于获取所述第一平台产品标识与查找的所述第二平台产品标识的关联度,根据所述关联度确定目标第二平台产品标识;
    其中,所述关联度是根据第一平台和第二平台的共同购买人的产品购买行为计算得到的;及
    跨平台产品推送模块,用于将所述目标第二平台产品标识推送至请求终端。
  7. 根据权利要求6所述的装置,其特征在于,还包括:
    共同购买人确定模块,用于确定所述第一平台和所述第二平台的共同购买人,所述共同购买人对应的第一平台用户标识关联第一平台产品标识,所 述共同购买人对应的第二平台用户标识关联第二平台产品标识;
    产品间组合模块,用于根据所有所述共同购买人关联的所述第一平台产品标识和所述第二平台产品标识确定所述第一平台和所述第二平台的所有跨平台产品间组合,以及每个产品间组合对应的购买次数;及
    关联度计算模块,用于根据所述产品间组合对应的购买次数计算每个所述产品间组合中的所述第一平台产品标识与所述第二平台产品标识之间的关联度。
  8. 根据权利要求6或7所述的装置,其特征在于,所述用户标识对应多个第一平台产品标识;所述目标产品标识确定模块,还用于计算关联的每一个所述第二平台产品标识对应的全量关联度,所述全量关联度为所述用户标识对应的每一个所述第一平台产品标识与所述第二平台产品标识的关联度的加和;根据所述全量关联度,从关联的所有所述第二平台产品标识中确定目标第二平台产品标识。
  9. 根据权利要求6所述的装置,其特征在于,还包括:
    聚类推荐模块,用于若未查找到所述用户标识对应的第一平台产品标识,则查找所述用户标识所属聚类;及获取所述聚类所偏好的产品标识,将所述产品标识推送至产品推荐请求终端,其中,所述聚类所偏好的产品标识包括第一平台产品标识和/或第二平台产品标识。
  10. 根据权利要求6所述的装置,其特征在于,还包括:
    分模式推荐模块,用于接收产品推荐请求,所述产品推荐请求中携带用户标识和产品推荐模式,其中,所述产品推荐模式包括平台内推荐、跨平台推荐、平台内和跨平台推荐;
    若所述产品推荐模式为平台内推荐,则获取待推荐的平台内产品标识;
    若所述产品推荐模式为跨平台推荐,则获取待推荐的关联平台的产品标识;
    若所述产品推荐模式为平台内和跨平台推荐,则获取待推荐的平台内产品标识和待推荐的关联平台的产品标识。
  11. 一种服务器,包括存储器及一个或多个处理器,所述存储器中储存有计算机可读指令,所述计算机可读指令被所述一个或多个处理器执行时, 使得所述一个或多个处理器执行以下步骤:
    接收产品推荐请求,所述产品推荐请求中携带用户标识;
    获取所述用户标识对应的第一平台产品标识;
    查找与所述第一平台产品标识关联的所有第二平台产品标识;
    获取所述第一平台产品标识与查找的所述第二平台产品标识的关联度,根据所述关联度确定目标第二平台产品标识;
    其中,所述关联度是根据第一平台和第二平台的共同购买人的产品购买行为计算得到的;及
    将所述目标第二平台产品标识推送至请求终端。
  12. 根据权利要求11所述的服务器,其特征在于,所述处理器执行所述计算机可读指令时还执行以下步骤:
    确定所述第一平台和所述第二平台的共同购买人,所述共同购买人对应的第一平台用户标识关联第一平台产品标识,所述共同购买人对应的第二平台用户标识关联第二平台产品标识;
    根据所有所述共同购买人关联的所述第一平台产品标识和所述第二平台产品标识确定所述第一平台和所述第二平台的所有跨平台产品间组合,以及每个产品间组合对应的购买次数;及
    根据所述产品间组合对应的购买次数计算每个所述产品间组合中的所述第一平台产品标识与所述第二平台产品标识之间的关联度。
  13. 根据权利要求11或12所述的服务器,其特征在于,所述用户标识对应多个第一平台产品标识;
    所述处理器执行所述根据关联度确定所述目标第二平台产品标识时,还执行以下步骤:
    计算关联的每个所述第二平台产品标识对应的全量关联度,所述全量关联度是所述用户标识对应的每个所述第一平台产品标识与所述第二平台产品标识的所述关联度的加和;及
    根据所述全量关联度从关联的所有所述第二平台产品标识中确定目标第二平台产品标识。
  14. 根据权利要求11所述的服务器,其特征在于,所述处理器执行所述 计算机可读指令时还执行以下步骤:
    若未查找到所述用户标识对应的第一平台产品标识,则查找所述用户标识所属聚类;及
    获取所述聚类所偏好的产品标识,将所述产品标识推送至产品推荐请求终端,其中,所述聚类所偏好的产品标识包括第一平台产品标识和/或第二平台产品标识。
  15. 根据权利要求11所述的服务器,其特征在于,所述处理器执行所述计算机可读指令时还执行以下步骤:
    接收产品推荐请求,所述产品推荐请求中携带用户标识和产品推荐模式,其中,所述产品推荐模式包括平台内推荐、跨平台推荐、平台内和跨平台推荐;
    若所述产品推荐模式为平台内推荐,则获取待推荐的平台内产品标识;
    若所述产品推荐模式为跨平台推荐,则获取待推荐的关联平台的产品标识;
    若所述产品推荐模式为平台内和跨平台推荐,则获取待推荐的平台内产品标识和待推荐的关联平台的产品标识。
  16. 一个或多个存储有计算机可读指令的非易失性计算机可读存储介质,所述计算机可读指令被一个或多个处理器执行时,使得所述一个或多个处理器执行以下步骤:
    接收产品推荐请求,所述产品推荐请求中携带用户标识;
    获取所述用户标识对应的第一平台产品标识;
    查找与所述第一平台产品标识关联的所有第二平台产品标识;
    获取所述第一平台产品标识与查找的所述第二平台产品标识的关联度,根据所述关联度确定目标第二平台产品标识;
    其中,所述关联度是根据第一平台和第二平台的共同购买人的产品购买行为计算得到的;及
    将所述目标第二平台产品标识推送至请求终端。
  17. 根据权利要求16所述的存储介质,其特征在于,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    确定所述第一平台和所述第二平台的共同购买人,所述共同购买人对应的第一平台用户标识关联第一平台产品标识,所述共同购买人对应的第二平台用户标识关联第二平台产品标识;
    根据所有所述共同购买人关联的所述第一平台产品标识和所述第二平台产品标识确定所述第一平台和所述第二平台的所有跨平台产品间组合,以及每个产品间组合对应的购买次数;及
    根据所述产品间组合对应的购买次数计算每个所述产品间组合中的所述第一平台产品标识与所述第二平台产品标识之间的关联度。
  18. 根据权利要求16或17所述的存储介质,其特征在于,所述用户标识对应多个第一平台产品标识;
    所述根据关联度确定所述目标第二平台产品标识被所述处理器执行时还执行以下步骤:
    计算关联的每个所述第二平台产品标识对应的全量关联度,所述全量关联度是所述用户标识对应的每个所述第一平台产品标识与所述第二平台产品标识的所述关联度的加和;及
    根据所述全量关联度从关联的所有所述第二平台产品标识中确定目标第二平台产品标识。
  19. 根据权利要求16所述的存储介质,其特征在于,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    若未查找到所述用户标识对应的第一平台产品标识,则查找所述用户标识所属聚类;及
    获取所述聚类所偏好的产品标识,将所述产品标识推送至产品推荐请求终端,其中,所述聚类所偏好的产品标识包括第一平台产品标识和/或第二平台产品标识。
  20. 根据权利要求16所述的存储介质,其特征在于,所述计算机可读指令被所述处理器执行时还执行以下步骤:
    接收产品推荐请求,所述产品推荐请求中携带用户标识和产品推荐模式,其中,所述产品推荐模式包括平台内推荐、跨平台推荐、平台内和跨平台推荐;
    若所述产品推荐模式为平台内推荐,则获取待推荐的平台内产品标识;
    若所述产品推荐模式为跨平台推荐,则获取待推荐的关联平台的产品标识;
    若所述产品推荐模式为平台内和跨平台推荐,则获取待推荐的平台内产品标识和待推荐的关联平台的产品标识。
PCT/CN2018/123342 2018-09-18 2018-12-25 跨平台产品推荐方法、装置、服务器和存储介质 WO2020056973A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811089205.6 2018-09-18
CN201811089205.6A CN109447731A (zh) 2018-09-18 2018-09-18 跨平台产品推荐方法、装置、计算机设备和存储介质

Publications (1)

Publication Number Publication Date
WO2020056973A1 true WO2020056973A1 (zh) 2020-03-26

Family

ID=65532764

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/123342 WO2020056973A1 (zh) 2018-09-18 2018-12-25 跨平台产品推荐方法、装置、服务器和存储介质

Country Status (2)

Country Link
CN (1) CN109447731A (zh)
WO (1) WO2020056973A1 (zh)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112837107B (zh) * 2019-11-22 2023-10-24 上海哔哩哔哩科技有限公司 跨平台商品推荐方法、装置以及计算机设备
CN111031364B (zh) * 2019-12-30 2020-11-06 黑龙江锋速网络科技有限公司 一种网络产品销售的大数据融合方法
CN113313601A (zh) * 2020-02-26 2021-08-27 京东数字科技控股股份有限公司 产品组合的推荐方法、装置及***、存储介质、电子装置
CN113313597B (zh) * 2020-02-26 2023-09-26 京东科技控股股份有限公司 产品组合的推荐方法、装置及***、存储介质、电子装置
CN111680224A (zh) * 2020-04-22 2020-09-18 威比网络科技(上海)有限公司 跨平台课程推送方法、装置、电子设备、存储介质
CN111753210B (zh) * 2020-05-27 2021-05-18 浙江口碑网络技术有限公司 资源推送方法、装置、计算机设备及计算机可读存储介质
CN112364247B (zh) * 2020-11-20 2022-03-04 北京五八信息技术有限公司 一种信息处理方法及装置
CN113487390B (zh) * 2021-08-02 2024-02-23 深圳市唯忆珠宝科技有限公司 珠宝卖家推荐方法、装置、设备和存储介质
CN113742576B (zh) * 2021-08-10 2024-04-26 深圳市东信时代信息技术有限公司 基于跨平台的内容推荐方法、装置、设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130110915A1 (en) * 2008-07-24 2013-05-02 Alibaba Group Holding Limited Correlated information recommendation
CN104463630A (zh) * 2014-12-11 2015-03-25 新一站保险代理有限公司 一种基于网购保险产品特性的产品推荐方法及***
CN105893383A (zh) * 2014-12-17 2016-08-24 深圳楼兰辉煌科技有限公司 基于关联规则推荐算法的车联网信息智能推送方法及***
CN107767154A (zh) * 2016-08-18 2018-03-06 中国电信股份有限公司 信息推送方法、平台和***
CN108038217A (zh) * 2017-12-22 2018-05-15 北京小度信息科技有限公司 信息推荐方法及装置

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105279204B (zh) * 2014-07-25 2019-04-09 阿里巴巴集团控股有限公司 信息推送方法和装置
CN106330846A (zh) * 2015-07-03 2017-01-11 阿里巴巴集团控股有限公司 跨平台的对象推荐方法及装置
US10068213B2 (en) * 2015-09-09 2018-09-04 Mastercard International Incorporated Systems and methods for facilitating cross-platform purchase redirection
US10715849B2 (en) * 2016-07-27 2020-07-14 Accenture Global Solutions Limited Automatically generating a recommendation based on automatic aggregation and analysis of data
CN107341222B (zh) * 2017-06-28 2020-04-07 清华大学 跨平台主题关联方法、装置及其设备

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130110915A1 (en) * 2008-07-24 2013-05-02 Alibaba Group Holding Limited Correlated information recommendation
CN104463630A (zh) * 2014-12-11 2015-03-25 新一站保险代理有限公司 一种基于网购保险产品特性的产品推荐方法及***
CN105893383A (zh) * 2014-12-17 2016-08-24 深圳楼兰辉煌科技有限公司 基于关联规则推荐算法的车联网信息智能推送方法及***
CN107767154A (zh) * 2016-08-18 2018-03-06 中国电信股份有限公司 信息推送方法、平台和***
CN108038217A (zh) * 2017-12-22 2018-05-15 北京小度信息科技有限公司 信息推荐方法及装置

Also Published As

Publication number Publication date
CN109447731A (zh) 2019-03-08

Similar Documents

Publication Publication Date Title
WO2020056973A1 (zh) 跨平台产品推荐方法、装置、服务器和存储介质
WO2021004333A1 (zh) 基于知识图谱的事件处理方法、装置、设备和存储介质
US20240223480A1 (en) Systems and methods for social graph data analytics to determine connectivity within a community
US9990435B2 (en) Controlling access of user information using social-networking information
US10728361B2 (en) System for association of customer information across subscribers
WO2020186786A1 (zh) 文件处理方法、装置、计算机设备和存储介质
US8799306B2 (en) Recommendation of search keywords based on indication of user intention
WO2021012790A1 (zh) 页面数据生成方法、装置、计算机设备及存储介质
US9336314B2 (en) Dynamic facet ordering for faceted search
US20060004789A1 (en) Method of sharing social network information with existing user databases
CN110609737B (zh) 关联数据查询方法、装置、计算机设备和存储介质
CN111046237B (zh) 用户行为数据处理方法、装置、电子设备及可读介质
CN110555164B (zh) 群体兴趣标签的生成方法、装置、计算机设备和存储介质
US9275125B1 (en) System for organizing data from a plurality of users to create individual user profiles
WO2016178655A1 (en) Secure multi-party information retrieval
WO2020108152A1 (zh) 身份数据的防误用方法及装置、电子设备
CN109325796B (zh) ***筛选方法、装置、计算机设备及存储介质
US20200356994A1 (en) Systems and methods for reducing false positives in item detection
US20190294594A1 (en) Identity Data Enhancement
CA3210310A1 (en) Systems and methods for secure storage of sensitive data
CN111047336A (zh) 用户标签推送、用户标签展示方法、装置和计算机设备
CN113392138B (zh) 一种隐私数据的统计分析方法、装置、服务器和存储介质
WO2020134990A1 (zh) 产品信息的查询方法、装置、计算机设备及存储介质
CN114662007B (zh) 数据推荐方法、装置、计算机设备和存储介质
CN114240683A (zh) 群组创建方法、装置、计算机设备和存储介质

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: 18934395

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18934395

Country of ref document: EP

Kind code of ref document: A1