WO2019085327A1 - 电子装置、产品推荐方法、系统和计算机可读存储介质 - Google Patents

电子装置、产品推荐方法、系统和计算机可读存储介质 Download PDF

Info

Publication number
WO2019085327A1
WO2019085327A1 PCT/CN2018/076114 CN2018076114W WO2019085327A1 WO 2019085327 A1 WO2019085327 A1 WO 2019085327A1 CN 2018076114 W CN2018076114 W CN 2018076114W WO 2019085327 A1 WO2019085327 A1 WO 2019085327A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
target
target user
degree
preset
Prior art date
Application number
PCT/CN2018/076114
Other languages
English (en)
French (fr)
Inventor
黄博
王建明
肖京
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 WO2019085327A1 publication Critical patent/WO2019085327A1/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
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking

Definitions

  • the present application relates to the field of computer artificial intelligence, and in particular, to an electronic device, a product recommendation method, a system, and a computer readable storage medium.
  • the traditional recommendation system is usually a content-based association rule recommendation model, that is, based on the previous purchase or browsing information of the target user, recommending other products with high similarity to the previously purchased or browsed products to the target user, which is a great limitation.
  • Sex is to require the target user to have a similar product purchase or browsing history, and the product that the target user does not understand cannot be accurately recommended.
  • the present application provides an electronic device, a product recommendation method, a system, and a computer readable storage medium, which are intended to achieve a targeted and accurate recommendation to a target user for a product that the target user does not understand.
  • a first aspect of the present application provides an electronic device including a memory, a processor, and a memory recommendation system stored on the processor, where the product recommendation system is executed by the processor The following steps are implemented:
  • the related user of the target user is obtained based on the social network, and the target user and each associated user are regarded as nodes, and the directed edge of the target user to the associated user is used as the target user's trust degree to the associated user, and the target user and the associated user are established.
  • Trust relationship network diagram
  • the target products are ranked in descending order according to the degree of interest of the target users on each target product, and the top N target products of the ranked list are selected and recommended to the target users.
  • a second aspect of the present application provides a product recommendation method, the method comprising the steps of:
  • the related user of the target user is obtained based on the social network, and the target user and each associated user are regarded as nodes, and the directed edge of the target user to the associated user is used as the target user's trust degree to the associated user, and the target user and the associated user are established.
  • Trust relationship network diagram
  • the target products are ranked in descending order according to the degree of interest of the target users on each target product, and the top N target products of the ranked list are selected and recommended to the target users.
  • a third aspect of the present application provides a product recommendation system, where the product recommendation system includes:
  • An obtaining module configured to acquire an associated user of the target user based on the social network, use the target user and each associated user as the node, and use the directed edge of the target user to the associated user as the target user's trust degree to the associated user, and establish the target user and a network diagram of trust relationships between associated users;
  • a first calculating module configured to calculate, according to a preset rule, a trust degree of the target user to each associated user, and assign a corresponding weight to the corresponding directed edge in the network diagram of the trust relationship according to the calculated trust degree;
  • a selection module configured to select, according to the trust relationship network diagram, the top K associated users with the highest degree of trust of the target user, where K is a preset value;
  • a second calculation module configured to acquire historical preset behavior data of the first K associated users for the target product for each target product, and according to the trustworthiness of the target user to the top K associated users And the historical preset behavior data of the first K users, and calculating, by using a preset calculation formula, the degree of interest of the target user on the target product;
  • the recommendation module is configured to perform a descending ranking of each target product according to the target user's interest tendency toward each target product, and select the top N target products of the ranked list to recommend to the target user.
  • a fourth aspect of the present application provides a computer readable storage medium storing a product recommendation system, the product recommendation system being executable by at least one processor to cause the at least one processor to execute as follows step:
  • the related user of the target user is obtained based on the social network, and the target user and each associated user are regarded as nodes, and the directed edge of the target user to the associated user is used as the target user's trust degree to the associated user, and the target user and the associated user are established.
  • Trust relationship network diagram
  • the target products are ranked in descending order according to the degree of interest of the target users on each target product, and the top N target products of the ranked list are selected and recommended to the target users.
  • the associated user of the target user is obtained based on the social network, and the network diagram of the trust relationship between the target user and each associated user is constructed according to the trust degree of the target user and each associated user; the target user has the highest trust degree.
  • the first K associated users according to the trustworthiness of the target K users and the historical preset behavior data of the top K users, using the preset calculation formula to calculate the target user
  • the degree of interest of the target product the target users are ranked in descending order according to the degree of interest of each target product, and the top N target products of the ranked list are selected and recommended to the target user.
  • the solution achieves a targeted and accurate recommendation to the target user for products that the target user does not understand.
  • FIG. 1 is a schematic flow chart of an embodiment of a product recommendation method according to the present application.
  • FIG. 2 is a schematic diagram of an operating environment of an embodiment of a product recommendation system of the present application
  • FIG. 3 is a block diagram of a program of an embodiment of a product recommendation system of the present application.
  • FIG. 1 is a schematic flowchart of an embodiment of a product recommendation method according to the present application.
  • the product recommendation method includes:
  • Step S10 Acquire an associated user of the target user based on the social network, use the target user and each associated user as the node, and use the directed edge of the target user to the associated user as the target user's trust degree to the associated user, and establish the target user and the associated user. a network diagram of trust relationships;
  • the social network may be a microblog, a bar, or the like, and the associated user of the target user is, for example, a group of people that the target user pays attention to.
  • the network diagram of the trust relationship between the target user and all the associated users is constructed.
  • the target user and each associated user respectively correspond to the trust relationship.
  • Step S20 Calculate the trust degree of the target user to each associated user according to the preset rule, and assign the corresponding weight to the corresponding directed edge in the trust relationship network map according to the calculated trust degree;
  • the trust degree of the target user to each associated user is calculated according to a preset rule, and each directed edge in the trust relationship network graph is calculated according to the calculated proportional relationship of each trust degree.
  • the weight given to the directed edge corresponding to the associated user with high trust is high, and the weight given by the directed edge corresponding to the associated user with low trust is low.
  • Step S30 Select, according to the trust relationship network diagram, the top K associated users with the highest trust degree of the target user, where K is a preset value;
  • the top K related users with the highest trust degree of the target user are found, that is, the associated users corresponding to the top K trust degree of the descending order of trust, wherein K is The default value (for example, 10).
  • Step S40 Obtain historical preset behavior data of the first K associated users for the target product for each target product, and according to the trust degree of the target user to the top K associated users and the former K-user historical preset behavior data, using a preset calculation formula to calculate the degree of interest of the target user to the target product;
  • the target product is a product or advertisement that the target user does not understand.
  • the historical behavior record of the top K associated users for the target product is obtained by querying historical behavior records of the first K associated users in the social network, and the historical preset behavior data is obtained. Including: purchase, click to browse, collection, etc. And according to the trustworthiness of the target K users and the historical preset behavior data of the first K users, using a preset calculation formula to calculate the interest tendency of the target user to the target product. degree.
  • step S50 the target products are ranked in descending order according to the degree of interest of the target users for each target product, and the top N target products of the ranked list are selected and recommended to the target users.
  • the target products are ranked in descending order, that is, the target product with the highest interest rate of the target user is ranked first; the top N is selected from the ranked list (for example, 3)
  • the target product is recommended to the target user, that is, the top N target products with the highest target interest preference are recommended to the target user.
  • the associated user of the target user is obtained based on the social network, and according to the trust degree of the target user and each associated user, a network diagram of the trust relationship between the target user and each associated user is constructed; and the trust degree of the target user is selected.
  • the highest pre-K associated users according to the trustworthiness of the target K users and the historical preset behavior data of the first K users, using the preset calculation formula to calculate the target
  • the degree of interest of the user to the target product ranking the target products in descending order according to the degree of interest of the target users on each target product, and selecting the top N target products of the ranked list to recommend to the target users.
  • the solution achieves a targeted and accurate recommendation to the target user for products that the target user does not understand.
  • the preset rule is:
  • the social circle overlap degree, the determined interaction frequency, and the determined influence force are weighted and summed according to a preset weight ratio to obtain the trust degree of the target user to the associated user.
  • the social circle overlap degree is the direct trust degree of the target user and the associated user, which may be obtained by first acquiring the crowd set B that the related user pays attention to and the crowd set A that the target user pays attention to; Calculation formula (ie, the similarity formula), the degree of overlap between the associated user and the target user's attention, that is, the social circle overlap degree, is calculated.
  • the frequency of interaction between the two in a preset time period for example, the associated user responds to the target user and responds to the target user within one month The number of times.
  • Determining the influence of the user in the social network for example, determining the influence value according to whether the associated user is a large V and the number of fans of the associated user in the social network.
  • the authority is trusted. Degree and impact are higher for people, so if the associated user is a large V, the influence is increased by a (the first preset value), if the number of fans is ranked in the previous preset name (for example, 1000), the impact Force b (second preset value); and so on.
  • the social circle overlap (denoted as D), the interaction frequency (denoted as E), and the influence (denoted as F) are obtained according to a pre-set weight ratio, and the target user's trust degree to the associated user is obtained by weighted summation;
  • the preset weight ratios of D, E, and F are: 4:4:2, and the final target user's trust for the associated user is: 4D+4E+2F.
  • the preset calculation formula is:
  • P(u,i) is the degree of interest of the target user u on the target product i
  • t(u,K) is the set containing the K related users
  • N(i) is the target product i
  • the set of associated users of the historical preset behavior, T uv represents the trust degree of the target user u to the associated user v
  • r vi is the preset historical behavior value of the associated user v to the target product i.
  • the associated user v has a preset historical behavior value r vi for the target product, and the following manner may be adopted: 1.
  • the present application also proposes a product recommendation system.
  • FIG. 2 is a schematic diagram of an operating environment of a preferred embodiment of the product recommendation system 10 of the present application.
  • the product recommendation system 10 is installed and operated in the electronic device 1.
  • the electronic device 1 may be a computing device such as a desktop computer, a notebook, a palmtop computer, and a server.
  • the electronic device 1 may include, but is not limited to, a memory 11, a processor 12, and a display 13.
  • Figure 2 shows only the electronic device 1 with components 11-13, but it should be understood that not all illustrated components may be implemented, and more or fewer components may be implemented instead.
  • the memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, such as a hard disk or memory of the electronic device 1.
  • the memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in hard disk equipped on the electronic device 1, a smart memory card (SMC), and a secure digital (SD). Card, flash card, etc.
  • the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device.
  • the memory 11 is used to store application software and various types of data installed in the electronic device 1, such as program codes of the product recommendation system 10.
  • the memory 11 can also be used to temporarily store data that has been output or is about to be output.
  • the processor 12 in some embodiments, may be a Central Processing Unit (CPU), microprocessor or other data processing chip for running program code or processing data stored in the memory 11, such as executing a product recommendation system. 10 and so on.
  • CPU Central Processing Unit
  • microprocessor or other data processing chip for running program code or processing data stored in the memory 11, such as executing a product recommendation system. 10 and so on.
  • the display 13 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch sensor, or the like in some embodiments.
  • the display 13 is for displaying information processed in the electronic device 1 and a user interface for displaying visualization, such as a business customization interface or the like.
  • the components 11-13 of the electronic device 1 communicate with one another via a system bus.
  • FIG. 3 is a program module diagram of a preferred embodiment of the product recommendation system 10 of the present application.
  • the product recommendation system 10 can be partitioned into one or more modules, one or more modules being stored in the memory 11 and being processed by one or more processors (the processor 12 in this embodiment). Execute to complete this application.
  • the product recommendation system 10 can be divided into an acquisition module 101, a first calculation module 102, a selection module 103, a second calculation module 104, and a recommendation module 105.
  • a module referred to in this application refers to a series of computer program instruction segments capable of performing a specific function, and is more suitable than the program for describing the execution process of the product recommendation system 10 in the electronic device 1, wherein:
  • the obtaining module 101 is configured to acquire an associated user of the target user based on the social network, use the target user and each associated user as the node, and use the directed edge of the target user to the associated user as the target user's trust degree to the associated user, and establish a target user. a network diagram of trust relationships with associated users;
  • the social network may be a microblog, a bar, or the like, and the associated user of the target user is, for example, a group of people that the target user pays attention to.
  • the network diagram of the trust relationship between the target user and all the associated users is constructed.
  • the target user and each associated user respectively correspond to the trust relationship.
  • the first calculating module 102 is configured to calculate, according to a preset rule, the trust degree of the target user to each associated user, and assign the corresponding weight to the corresponding directed edge in the trust relationship network map according to the calculated trust degree. ;
  • the trust degree of the target user to each associated user is calculated according to a preset rule, and each directed edge in the trust relationship network graph is calculated according to the calculated proportional relationship of each trust degree.
  • the weight given to the directed edge corresponding to the associated user with high trust is high, and the weight given by the directed edge corresponding to the associated user with low trust is low.
  • the selecting module 103 is configured to select, according to the trust relationship network diagram, the top K associated users with the highest trust degree of the target user, where K is a preset value;
  • the top K related users with the highest trust degree of the target user are found, that is, the associated users corresponding to the top K trust degree of the descending order of trust, wherein K is The default value (for example, 10).
  • the second calculation module 104 is configured to acquire, for each target product, historical preset behavior data of the first K associated users for the target product, and according to the trust of the target user to the top K associated users. And the historical preset behavior data of the first K users, and calculating, by using a preset calculation formula, the degree of interest of the target user on the target product;
  • the target product is a product or advertisement that the target user does not understand.
  • the historical behavior record of the top K associated users for the target product is obtained by querying historical behavior records of the first K associated users in the social network, and the historical preset behavior data is obtained. Including: purchase, click to browse, collection, etc. And according to the trustworthiness of the target K users and the historical preset behavior data of the first K users, using a preset calculation formula to calculate the interest tendency of the target user to the target product. degree.
  • the recommendation module 105 is configured to perform a descending ranking of each target product according to the degree of interest of the target user to each target product, and select the top N target products of the ranked list to recommend to the target user.
  • the target products are ranked in descending order, that is, the target product with the highest interest rate of the target user is ranked first; the top N is selected from the ranked list (for example, 3)
  • the target product is recommended to the target user, that is, the top N target products with the highest target interest degree are recommended to the target user.
  • the associated user of the target user is obtained based on the social network, and according to the trust degree of the target user and each associated user, a network diagram of the trust relationship between the target user and each associated user is constructed; and the trust degree of the target user is selected.
  • the highest pre-K associated users according to the trustworthiness of the target K users and the historical preset behavior data of the first K users, using the preset calculation formula to calculate the target
  • the degree of interest of the user to the target product ranking the target products in descending order according to the degree of interest of the target users on each target product, and selecting the top N target products of the ranked list to recommend to the target users.
  • the solution achieves a targeted and accurate recommendation to the target user for products that the target user does not understand.
  • the preset rule is:
  • the social circle overlap degree, the determined interaction frequency, and the determined influence force are weighted and summed according to a preset weight ratio to obtain the trust degree of the target user to the associated user.
  • the social circle overlap degree is the direct trust degree of the target user and the associated user, which may be obtained by first acquiring the crowd set B that the related user pays attention to and the crowd set A that the target user pays attention to; Calculation formula (ie, the similarity formula), the degree of overlap between the associated user and the target user concerned, that is, the degree of overlap of the social circle is calculated.
  • the frequency of interaction between the two in a preset time period for example, the associated user responds to the target user and responds to the target user within one month The number of times.
  • Determining the influence of the user in the social network for example, determining the influence value according to whether the associated user is a large V and the number of fans of the associated user in the social network.
  • the authority is trusted. Degree and impact are higher for people, so if the associated user is a large V, the influence is increased by a (the first preset value), if the number of fans is ranked in the previous preset name (for example, 1000), the impact Force b (second preset value); and so on.
  • the social circle overlap (denoted as D), the interaction frequency (denoted as E), and the influence (denoted as F) are obtained according to a pre-set weight ratio, and the target user's trust degree to the associated user is obtained by weighted summation;
  • the preset weight ratios of D, E, and F are: 4:4:2, and the final target user's trust for the associated user is: 4D+4E+2F.
  • the preset calculation formula is:
  • P(u,i) is the degree of interest of the target user u on the target product i
  • t(u,K) is the set containing the K related users
  • N(i) is the target product i
  • the set of associated users of the historical preset behavior, T uv represents the trust degree of the target user u to the associated user v
  • r vi is the preset historical behavior value of the associated user v to the target product i.
  • the associated user v has a preset historical behavior value r vi for the target product, and the following manner may be adopted: 1.
  • the present application further provides a computer readable storage medium storing a product recommendation system, the product recommendation system being executable by at least one processor to cause the at least one processor to execute The product recommendation method in any of the above embodiments.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Marketing (AREA)
  • General Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • General Health & Medical Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • Primary Health Care (AREA)
  • Human Resources & Organizations (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Development Economics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

一种电子装置、产品推荐方法、系统和计算机可读存储介质，该方法包括：基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图（S10）；按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋予与所述信任关系网络图中对应的有向边（S20）；基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值（S30）；针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度（S40）；按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户（S50）。所述方法能够有针对性地、准确地向目标用户推荐产品。

Description

电子装置、产品推荐方法、系统和计算机可读存储介质
本申请基于巴黎公约申明享有2017年11月1日递交的申请号为CN 201711058995.7、名称为“电子装置、产品推荐方法和计算机可读存储介质”中国专利申请的优先权，该中国专利申请的整体内容以参考的方式结合在本申请中。
技术领域
本申请涉及计算机人工智能领域，特别涉及一种电子装置、产品推荐方法、系统和计算机可读存储介质。
背景技术
传统的推荐系统通常为基于内容的关联规则推荐模型，也就是基于目标用户之前的购买或者浏览信息，向目标用户推荐与之前购买或者浏览的产品相似度高的其他产品，该方法很大的局限性就是要求目标用户有之前类似的产品购买或浏览记录，针对目标用户未了解的产品则无法准确推荐。
发明内容
本申请提供一种电子装置、产品推荐方法、系统和计算机可读存储介质，旨在实现有针对性的、准确的向目标用户推荐目标用户未了解的产品。
本申请第一方面提供一种电子装置，该电子装置包括存储器、处理器，所述存储器上存储有可在所述处理器上运行的产品推荐系统，所述产品推荐系统被所述处理器执行时实现如下步骤：
基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值；
针对每个目标产品，分别获取所述前K个关联用户对该目标产 品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
本申请第二方面提供一种产品推荐方法，该方法包括步骤：
基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值；
针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
本申请第三方面提供一种产品推荐系统，所述产品推荐系统包括：
获取模块，用于基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
第一计算模块，用于按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
选择模块，用于基于所述信任关系网络图，选出该目标用户的信 任度最高的前K个关联用户，K为预设值；
第二计算模块，用于针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
推荐模块，用于按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
本申请第四方面提供一种计算机可读存储介质，所述计算机可读存储介质存储有产品推荐系统，所述产品推荐系统可被至少一个处理器执行，以使所述至少一个处理器执行如下步骤：
基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值；
针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
本申请技术方案，基于社交网络获取目标用户的关联用户，并根据目标用户与各个关联用户的信任度，构建完成目标用户与各个关联用户之间的信任关系网络图；选取目标用户的信任度最高的前K个关 联用户，根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。与现有技术相比，本方案实现了有针对性的、准确的向目标用户推荐目标用户未了解的产品。
附图说明
图1为本申请产品推荐方法一实施例的流程示意图；
图2为本申请产品推荐系统一实施例的运行环境示意图；
图3为本申请产品推荐系统一实施例的程序模块图。
具体实施方式
以下结合附图对本申请的原理和特征进行描述，所举实例只用于解释本申请，并非用于限定本申请的范围。
如图1所示，图1为本申请产品推荐方法一实施例的流程示意图。
本实施例中，该产品推荐方法包括：
步骤S10，基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
所述社交网络可以为微博、贴吧等，目标用户的关联用户例如为目标用户关注的人群。基于所述社交网络获取到目标用户的各个关联用户后，构建目标用户与其所有关联用户之间的信任关系网络图，该信任关系网络图中，目标用户、每个关联用户分别对应为该信任关系网络图中的节点，以目标用户的节点到每个关联用户的节点的有向边代表该目标用户对该关联用户的信任度。
步骤S20，按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
在构建完信任关系网络图后，根据预设规则计算出该目标用户分 别对各个关联用户的信任度，按照计算出的各个信任度的比例关系对所述信任关系网络图中的各个有向边对应赋予权重，即信任度高的关联用户对应的有向边所赋予的权重就高，信任度低的关联用户对应的有向边所赋予的权重就低。
步骤S30，基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值；
基于该信任关系网络图，根据各个有向边的权重，找出该目标用户的信任度最高的前K个关联用户，即信任度降序排名的前K名信任度对应的关联用户，其中K为预设值(例如10)。
步骤S40，针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
目标产品为该目标用户未了解的产品或广告等。对每一个目标产品，通过查询社交网络中所述前K个关联用户的历史行为记录，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，所述历史预设行为数据包括：购买、点击浏览、收藏等。再根据该目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度。
步骤S50，按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
根据得出的该目标用户对各个目标产品的兴趣倾向度，对各个目标产品进行降序排名，即目标用户兴趣倾向度最高的目标产品排名最前；从该排名榜中选出前N(例如3)名目标产品推荐给该目标用户，也就是将目标用户兴趣倾向度最高的前N个目标产品推荐给目标用户。
本实施例技术方案，基于社交网络获取目标用户的关联用户，并根据目标用户与各个关联用户的信任度，构建完成目标用户与各个关 联用户之间的信任关系网络图；选取目标用户的信任度最高的前K个关联用户，根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。与现有技术相比，本方案实现了有针对性的、准确的向目标用户推荐目标用户未了解的产品。
优选地，本实施例中，所述预设规则为：
1、计算关联用户与所述目标用户的社交圈重叠度；
2、确定关联用户与所述目标用户在预设时间段内的互动频次；
3、确定关联用户在社交网络中的影响力；
4、将所述社交圈重叠度、确定的互动频次和确定的影响力按预设权重比例加权求和得到目标用户对所述关联用户的信任度。
其中，社交圈重叠度也就是目标用户与关联用户的直接信任度，其可采取以下方式得出：先获取所述关联用户关注的人群集合B和所述目标用户关注的人群集合A；然后根据计算公式
Figure PCTCN2018076114-appb-000001
(即相似性公式)，计算出关联用户与所述目标用户的关注的人群的重叠程度，即社交圈重叠度。
确定关联用户与该目标用户的互动关系程度，通过社交系统的互动记录中查询得出，两者在预设时间段内的互动频次；比如关联用户近一个月内回复目标用户及被目标用户回复的次数。
关联用户在社交网络中的影响力的确定，例如，可根据该关联用户是否为大V以及该关联用户在社交网络中的粉丝数量排名来确定其影响力值，一般来说，权威的可信度和影响对人们来说更高，因此，若该关联用户是大V，则影响力加a(第一预设数值)，若其粉丝数量排名在前预设名(例如1000)内，影响力加b(第二预设数值)；等等。
最后，将社交圈重叠度(记为D)、互动频次(记为E)和影响 力(记为F)按照预先设置的权重比例，通过加权求和得到目标用户对该关联用户的信任度；例如，D、E和F的预设权重比例为：4:4:2，则最终得到的目标用户对该关联用户的信任度为：4D+4E+2F。
优选地，本实施例中，所述预设的计算公式为：
Figure PCTCN2018076114-appb-000002
其中，P(u，i)就是代表目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。其中，关联用户v对目标产品有预设历史行为值r vi确定可采取以下方式：1、关联用户v对目标产品有预设历史行为(例如，有过购买、点击浏览或收藏)，则r vi＝1；2、关联用户v对目标产品有预设历史行为，该预设历史行为是购买时，r vi＝1，该预设历史行为是点击浏览或收藏时，r vi＝0.8；等等。
此外，本申请还提出一种产品推荐系统。
请参阅图2，是本申请产品推荐系统10较佳实施例的运行环境示意图。
在本实施例中，产品推荐系统10安装并运行于电子装置1中。电子装置1可以是桌上型计算机、笔记本、掌上电脑及服务器等计算设备。该电子装置1可包括，但不仅限于，存储器11、处理器12及显示器13。图2仅示出了具有组件11-13的电子装置1，但是应理解的是，并不要求实施所有示出的组件，可以替代的实施更多或者更少的组件。
存储器11在一些实施例中可以是电子装置1的内部存储单元，例如该电子装置1的硬盘或内存。存储器11在另一些实施例中也可以是电子装置1的外部存储设备，例如电子装置1上配备的插接式硬盘，智能存储卡(Smart Media Card,SMC)，安全数字(Secure Digital,SD)卡，闪存卡(Flash Card)等。进一步地，存储器11还可以既包括电子装置1的内部存储单元也包括外部存储设备。存储器11用于 存储安装于电子装置1的应用软件及各类数据，例如产品推荐系统10的程序代码等。存储器11还可以用于暂时地存储已经输出或者将要输出的数据。
处理器12在一些实施例中可以是一中央处理器(Central Processing Unit,CPU)，微处理器或其他数据处理芯片，用于运行存储器11中存储的程序代码或处理数据，例如执行产品推荐系统10等。
显示器13在一些实施例中可以是LED显示器、液晶显示器、触控式液晶显示器以及OLED(Organic Light-Emitting Diode，有机发光二极管)触摸器等。显示器13用于显示在电子装置1中处理的信息以及用于显示可视化的用户界面，例如业务定制界面等。电子装置1的部件11-13通过系统总线相互通信。
请参阅图3，是本申请产品推荐系统10较佳实施例的程序模块图。在本实施例中，产品推荐系统10可以被分割成一个或多个模块，一个或者多个模块被存储于存储器11中，并由一个或多个处理器(本实施例为处理器12)所执行，以完成本申请。例如，在图3中，产品推荐系统10可以被分割成获取模块101、第一计算模块102、选择模块103、第二计算模块104及推荐模块105。本申请所称的模块是指能够完成特定功能的一系列计算机程序指令段，比程序更适合于描述产品推荐系统10在电子装置1中的执行过程，其中：
获取模块101，用于基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
所述社交网络可以为微博、贴吧等，目标用户的关联用户例如为目标用户关注的人群。基于所述社交网络获取到目标用户的各个关联用户后，构建目标用户与其所有关联用户之间的信任关系网络图，该信任关系网络图中，目标用户、每个关联用户分别对应为该信任关系网络图中的节点，以目标用户的节点到每个关联用户的节点的有向边代表该目标用户对该关联用户的信任度。
第一计算模块102，用于按预设规则分别计算出目标用户对各个 关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
在构建完信任关系网络图后，根据预设规则计算出该目标用户分别对各个关联用户的信任度，按照计算出的各个信任度的比例关系对所述信任关系网络图中的各个有向边对应赋予权重，即信任度高的关联用户对应的有向边所赋予的权重就高，信任度低的关联用户对应的有向边所赋予的权重就低。
选择模块103，用于基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值；
基于该信任关系网络图，根据各个有向边的权重，找出该目标用户的信任度最高的前K个关联用户，即信任度降序排名的前K名信任度对应的关联用户，其中K为预设值(例如10)。
第二计算模块104，用于针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
目标产品为该目标用户未了解的产品或广告等。对每一个目标产品，通过查询社交网络中所述前K个关联用户的历史行为记录，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，所述历史预设行为数据包括：购买、点击浏览、收藏等。再根据该目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度。
推荐模块105，用于按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
根据得出的该目标用户对各个目标产品的兴趣倾向度，对各个目标产品进行降序排名，即目标用户兴趣倾向度最高的目标产品排名最前；从该排名榜中选出前N(例如3)名目标产品推荐给该目标用户， 也就是将目标用户兴趣倾向度最高的前N个目标产品推荐给目标用户。
本实施例技术方案，基于社交网络获取目标用户的关联用户，并根据目标用户与各个关联用户的信任度，构建完成目标用户与各个关联用户之间的信任关系网络图；选取目标用户的信任度最高的前K个关联用户，根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。与现有技术相比，本方案实现了有针对性的、准确的向目标用户推荐目标用户未了解的产品。
优选地，本实施例中，所述预设规则为：
1、计算关联用户与所述目标用户的社交圈重叠度；
2、确定关联用户与所述目标用户在预设时间段内的互动频次；
3、确定关联用户在社交网络中的影响力；
4、将所述社交圈重叠度、确定的互动频次和确定的影响力按预设权重比例加权求和得到目标用户对所述关联用户的信任度。
其中，社交圈重叠度也就是目标用户与关联用户的直接信任度，其可采取以下方式得出：先获取所述关联用户关注的人群集合B和所述目标用户关注的人群集合A；然后根据计算公式
Figure PCTCN2018076114-appb-000003
(即相似性公式)，计算出关联用户与所述目标用户关注的人群的重叠程度，即社交圈重叠度。
确定关联用户与该目标用户的互动关系程度，通过社交系统的互动记录中查询得出，两者在预设时间段内的互动频次；比如关联用户近一个月内回复目标用户及被目标用户回复的次数。
关联用户在社交网络中的影响力的确定，例如，可根据该关联用户是否为大V以及该关联用户在社交网络中的粉丝数量排名来确定其影响力值，一般来说，权威的可信度和影响对人们来说更高，因此， 若该关联用户是大V，则影响力加a(第一预设数值)，若其粉丝数量排名在前预设名(例如1000)内，影响力加b(第二预设数值)；等等。
最后，将社交圈重叠度(记为D)、互动频次(记为E)和影响力(记为F)按照预先设置的权重比例，通过加权求和得到目标用户对该关联用户的信任度；例如，D、E和F的预设权重比例为：4:4:2，则最终得到的目标用户对该关联用户的信任度为：4D+4E+2F。
优选地，本实施例中，所述预设的计算公式为：
Figure PCTCN2018076114-appb-000004
其中，P(u，i)就是代表目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。其中，关联用户v对目标产品有预设历史行为值r vi确定可采取以下方式：1、关联用户v对目标产品有预设历史行为(例如，有过购买、点击浏览或收藏)，则r vi＝1；2、关联用户v对目标产品有预设历史行为，该预设历史行为是购买时，r vi＝1，该预设历史行为是点击浏览或收藏时，r vi＝0.8；等等。
进一步地，本申请还提出一种计算机可读存储介质，所述计算机可读存储介质存储有产品推荐系统，所述产品推荐系统可被至少一个处理器执行，以使所述至少一个处理器执行上述任一实施例中的产品推荐方法。
以上所述仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是在本申请的申请构思下，利用本申请说明书及附图内容所作的等效结构变换，或直接/间接运用在其他相关的技术领域均包括在本申请的专利保护范围内。

Claims (20)

  1. 一种电子装置，其特征在于，所述电子装置包括存储器、处理器，所述存储器上存储有可在所述处理器上运行的产品推荐系统，所述产品推荐系统被所述处理器执行时实现如下步骤：
    基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
    按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
    基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值；
    针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
    按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
  2. 如权利要求1所述的电子装置，其特征在于，所述预设的计算公式为：
    Figure PCTCN2018076114-appb-100001
    其中，P(u，i)为目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。
  3. 如权利要求1所述的电子装置，其特征在于，所述预设规则为：
    计算关联用户与所述目标用户的社交圈重叠度；
    确定关联用户与所述目标用户在预设时间段内的互动频次；
    确定关联用户在社交网络中的影响力；
    将所述社交圈重叠度、确定的互动频次和确定的影响力按预设权重比例加权求和得到目标用户对所述关联用户的信任度。
  4. 如权利要求3所述的电子装置，其特征在于，所述预设的计算公式为：
    Figure PCTCN2018076114-appb-100002
    其中，P(u，i)为目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。
  5. 如权利要求3所述的电子装置，其特征在于，所述计算关联用户与所述目标用户的社交圈重叠度的步骤包括：
    获取所述关联用户关注的人群集合B和所述目标用户关注的人群集合A；
    根据计算公式
    Figure PCTCN2018076114-appb-100003
    计算出关联用户与所述目标用户的社交圈重叠度S AB。
  6. 如权利要求5所述的电子装置，其特征在于，所述预设的计算公式为：
    Figure PCTCN2018076114-appb-100004
    其中，P(u，i)为目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。
  7. 一种产品推荐方法，其特征在于，该方法包括步骤：
    基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关 联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
    按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
    基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值；
    针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
    按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
  8. 如权利要求7所述的产品推荐方法，其特征在于，所述预设的计算公式为：
    Figure PCTCN2018076114-appb-100005
    其中，P(u，i)为目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。
  9. 如权利要求7所述的产品推荐方法，其特征在于，所述预设规则为：
    计算关联用户与所述目标用户的社交圈重叠度；
    确定关联用户与所述目标用户在预设时间段内的互动频次；
    确定关联用户在社交网络中的影响力；
    将所述社交圈重叠度、确定的互动频次和确定的影响力按预设权重比例加权求和得到目标用户对所述关联用户的信任度。
  10. 如权利要求9所述的产品推荐方法，其特征在于，所述预设的计算公式为：
    Figure PCTCN2018076114-appb-100006
    其中，P(u，i)为目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。
  11. 如权利要求9所述的产品推荐方法，其特征在于，所述计算关联用户与所述目标用户的社交圈重叠度的步骤包括：
    获取所述关联用户关注的人群集合B和所述目标用户关注的人群集合A；
    根据计算公式
    Figure PCTCN2018076114-appb-100007
    计算出关联用户与所述目标用户的社交圈重叠度S AB。
  12. 如权利要求11所述的产品推荐方法，其特征在于，所述预设的计算公式为：
    Figure PCTCN2018076114-appb-100008
    其中，P(u，i)为目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。
  13. 一种产品推荐系统，其特征在于，所述产品推荐系统包括：
    获取模块，用于基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
    第一计算模块，用于按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
    选择模块，用于基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值；
    第二计算模块，用于针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
    推荐模块，用于按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
  14. 如权利要求13所述的产品推荐系统，其特征在于，所述预设的计算公式为：
    Figure PCTCN2018076114-appb-100009
    其中，P(u，i)为目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。
  15. 如权利要求13所述的产品推荐系统，其特征在于，所述预设规则为：
    计算关联用户与所述目标用户的社交圈重叠度；
    确定关联用户与所述目标用户在预设时间段内的互动频次；
    确定关联用户在社交网络中的影响力；
    将所述社交圈重叠度、确定的互动频次和确定的影响力按预设权重比例加权求和得到目标用户对所述关联用户的信任度。
  16. 如权利要求15所述的产品推荐系统，其特征在于，所述获取模块通过获取所述关联用户关注的人群集合B和所述目标用户关注的人群集合A后，根据计算公式
    Figure PCTCN2018076114-appb-100010
    计算出关联用户与所述目标用户的社交圈重叠度S AB。
  17. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质存储有产品推荐系统，所述产品推荐系统可被至少一个处理器执行，以使所述至少一个处理器执行如下步骤：
    基于社交网络获取目标用户的关联用户，将目标用户和各个关联用户均作为节点，以目标用户至关联用户的有向边作为目标用户对关联用户的信任度，建立目标用户与关联用户之间的信任关系网络图；
    按预设规则分别计算出目标用户对各个关联用户的信任度，并根据计算出的信任度将对应的权重赋与所述信任关系网络图中对应的有向边；
    基于所述信任关系网络图，选出该目标用户的信任度最高的前K个关联用户，K为预设值；
    针对每个目标产品，分别获取所述前K个关联用户对该目标产品的历史预设行为数据，并根据所述目标用户对所述前K个关联用户的信任度及所述前K个用户的历史预设行为数据，利用预设的计算公式计算得出所述目标用户对该目标产品的兴趣倾向度；
    按所述目标用户对各个目标产品的兴趣倾向度对各个目标产品进行降序排名，选取排名榜的前N名目标产品推荐给所述目标用户。
  18. 如权利要求17所述的计算机可读存储介质，其特征在于，所述预设的计算公式为：
    Figure PCTCN2018076114-appb-100011
    其中，P(u，i)为目标用户u对目标产品i的兴趣倾向度，t(u,K)为包含所述K个关联用户的集合，N(i)为对目标产品i有过的历史预设行为的关联用户集合，T uv代表目标用户u对关联用户v的信任度，r vi为关联用户v对目标产品i的预设历史行为值。
  19. 如权利要求17所述的计算机可读存储介质，其特征在于，所述预设规则为：
    计算关联用户与所述目标用户的社交圈重叠度；
    确定关联用户与所述目标用户在预设时间段内的互动频次；
    确定关联用户在社交网络中的影响力；
    将所述社交圈重叠度、确定的互动频次和确定的影响力按预设权重比例加权求和得到目标用户对所述关联用户的信任度。
  20. 如权利要求19所述的计算机可读存储介质，其特征在于，所述计算关联用户与所述目标用户的社交圈重叠度的步骤包括：
    获取所述关联用户关注的人群集合B和所述目标用户关注的人群集合A；
    根据计算公式
    Figure PCTCN2018076114-appb-100012
    计算出关联用户与所述目标用户的社交圈重叠度S AB。
PCT/CN2018/076114 2017-11-01 2018-02-10 电子装置、产品推荐方法、系统和计算机可读存储介质 WO2019085327A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711058995.7 2017-11-01
CN201711058995.7A CN107909428A (zh) 2017-11-01 2017-11-01 电子装置、产品推荐方法和计算机可读存储介质

Publications (1)

Publication Number Publication Date
WO2019085327A1 true WO2019085327A1 (zh) 2019-05-09

Family

ID=61843320

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/076114 WO2019085327A1 (zh) 2017-11-01 2018-02-10 电子装置、产品推荐方法、系统和计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN107909428A (zh)
WO (1) WO2019085327A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112995256A (zh) * 2019-12-13 2021-06-18 腾讯科技（深圳）有限公司 行为数据处理方法及相关设备

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109727056B (zh) * 2018-07-06 2023-04-18 平安科技（深圳）有限公司 金融机构推荐方法、设备、存储介质及装置
CN108985895B (zh) * 2018-07-10 2021-09-14 西南科技大学 一种获取电子商务中商家信誉值的方法
CN109800356A (zh) * 2019-01-23 2019-05-24 国信优易数据有限公司 一种信息资源推荐方法、装置、设备及存储介质
CN109886823A (zh) * 2019-02-25 2019-06-14 北京奇艺世纪科技有限公司 一种社交圈的推荐方法及装置
CN110457573B (zh) * 2019-07-04 2024-05-07 平安科技（深圳）有限公司 产品推荐方法、装置、计算机设备及存储介质
CN112328865B (zh) * 2019-07-31 2024-01-05 北京达佳互联信息技术有限公司 信息处理和推荐方法、装置、设备及存储介质
CN110737846B (zh) * 2019-10-28 2022-05-31 北京字节跳动网络技术有限公司 一种社交界面推荐方法、装置、电子设备及存储介质
CN110738554B (zh) * 2019-10-28 2023-01-24 拉扎斯网络科技（上海）有限公司 任务处理方法、装置、电子设备及计算机可读存储介质
CN112861015A (zh) * 2019-11-27 2021-05-28 北京达佳互联信息技术有限公司 一种应用程序中账户关联信息获取方法、装置及电子设备
CN113449177A (zh) * 2020-03-27 2021-09-28 北京达佳互联信息技术有限公司 信息推荐方法、装置、电子设备及存储介质
CN112084423B (zh) * 2020-08-31 2023-12-12 腾讯科技（深圳）有限公司 一种信息发布平台推荐方法及装置
CN112269926A (zh) * 2020-10-21 2021-01-26 深圳思为科技有限公司 数据推送方法及相关装置
CN113127749A (zh) * 2021-05-18 2021-07-16 北京大米科技有限公司 一种内容推荐的方法、装置、可读存储介质和电子设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103995823A (zh) * 2014-03-25 2014-08-20 南京邮电大学 一种基于社交网络的信息推荐方法
CN104317959A (zh) * 2014-11-10 2015-01-28 北京字节跳动网络技术有限公司 基于社交平台的数据挖掘方法及装置
CN105677881A (zh) * 2016-01-12 2016-06-15 腾讯科技（深圳）有限公司 一种信息推荐方法、装置及服务器
CN105956201A (zh) * 2016-06-22 2016-09-21 Tcl集团股份有限公司 一种社会化网络的视频推荐方法和装置

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4720853B2 (ja) * 2008-05-19 2011-07-13 ソニー株式会社 情報処理装置、情報処理方法、およびプログラム

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103995823A (zh) * 2014-03-25 2014-08-20 南京邮电大学 一种基于社交网络的信息推荐方法
CN104317959A (zh) * 2014-11-10 2015-01-28 北京字节跳动网络技术有限公司 基于社交平台的数据挖掘方法及装置
CN105677881A (zh) * 2016-01-12 2016-06-15 腾讯科技（深圳）有限公司 一种信息推荐方法、装置及服务器
CN105956201A (zh) * 2016-06-22 2016-09-21 Tcl集团股份有限公司 一种社会化网络的视频推荐方法和装置

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112995256A (zh) * 2019-12-13 2021-06-18 腾讯科技（深圳）有限公司 行为数据处理方法及相关设备
CN112995256B (zh) * 2019-12-13 2022-08-26 腾讯科技（深圳）有限公司 行为数据处理方法及相关设备

Also Published As

Publication number Publication date
CN107909428A (zh) 2018-04-13

Similar Documents

Publication Publication Date Title
WO2019085327A1 (zh) 电子装置、产品推荐方法、系统和计算机可读存储介质
US9965522B2 (en) Mobile application search ranking
US20220092446A1 (en) Recommendation method, computing device and storage medium
WO2018121700A1 (zh) 基于已安装应用来推荐应用信息的方法、装置、终端设备及存储介质
WO2018053966A1 (zh) 点击率预估
US11151493B2 (en) Infrastructure benchmarking based on dynamic cost modeling
US10095695B2 (en) Dynamically determining the relatedness of web objects
WO2020238502A1 (zh) 物品推荐方法及装置、电子设备及存储介质
JP2013511085A (ja) 検索方法およびシステム
US10630788B2 (en) Data packet transmission optimization of data used for content item selection
CN112136127A (zh) 用于搜索操作输出元素的动作指示符
US20190220909A1 (en) Collaborative Filtering to Generate Recommendations
EP2778979A1 (en) Search result ranking by brand
WO2019071890A1 (zh) 产品推荐装置、方法及计算机可读存储介质
US10474670B1 (en) Category predictions with browse node probabilities
US20170357999A1 (en) Method and system for providing ranking information using effect analysis data of information data
US11017452B2 (en) Concerted learning and multi-instance sequential prediction tree
US20150221014A1 (en) Clustered browse history
WO2020056976A1 (zh) 搜索结果优化排序方法、装置、程序及计算机可读存储介质
US20170193539A1 (en) Time-value estimation method and system for sharing environment
US10387934B1 (en) Method medium and system for category prediction for a changed shopping mission
JP2019191686A (ja) 情報処理装置、情報処理システム、情報処理方法、およびプログラム
CN107786376B (zh) 内容推送方法、装置及计算机设备
CN112182386B (zh) 一种基于知识图谱的目标推荐方法及装置
CN109857501B (zh) 一种app的页面展示方法、装置、存储介质和服务器

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: 18872861

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 16/10/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 18872861

Country of ref document: EP

Kind code of ref document: A1