Recommendation system, management server, method for controlling management server, and control program for management server

Abstract

To achieve a highly accurate recommendation service between business operators while protecting personal information.SOLUTION: A management service according to an embodiment of the present disclosure includes: a setting part for setting common users who use both a first service and a second service different from the first service; and a communication part for transmitting information on a commodity on the first service to be recommended to a second similar user similar to the common users in the second service, specified on the basis of a use history in the second service of the common users on the basis of a use history of a first similar user similar to the common users in the first service, specified on the basis of a use history in the first user of the common users.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present disclosure relates to a recommendation system, a management server, a control method for the management server, and a control program for the management server.

Background Art

[0002] Conventionally, in electronic commerce (E-commerce) where businesses buy and sell goods and services on the Internet, collaborative filtering is known as a technique for recommending goods and services according to user preferences (for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] If it is possible to integrate the data of each business operator (such as the purchase history of users) among multiple business operators and execute collaborative filtering, it is considered that the recommendation accuracy will be improved compared to the case using only the data of one's own company. However, passing one's own data to the data of other companies for processing is not preferable from the perspective of personal information protection. Although the use of PPDM (Privacy Preserving Data Mining) can also be considered, it has been difficult to assert that it does not conflict with the Personal Information Protection Law.

Means for Solving the Problems

[0005] A recommendation system for recommending products provided by a first service, which is a service of a first operator, to a user who receives the provision of products of a second operator using a second service, which is a service of a second operator, includes a management server, a first server related to the first operator, and a second server related to the second operator, each of which includes a communication unit for transmitting and receiving data to and from each other. The management server includes a setting unit that receives settings of common users who use both the first service and the second service. The communication unit transmits information about the common users to the first server and the second server. The first server extracts, based on the usage history of the first service by first users who use the first service and include the common users, first similar users among the first users who are similar to the common users in terms of the usage tendency of the first service, and extracts products received by the first similar users using the first service as recommended products. The second server includes a second extraction unit that extracts second similar users among the second users who are similar to the common users in terms of the usage tendency of the second service, based on the usage history of the second service by second users who use the second service and include the common users, and a provision unit that provides information about the recommended products in the first service to the second similar users.

[0006] In a recommendation system according to an embodiment of the present disclosure, the management server further includes an instruction generation unit that generates various instructions for the first server and the second server. The generation unit, in response to receiving a first request for recommending products of a first operator from the first server, generates an instruction for extracting recommended products based on common users for the first server, and generates an instruction for extracting second similar users for common users for the second server related to a second service from which a common user associated with the same common user as the first service has been extracted. The communication unit receives an access destination for information on the recommended products from the first server, and transmits information on the relay destination for relaying to the access destination and the instruction for extracting the second similar users to the second server. In the second server, the providing unit may provide information on the relay destination as information on the recommended products in the first service to the second similar users.

[0007] In a recommendation system according to an embodiment of the present invention, the first server and the second server each include a storage unit that stores information on other operators for which a common user is registered as a common user. The management server further includes a generation unit that generates various instructions for the first server and the second server. The generation unit, in response to receiving from the first server a first request for recommending products of a first operator and information on the recommended products generated by the first server based on common users, generates an instruction for extracting second similar users for common users for the second server related to a second service with which the same common user as the first service is associated. The communication unit may transmit the information on the recommended products and the instruction for extracting the second similar users to the second server.

[0008] In a recommendation system according to an embodiment of the present disclosure, the management server may further include a reward setting unit that associates a predetermined reward with a common user.

[0009] In the second server of a recommendation system according to an embodiment of the present disclosure, the providing unit may provide the recommended products in the first service as advertisements to the communication terminals of the second similar users.

[0010] The first server of the recommendation system according to an embodiment of the present invention may transmit information on a first similar user on which a predetermined anonymization process has been performed, together with information on a recommended product, to a management server by means of a communication unit.

[0011] The first server of the recommendation system according to an embodiment of the present invention may transmit information on a recommended product on which a predetermined randomization process has been performed to a management server by means of a communication unit.

[0012] In a management server of a recommendation system according to an embodiment of the present disclosure, a reward setting unit may distribute the profit obtained by a first business operator between the first business operator and a second business operator when a recommended product is used in a first service by a second similar user who has been provided with the recommended product.

[0013] In a management server of a recommendation system according to an embodiment of the present invention, a setting unit may set, for a single common user, a plurality of services provided by a plurality of business operators respectively in association with each other.

[0014] A management server according to an embodiment of the present invention includes: a setting unit that sets a common user who uses both a first service and a second service different from the first service; and a communication unit that transmits information on a product on the first service to be recommended to a second similar user who is similar to the common user in the second service, based on the usage history of the common user in the second service, which is specified based on the usage history of a first similar user who is similar to the common user in the first service, which is specified based on the usage history of the common user in the first service.

[0015] A control method for a management server according to an embodiment of the present disclosure includes: a step of setting a common user who uses both a first service and a second service different from the first service by the management server; and based on the usage history of the common user in the first service, based on the usage history of a first similar user who is similar to the common user in the first service, and based on the usage history of the common user in the second service, sending information on products on the first service to be recommended to a second similar user who is similar to the common user in the second service.

[0016] A control program for a management server according to an embodiment of the present disclosure causes the management server to realize: a function of setting a common user who uses both a first service and a second service different from the first service; and a function of sending information on products on the first service to be recommended to a second similar user who is similar to the common user in the second service, based on the usage history of the common user in the first service, based on the usage history of a first similar user who is similar to the common user in the first service, and based on the usage history of the common user in the second service.

Brief Description of Drawings

[0017]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Mode for Carrying Out the Invention

[0018] Hereinafter, an embodiment of the invention according to the present disclosure (also referred to as the present invention) will be described with reference to the drawings. Note that the drawings are merely examples, and the present invention is not limited to those shown in the drawings. For example, the number of management servers (information processing devices), first operator servers (first servers), second operator servers (second servers), database servers, users, etc., data sets (tables), sequence diagrams, flowcharts, and display screens are merely examples, and the present invention is not limited thereto.

[0019] <System Configuration> FIG. 1 is a diagram showing a configuration example of a recommendation system according to an embodiment of the present invention. The recommendation system 600 may be an information processing system related to a recommendation service that can recommend its own products to users of other companies' services among a plurality of businesses. Note that the "business" may be a company, corporation, individual, group, etc. that provides its own products in exchange for consideration from users, and as long as it can acquire and accumulate the usage history of users, the services of the business are not limited to those provided on the Internet. For example, even in a form where a business sells products at a physical store, if it can acquire the products purchased by a user together with the user's information, the recommendation system according to an embodiment of the present invention may be used. For example, combinations of a business and its products may include a video distribution business and content, a clothing manufacturer and clothing, a bookstore and books, a restaurant and cuisine, an online shopping site and items that can be purchased at the site, etc., but are not limited thereto. Also, the "product" is not limited to tangible goods and may include so-called intangible "services". An example of the "service" may be a cleaning business or a transportation business (taxi, train, etc.). Hereinafter, the service of the first business may also be referred to as the "first service", and the service of the second business may also be referred to as the "second service".

[0020] The "usage history" is a record of a predetermined period during which a user has used the service of a business, and may include the purchase history of products such as the name of the purchased product and the purchase time, and the evaluation (rating) of the purchased product.

[0021] The recommendation system 600 may include a management server (information processing device) 100, a first business server (first server) 200P, a second business server (second server) 200Q, and a database server 400.

[0022] The first operator server 200P and the second operator server 200Q may each be a server (information processing device) that performs various arithmetic processes related to the first operator and a second operator different from the first operator. Here, the side that recommends its own products to users of services provided by other companies is defined as the first operator, and the side that receives recommendations for products of other companies is defined as the second operator. Note that in FIG. 1, only two operators are shown, but there may be two or more operators. Also, in an embodiment of the present invention, the hardware configurations and functional configurations of the first operator server 200P and the second operator server 200Q may be the same and they may be able to recommend each other's products. However, for the sake of convenience in explanation, they are distinguished here. Hereinafter, when there is no need to distinguish between the first operator server 200P and the second operator server 200Q, they are also referred to as "operator server 200". Note that various processes in the case where the first operator or the second operator provides services on the Internet, such as content distribution and e-commerce platform services, may be executed by a service server (not shown) different from the operator server 200.

[0023] The management server 100 may be a server on the provider side that provides a recommendation service realized by the recommendation system 600 according to an embodiment of the present invention. The management server 100 may perform various arithmetic processes in realizing the recommendation system 600 in which recommendations are mutually made among a plurality of operators.

[0024] The management server 100 may be connected to the first operator server 200P and the second operator server 200Q via the network 500. The network 500 may include a wireless network or a wired network. Specifically, for example, the network 500 may be a wireless LAN (WLAN), a wide area network (WAN), ISDNs (integrated service digital networks), wireless LANs, CDMA (code division multiple access), LTE (long term evolution), LTE-Advanced, 4G (4th generation communication), 5G (5th generation communication), and a mobile communication system such as 6G (6th generation communication) and later. Also, the network 500 may be a combination of these.

[0025] In addition, in FIG. 1, the management server 100, the first operator server 200P, and the second operator server 200Q are shown individually, but it is not limited to this. That is, each function described as being provided in the management server 100, the first operator server 200P, and the second operator server 200Q may be realized by a plurality of servers. Also, the management server 100, the first operator server 200P, and the second operator server 200Q may be, for example, a distributed server system that cooperates by communicating via a network, or a so-called cloud server. That is, the management server 100, the first operator server 200P, and the second operator server 200Q are not limited to physical servers and may include virtual servers by software.

[0026] The database server 400 may store (memorize) various data (information) used by the management server 100 in the recommendation system 600. In FIG. 1, only one database server 400 is shown separately from the management server 100, but the database server 400 may be integrated with the management server 100. That is, the database server 400 may be the volatile memory or non-volatile memory of the management server 100. Further, the database server 400 may be composed of a plurality of storage devices. The database server 400 may be connected to the management server 100 via a dedicated internal network different from the network 500, or may be connected to the management server 100 via the network 500.

[0027] <Overview of the Recommendation System> Here, with continued reference to FIG. 1, an overview of the recommendation system 600 according to an embodiment of the present invention will be described. Hereinafter, the first operator is also referred to as "Company P" and the second operator is also referred to as "Company Q". Company P and Company Q may each have users who use their own services. In the example of FIG. 1, the users of Company P may be the set 300P including Users 10 and 30, and the users of Company Q may be the set 300Q including Users 20 and 30. As shown in FIG. 1, the sets 300P and 300Q may include User 30 who belongs to both sets. User 30 is a user who is registered as a user in both the services of Company P and Company Q and satisfies a predetermined condition described later, and is hereinafter referred to as a "common user". Further, although details will be described later, the similar users extracted on the first operator side correspond to "first similar users", and the similar users extracted on the second operator side correspond to "second similar users".

[0028] Here, the predetermined conditions related to the common user 30 may refer to the disclosure to a third party that the common user 30 is a user of both Company P and Company Q and has been approved by the common user 30. For example, the common user 30 may be a user who has limitedly approved the disclosure of personal information. For example, the common user 30 may be a user who has permitted a provider of a recommendation service, which is a third party different from Company P and Company Q, to handle their personal information. Thereby, the management server 100 may be able to create a database of FIG. 3 described later in which the common user corresponds to which user in the services of Company P and Company Q is stored. Note that, even if a user is registered as a user in both the services of Company P and Company Q, if they have not permitted the disclosure of their personal information, they do not correspond to the "common user" here. Here, the personal information that the common user has approved to disclose is information that can identify that the user is a common user in Company P or Company Q, and may be, for example, an account name, email address, phone number, name, date of birth, gender, etc. used at the time of registration for each service, but is not limited thereto. Note that the approval of the common user will be described later.

[0029] According to the recommendation system 600 according to an embodiment of the present invention, by using common users existing across a plurality of different services, products of the company's own service may be recommended to users of other companies' services. At this time, user-based collaborative filtering may be used for extracting products to be recommended. More specifically, among the set 300P of users of Company P, a first similar user similar to the common user 30 in the usage tendency of the service of Company P may be extracted by the first operator server 200P. Also, among the set 300Q of users of Company Q, a second similar user similar to the common user 30 in the usage tendency of the service of Company Q may be extracted by the second operator server 200Q. The management server 100 may obtain information regarding recommended products of Company P set based on the usage history of the service of Company P by the first similar user from the first operator server 200P. Then, the management server 100 may notify the second operator server 200Q of the information regarding the recommended products of Company P as recommended products to the second similar user who is a user of the service of Company Q.

[0030] As described above, in the recommendation system according to an embodiment of the present application, recommendations for the services of Company P may be made based on the information accumulated in the services of Company Q (that is, the usage history of Company Q's services by common users). In this way, the method of making recommendations for the target service (here, the services of Company P) using the history accumulated in a service different from the target service (here, the services of Company Q) is defined as cross-domain recommendation (CDR).

[0031] Thus, in the recommendation system according to an embodiment of the present invention, the information transmitted from one operator to another operator is only the information identifying which user of the services of Company P and Company Q the common user is, specified by the information that the common user has pre-approved for disclosure, and the information regarding the recommended products. Therefore, the present application can make accurate recommendations without conflicting with the personal information protection law while using the CDR method.

[0032] <Database> Here, the database referred to by the management server 100 will be described. FIG. 3 may be an example of a table regarding common users stored in the database server 400. The common user information table TB10 may store, in association with a common user ID (IDentifier) (an example of an identifier) for uniquely identifying a common user, the services registered as the common user and the identification information in the services. For example, a user identified by the common user ID "com_01" may be registered with the identification information "***_P03" in the service of Company P, the identification information "***_Q15" in the service of Company Q, and the identification information "***_S40" in the service of Company S. Also, a user identified by the common user ID "com_02" may be registered with the identification information "***_R10" in the service of Company R and the identification information "***_S02" in the service of Company S. Note that the figure is an example, and the format of the identification information is not limited to the illustrated one as long as the common user can be uniquely identified in the services of each company, and the table may be configured in any manner.

[0033] Next, although the details will be described later, with reference to FIG. 2, the hardware configuration and functional configuration of the operator server 200 and the management server 100 according to an embodiment of the present invention will be briefly described.

[0034] <Operator Server> In FIG. 2, the hardware configuration and functional configuration of the first operator server 200P and the second operator server 200Q are the same, and it may be on the side of recommending its own products or on the side of being recommended other companies' products. Therefore, in FIG. 2, only the first operator server 200P is shown as a block diagram of the operator server 200. Also, hereinafter, the operator related to the operator server 200 will be referred to as "own company", and the service provided by the operator will be referred to as "own company's service".

[0035] (1) Hardware Configuration of Operator Server The operator server 200 may include a control unit 210, a communication unit 220, and a storage unit 270.

[0036] The control unit 210 is typically a processor, including a central processing unit (CPU), a micro processing unit (MPU), a graphics processing unit (GPU), a microprocessor, a processor core, a multiprocessor, an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), etc., and may be implemented by a logic circuit (hardware) formed on an integrated circuit (IC (Integrated Circuit) chip, LSI (Large Scale Integration)) or a dedicated circuit.

[0037] The storage unit 270 may store various programs and various data required for the operation of the operator server 200. The storage unit 170 may include, for example, a hard disk drive (HDD), a solid state drive (SSD), a flash memory, etc. Further, the storage unit 270 may include a memory that provides a working area for the control unit 210.

[0038] The communication unit 220 may be implemented as hardware such as a network adapter, communication software, or a combination thereof. The communication unit 220 may transmit and receive various data to and from the management server 100 via the network 500. For example, when the communication unit 220 recommends the company's products to users of another company's service, the communication unit 220 may transmit information about products that similar users in the company's service, who are similar to the common users between the company and the other company, have received offers for using the company's service, to the management server 100. That is, when recommending the company's products, the communication unit corresponds to the "first communication unit". Also, when the communication unit 220 is recommended products of another company, the communication unit 220 may receive information about the recommended products of the other company from the management server 100. That is, when recommended products of another company, the communication unit 220 corresponds to the "second communication unit". Although not shown in FIG. 2, the communication unit 220 of the operator server 200 may be communicable with a communication terminal owned by a user registered in the company's service via the network 500.

[0039] (2) Functional Configuration of Operator Server The operator server 200 may include an extraction unit 211, a provision unit 212, and an acquisition unit 213 as functions realized by the control unit 210. In FIG. 2, functional units that are not essential for the embodiments described hereinafter may not be provided. Also, the functions or processes of each functional unit may be realized by machine learning or AI (Artificial Intelligence) within the achievable range. Note that a part of the various processes described hereinafter as being executed by the operator server 200 may be executed by the management server 100.

[0040] The extraction unit 211 may extract similar users among the users of the company's service who are similar to the common users in terms of the usage tendency of the company's service, based on the usage history of the company's service by users who use the company's service, including common users. The "usage tendency of the service" refers to the preferences and characteristics of users when using the service, and may be, for example, the preference for products to be purchased. Further, the extraction unit 211 may extract products that the similar users have received offers for using the company's service.

[0041] When receiving a recommendation from another company, the providing unit 212 may provide information on recommended products of the other company's service to similar users of the common users of the own company and the other company among the users of the own company's service, based on information on products provided by users of the other company using the other company's service.

[0042] The acquisition unit 213 may acquire various data from the management server 100, a user's communication terminal (not shown), and a service server via the network 500. For example, the acquisition unit 213 may acquire the usage history of a user from the service server.

[0043] <Management Server> Next, the hardware configuration and functional configuration of the management server 100 according to an embodiment of the present invention will be described.

[0044] (1) Hardware Configuration of the Management Server The management server 100 may include a control unit 110, a communication unit 120, and a storage unit 170.

[0045] The control unit 110 is typically a processor and may include a central processing unit (CPU), MPU, GPU, microprocessor, processor core, multiprocessor, ASIC, FPGA, etc., and may be realized by a logic circuit (hardware) formed on an integrated circuit (IC chip, LSI) or a dedicated circuit.

[0046] The storage unit 170 may store various programs and various data required for the operation of the management server 100. The storage unit 170 may include, for example, an HDD, SSD, flash memory, etc. Further, the storage unit 170 may include a memory that provides a working area for the control unit 110.

[0047] The communication unit 120 may be implemented as hardware such as a network adapter, communication software, or a combination thereof. The communication unit 120 may transmit and receive various data to and from the first operator server 200P and the second operator server 200Q via the network 500. For example, the communication unit 120 may transmit information regarding the common user 30 to the first operator server 200P and the second operator server 200Q. Note that the management server 100 may be communicable with the communication terminals owned by the users 10 and 20 and the common user 30 via the network 500.

[0048] (2) Functional Configuration of Management Server The management server 100 may include a setting unit 111, an extraction unit 112, a generation unit 113, an acquisition unit 114, and a reward setting unit 115 as functions realized by the control unit 110. In FIG. 2, functional units that are not essential for the embodiments described hereinafter may not be provided. Also, the functions or processes of each functional unit may be realized by machine learning or AI within a realizable range.

[0049] The setting unit 111 may receive settings of the common user 30 who uses both the first service and the second service. The extraction unit 112 may extract other services in which the common user exists for one service based on the common user information table TB10.

[0050] The generation unit 113 may generate various instructions for the operator server 200. For example, in response to receiving a request for recommending products of the first operator from the first operator server 200P, the generation unit 113 may generate an instruction for the first operator server 200P to extract similar users similar to common users in the usage tendency of the first operator's services. When the first operator is Company P, in FIG. 3, the common users are the common users with the common user IDs "com_01", "com_03", and "com_04" extracted by the extraction unit 112. Therefore, the generation unit 113 may instruct the first operator's service to extract similar users for the users registered with the identification information "***_P03", "***_P50", and "***_P07" respectively. That is, assuming that the same common users as those of Company P's service are extracted and associated, for example, for the second operator server 200Q related to Company Q's service, the generation unit 113 may generate an instruction to extract similar users for the users registered with the identification information "***_Q15" and "***_Q20" in Company Q's service. Also, assuming that the same common users as those of Company P's service are extracted and associated, for example, for the operator server (not shown) related to Company R's service, the generation unit 113 may generate an instruction to extract similar users for the users registered with the identification information "***_R30" and "***_R20" in Company R's service.

[0051] The acquisition unit 114 may acquire various data from the first operator server 200P, the second operator server 200Q, and the communication terminals of users (not shown) via the network 500.

[0052] So far, the example has been described by taking the mutual CDR between two operators as an example, but the number of operators is not limited to 2, and the number of operators may be 3 or more.

[0053] Here, when performing CDR mutually based on a common user, the range of operators used by the common user becomes the range in which CDR can be performed mutually. Usually, one user mainly uses the services of the operators they prefer, such as Company A for a shopping site, Company B for a travel site, Company C for a messaging service, and Company D for a communication carrier. Therefore, the number of services used by one common user is limited. Thus, as the number of operators performing CDR mutually increases, it is assumed that more common users will be required accordingly. A common user is a user who has consented, by agreeing to the terms of use or privacy policy, etc., to disclose their usage tendency in one operator to other operators. For a user, it requires a certain amount of resources to read, understand, and agree to the content of the terms of use, etc., so they are not always positive about agreeing to the terms of use of new services. Therefore, when a large number of common users are required, it is assumed that it may be difficult to secure sufficient common users.

[0054] In contrast, the reward setting unit 115 may associate a predetermined reward with the common user set by the setting unit 111. The predetermined reward may be, for example, something with a monetary value that can be used in the services registered by the common user. Thereby, it is possible to present a return for information provision to common user candidates, and increase the number of common users who allow information provision that can be a psychological hurdle. Also, since this enables securing a large number of common users, it becomes possible to increase the variety of operators used by common users, and more other recommended operators can be set.

[0055] <Sequence and Screen Example> The functions of the above-described management server 100 and operator server 200 will be described in detail based on the sequence diagram and screen examples. FIG. 4 shows the sequence between the management server 100 and operator servers 200 (first operator server 200P and second operator server 200Q) according to an embodiment of the present invention. FIGS. 5 and 7 show the screen displays of the user's communication terminal, and FIG. 6 may be an example of a schematic explanatory diagram of the extraction of similar users. Hereinafter, when distinguishing between the side where Company P recommends its own products and the side where Company Q is recommended Company P's products, and between the recommending side and the recommended side, explanations will be given with P and Q attached to the reference signs.

[0056] The setting unit 111 of the management server 100 may receive the setting of a common user (step P10). The setting of the common user may be, for example, a notification of whether to act as a common user is sent to the communication terminal of a user who uses either the first service or the second service, and may be set according to the response from the user to the notification. The screen D10 in FIG. 5 may be an example in which a notification M10 asking about the intention to act as a common user is displayed on the user's communication terminal. Note that the notification M10 may be generated by the generation unit 113 of the management server 100 and transmitted to the user's communication terminal via the operator server 200, or may be transmitted from the management server 100 to the user's communication terminal. When the user selects the "View Details" button on the screen D10, a screen for selecting a service to register as a common user may be displayed, and the user's identification information may be input. Alternatively, when the user becomes a common user in a service that has already been registered, a screen for inputting the user's identification information in the registered service may be displayed. The identification information input by the user may be transmitted to the management server 100. Note that information regarding the common user may be stored in the database server 400 as the common user information table TB10 in FIG. 3 (step P11).

[0057] The first operator server 200P may send a request for recommending its own products to the management server 100 (step P12). At this time, the first operator server 200P may include the access destination (URL) to the information on its recommended products in the request. Although details will be described later, in one embodiment, the information on the recommended products of Company P may be provided to the users of Company Q's service using various tags such as img tags in HTML and techniques such as iframes (inline frames). Therefore, the access destination to the information on the recommended products may be the URL of the site where the information on the recommended products of Company P can be viewed.

[0058] The extraction unit 112 of the management server 100 may refer to the common user information table TB10 to extract the operators in which the same common users as Company P exist. Here, Company Q in which common users with the same common user IDs "com_01" and "com_03" as Company P exist may be extracted. The identification information in Company P's service associated with those common users of Company P and Company Q may be sent to the first operator server 200P as the information of the common users. In the example of FIG. 3, it may be sent to the first operator server 200P that the users registered with the identification information "***_P03" and "***_P07" in Company P's service are common users (step P13). At this time, the generation unit 113 of the management server 100 may generate an extraction instruction for similar users of the common users in Company P's service for the products provided using Company P's service and send it to the first operator server 200P.

[0059] The extraction unit 211P of the first operator server 200P may extract similar users of common users received from the management server 100, and based on the similar users, extract recommended products (step P14). Here, the extraction of similar users will be described with reference to FIG. 6. The tables TB21 and TB31 in FIG. 6 may be rating tables in which ratings for products provided by each company are recorded as service usage histories of users of the services of Company P and Company Q, respectively. According to an embodiment of the present invention, the extraction of similar users may be performed on each of the side that recommends its own products and the side that receives product recommendations. Here, for the user with the common user ID "com_04" who is a common user of Company P and Company Q, it will be described that similar users are extracted by each of Company P and Company Q. That is, in the first operator server 200P of Company P, the extraction unit 211P may use the rating table TB21 of Company P to extract similar users of user P7 who is a common user from the users of the service of its own company. Also, in the second operator server 200Q of Company Q, the extraction unit 211Q may use the rating table TB22 of Company Q to extract similar users of user Q20 who is a common user from the users of the service of its own company.

[0060] Note that a recommendation algorithm in the recommendation system according to an embodiment of the present invention may use collaborative filtering, content-based filtering, or the like, and a method for extracting similar users in these algorithms may be an existing one. For example, for calculating the similarity between users, a Pearson correlation coefficient, a Spearman rank shift coefficient, a Kendall's τ shift coefficient, a cosine similarity, a Euclidean distance, a Hamming distance, or the like may be used according to the content of the usage history and the like. In addition, for calculating the similarity, a Jaccard coefficient, a Dice coefficient, a Simpson coefficient, or the like may be used.

[0061] The extraction unit 211P may calculate the similarity between user P7 and other users using the ratings in the table TB21. The similarity may be, for example, the Pearson correlation coefficient between users. The extraction unit 211P may extract users with a similarity equal to or greater than a predetermined threshold as similar users. In the example of FIG. 6, it is assumed that users P1, P2, etc. shown in the table TB22 are extracted as similar users. Note that, regardless of the similarity threshold, a predetermined number of similar users may be extracted in descending order of similarity.

[0062] The extraction unit 211P may further extract, as recommended products, the products (purchased products) that users P1 and P2, who are extracted as similar users, received services using the services of Company P. For example, the extraction unit 211P may extract, as recommended products, those with high ratings among the products purchased by users P1 and P2 extracted as similar users. In the example of FIG. 6, product PB rated "4.0" by user P1 and product PD rated "5.0" by user P2 are set as recommended products. Note that the number of recommended products is not limited to one for each similar user, and multiple recommended products may be set. Also, when the usage history indicates the presence or absence of purchase instead of a rating, the products most recently purchased by similar users may be extracted as recommended products.

[0063] The first operator server 200P may set information regarding the recommended products at the above-described access destination (step P15). That is, it may be set so that information regarding the recommended products is displayed by selecting the access destination. Then, the generation unit 113 of the management server 100 may transmit to the second operator server 200Q an extraction instruction for similar users of the common users associated with both Company P and Company Q and identified by the identification information in the services of Company Q, and a relay destination to the access destination of the recommended products (step P16). Here, the relay destination may be a URL for redirecting to the site regarding the recommended products of Company P described above (details will be described later).

[0064] In the second operator server 200Q, the extraction unit 211Q may extract similar users of common users (step P17). Further, the providing unit 212Q may transmit relay destination information to the extracted similar users (step P18).

[0065] These will be described with reference to FIG. 6. The extraction unit 211Q may calculate the similarity between the user Q20 and other users using the scores in the table TB31 and extract similar users. In the example of FIG. 6, it is assumed that the users Q2 and Q21 shown in the table TB32 are extracted as similar users. The providing unit 212Q may provide information regarding the recommended products PB and PD of Company P to the users Q2 and Q21 who are similar users in the service of Company Q. At this time, the information provided to the users Q2 and Q21 may be a web page or a notification in which the relay destination URL generated by the management server 100 is embedded as a redirect destination. That is, according to an embodiment of the present invention, before accessing the site for the recommended product of Company P, the redirect destination of the management server 100 exists as a relay. Therefore, the access destination of Company P is not transmitted to Company Q, and Company Q cannot recognize that the product recommended to its own users is a product of Company P. Therefore, the information of the operator in which the common user is registered can be kept secret.

[0066] FIG. 7 shows an example of a screen of a communication terminal of a similar user in the service of Company Q on which the recommended product of Company P is displayed. As shown in the screen D20, the information regarding the recommended product may be provided, for example, by a pop-up M20, but is not limited thereto. For example, it may be provided by email, text message, or the like.

[0067] Thus, according to an embodiment of the present invention, it may be possible to recommend own products among operators having common users. At this time, the information transmitted from one operator to another operator is only the information of which user the common user is in the services of one operator company and another operator, specified by the information that the common user has pre - promised to disclose. Therefore, accurate recommendations can be made without violating the Personal Information Protection Law.

[0068] In addition, the providing unit 212Q may display recommended products of other companies on the communication terminals of similar users in the form of advertisements. The form of advertisement may refer to being displayed on an application installed on the user's communication terminal to provide the company's own service. Thereby, the recommended products can be surely notified to the users.

[0069] Also, the reward setting unit 115 may distribute the profit obtained by the first operator when the recommended product is used in the first service by the similar user who has been provided with the recommended product, between the first operator and the second operator. That is, when a user of Q company's service purchases a product of P company in response to being recommended the product of P company, the profit obtained by P company thereby may be distributed to Q company. This may be realized by, for example, tracking or the like. Thereby, a fair recommendation system can be provided among the operators who use it.

[0070] <Flowchart of the management server> FIG. 8 shows an example of the control flow of the management server 100 according to an embodiment of the present invention. First, the setting unit 111 may set a common user who uses both the first service and a second service different from the first service (step T11). Next, based on the usage history of the common user in the second service, the communication unit 120 may transmit information on products on the first service to be recommended to the second similar users who are similar to the common user in the second service, based on the usage history of the first similar users who are similar to the common user in the first service, specified based on the usage history of the common user in the first service (step T12).

[0071] <Other embodiments> In the above, the cases where Company P and Company Q do not have information on their common users have been described. However, according to an embodiment of the present invention, information on common users among business operators may be stored, and each business operator may hold information on its own common users and information on other companies where they are registered. Therefore, in response to the request of a business operator who desires to recommend its own products, similar users can be extracted and recommended to users who may lead to the purchase of its own products.

[0072] This will be described with reference to FIGS. 9 and 10. In FIG. 9, the recommendation of Company P's products to Company Q is in a mode where it is made in response to a recommendation request for Company P's own products by Company P, and in FIG. 10, the recommendation of Company P's products to Company Q is in a mode where it is made in response to a recommendation request for products of other companies by Company Q.

[0073] A description will be given of FIG. 9. Since steps S10 and S11 are the same as steps P10 and P11 in FIG. 4, the description thereof will be omitted. The management server 100 may transmit information regarding common users to the first server 200P and the second server 200Q (step S12). Note that the information regarding common users transmitted here may refer to, for each company, information that can identify common users existing in the company and information regarding other company services in which the common users are registered as common users. For example, in the example of FIG. 3, for the first server 200P, it may be transmitted that the user registered with the identification information "***_P03" is a common user between "Company Q and Company S", the user registered with the identification information "***_P50" is a common user between "Company R", and the user registered with the identification information "***_P07" is a common user between "Company Q and Company R". Also, for the second server 200Q, it may be transmitted that the user registered with the identification information "***_Q15" is a common user between "Company P", the user registered with the identification information "***_Q20" is a common user between "Company P and Company R", and the user registered with the identification information "***_Q18" is a common user between "Company R and Company S". The first server 200P and the second server 200Q may each store information regarding their respective common users (steps S13 and S14).

[0074] The extraction unit 211P of the first operator server 200P may extract similar users of the common users received from the management server 100 and extract recommended products based on the similar users (step S15). Since these extractions are the same as step P14 in FIG. 4, the description thereof will be omitted.

[0075] The first operator server 200P may transmit information regarding a recommended product to the management server 100 as a recommendation request for its own products (step S16). The generation unit 113 of the management server 100 may transmit an extraction instruction for similar users of a common user identified by identification information in the service of Company Q, which is associated with the common users of Company P and Company Q, and information regarding the recommended product, to the second operator server 200Q (step S17). In the second operator server 200Q, the extraction unit 211Q may extract similar users of the common user and provide information regarding the recommended product in the service of Company P to the extracted similar users (step S18).

[0076] As described above, according to an embodiment of the present invention, it may be possible to recommend its own products between operators having common users. At this time, the information transmitted from one operator to another operator is only information as to which user in the services of one operator company and another operator the common user is, specified by information that the common user has previously permitted disclosure, and information regarding the recommended product. Therefore, accurate recommendations can be made without violating the Personal Information Protection Law.

[0077] Next, FIG. 10 will be described. Note that steps T10 to T14 are the same as steps S10 to S14 in FIG. 9, and thus the description thereof will be omitted. The second operator server 200Q may transmit a recommendation request for products of another company to the management server 100 (step T15). The generation unit 113 of the management server 100 may transmit an extraction instruction for similar users of a common user identified by identification information in the service of Company P, which is associated with the common users of Company P and Company Q, and the recommended product, to the first operator server 200P (step T16). In the first operator server 200P, the extraction unit 211P may extract similar users of the common user and extract the recommended product in the service of Company P (step T17). The communication unit 220P may transmit information regarding the extracted recommended product to the management server 100 (step T18).

[0078] The generation unit 113 of the management server 100 may transmit to the second operator server 200Q an instruction to extract similar users of a common user associated with common users of Company P and Company Q, which are identified by identification information in the service of Company Q, and information regarding recommended products (step T19). In the second operator server 200Q, the extraction unit 211Q may extract similar users of the common user and provide information regarding recommended products in the service of Company P to the extracted similar users (step T20).

[0079] Although the present invention has been described based on the drawings and examples, it should be noted that those skilled in the art can easily make various modifications and changes based on this disclosure. Therefore, it should be noted that these modifications and changes are included in the scope of the present invention. For example, the functions included in each component, each step, etc. can be rearranged so as not to be logically contradictory, and a plurality of components, steps, etc. can be combined into one or divided. Also, the configurations shown in the above embodiments may be combined as appropriate. For example, each component described as being provided in the management server 100 may be realized by being distributed among a plurality of servers. Also, each function described as being provided in the management server 100 may be executed by the operator server 200.

[0080] For example, in the above, only the aspect of providing information on recommended products to other companies has been described. However, after performing a predetermined anonymization process on the information of each company, information other than recommended products may be provided to other companies. For example, when providing the recommended products of Company P to Company Q, the information of the similar users used for extracting the recommended products may be transmitted to Company Q after a predetermined anonymization process. The information of the similar users may be, for example, the age, gender, preferences, etc. of the similar users. Also, the predetermined anonymization process may be, for example, k-anonymization, l-diversity, t-closeness, etc. Thereby, on the Company Q side, similar users closer to the similar users of Company P can be extracted, and the success rate of the recommendation can be increased.

[0081] Also, as further anonymization to prevent the identification of personal information, even for products not purchased by similar users, processing such that they are extracted as recommended products may be performed. For example, in the first operator server 200P, after changing the numerical information of the purchase history by random numbers, extraction of recommended products may be performed. That is, based on the concept of differential privacy, a predetermined randomization process may be performed on the recommended products to enhance the confidentiality of the data. In addition, the anonymization process may include processing of inserting data about virtual users (adding noise) or processing of excluding some data (sampling), etc. Thereby, the anonymity of the data exchanged between operators can be improved.

[0082] Also, in another embodiment, the allocation of recommended products to similar users extracted by the second operator who has obtained information about recommended products from the first operator may be set on the second operator side. Also, according to the number of recommended products obtained from the first operator, the number of similar users extracted by the second operator may be set. For example, the same number of similar users as the number of recommended products may be extracted, and one recommended product may be recommended to only one person. Alternatively, more similar users than the number of recommended products may be extracted, and one recommended product may be recommended to a plurality of similar users.

[0083] Also, in another embodiment, when an operator owns information about a common user, the operator may be able to obtain information about other operators registered as common users according to the permission of the common user. Thus, when an operator itself owns information about other operators who are registration destinations for a common user, product recommendations may be executed between operators without relying on the management server.

[0084] Furthermore, the management server 100 may obtain the number of similar users extracted by the first operator and the second operator from the first operator server 200P and the second operator server 200Q, respectively, and determine whether a recommendation is possible. For example, when the number of similar users extracted on the second operator side is significantly less than the number of similar users extracted on the first operator side, it may be determined that the recommendation is not established.

[0085] Also, the first service provider side may be able to select common users who are purchasing the products to be recommended. For example, based on the information of the common users obtained from the management server 100 in step S13 of FIG. 9 or step T13 of FIG. 10, the first service provider server 200P may be able to select common users.

[0086] The programs of the embodiments of the present disclosure may be provided in a state stored in a storage medium readable by an information processing apparatus. The storage medium can store the program in a "non-transitory tangible medium". The program includes, for example, software programs and control programs. When each functional unit of the management server 100 is realized by software, the management server 100 functions as the control unit 110, the setting unit 111, the extraction unit 112, the generation unit 113, the acquisition unit 114, and the reward setting unit 115 by executing the program loaded on the memory by the processor.

[0087] When appropriate, the storage medium can include one or more semiconductor-based or other integrated circuits (ICs) (e.g., field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), etc.), hard disk drives (HDDs), hybrid hard drives (HHDs), optical disks, optical disk drives (ODDs), magneto-optical disks, magneto-optical drives, floppy disks, floppy disk drives (FDDs), magnetic tapes, solid state drives (SSDs), RAM drives, secure digital cards or drives, any other appropriate storage medium, or any suitable combination of two or more of these. When appropriate, the storage medium may be volatile, non-volatile, or a combination of volatile and non-volatile.

[0088] Also, the program of the present disclosure may be provided to the management server 100 via any transmission medium capable of transmitting the program (such as a communication network or a broadcast wave).

[0089] In addition, each embodiment of the present disclosure can also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission. Note that the program of the present disclosure may be implemented using, for example, script languages such as JavaScript (registered trademark), Python, C language, Go language, Swift, Koltin, Java (registered trademark), etc.

[0090] According to each aspect of the present disclosure described above, by providing more opportunities for appropriate products to be recommended to users and for purchases to be made, it is possible to contribute to the achievement of Goal 8, "Decent Work and Economic Growth," of the Sustainable Development Goals (SDGs).

Description of Reference Numerals

[0091] 100 Management Server (Information Processing Device) 110 Control Unit 111 Setting Unit 112 Extraction Unit 113 Generation Unit 114 Acquisition Unit 115 Reward Setting Unit 120 Communication Unit 170 Storage Unit 200 Operator Server 210 Control Unit 211 Extraction Unit 212 Provision Unit 213 Acquisition Unit 300 User Set 400 Database Server 500 Network 600 Recommendation System 10, 20 Users 30 Common Users

Claims

1. A recommendation system for recommending a product provided by a first service, which is a service of a first operator, to a user who receives the provision of a product of a second operator by using a second service, which is a service of a second operator, comprising a management server, a first server related to a first operator, and a second server related to a second operator, each having a communication unit for transmitting and receiving data to and from each other, wherein the management server comprises a setting unit that receives settings of common users who use both the first service and the second service, transmits information about the common users to the first server and the second server through the communication unit, wherein the first server comprises a first extraction unit that extracts, based on the usage history of the first service by first users who use the first service and include the common users, first similar users among the first users who are similar to the common users in the usage tendency of the first service, and extracts, as recommended products, the products received by the first similar users using the first service, wherein the second server comprises a second extraction unit that extracts, based on the usage history of the second service by second users who use the second service and include the common users, second similar users among the second users who are similar to the common users in the usage tendency of the second service, and a providing unit that provides information about the recommended products in the first service to the second similar users, and is a recommendation system.

2. The management server further comprises an instruction generation unit that generates various instructions for the first server and the second server, wherein, in response to receiving a first request for recommending a product of the first operator from the first server, the generation unit generates an extraction instruction for the recommended products based on the common users for the first server, and generates an extraction instruction for second similar users for the common users for the second server related to the second service extracted assuming that the same common users as those related to the first service are associated, the communication unit receives an access destination for information about the recommended products from the first server and transmits, to the second server, information about a relay destination for relaying to the access destination and the extraction instruction for the second similar users, in the second server, The providing unit provides information regarding the relay destination to the second similar user as information regarding the recommended product in the first service. The recommendation system according to claim 1.

3. The first server and the second server each include a storage unit that stores information regarding other business operators in which the common user is registered as a common user. The management server further includes a generation unit that generates various instructions for the first server and the second server. In response to receiving, from the first server, a first request for recommending products of the first business operator and information regarding the recommended products generated by the first server based on the common user, the generation unit generates an extraction instruction for the second similar user for the common user to the second server related to the second service associated with the same common user as the first service. The communication unit transmits the information regarding the recommended product and the extraction instruction for the second similar user to the second server. The recommendation system according to claim 1.

4. The management server further includes a reward setting unit that associates a predetermined reward with the common user. The recommendation system according to any one of claims 1 to 3.

5. In the second server, the providing unit provides the recommended product in the first service to the communication terminal of the second similar user as an advertisement. The recommendation system according to any one of claims 1 to 3.

6. The first server transmits information regarding the first similar user subjected to predetermined anonymization processing to the management server by the communication unit together with the information regarding the recommended product. The recommendation system according to claim 1.

7. The first server transmits information regarding the recommended product subjected to predetermined randomization processing to the management server by the communication unit. The recommendation system according to claim 1.

8. In the management server, the reward setting unit distributes the profit obtained by the first business operator when the recommended product is used in the first service by the second similar user who has been provided with the recommended product between the first business operator and the second business operator. The recommendation system according to claim 4.

9. In the management server, the setting unit associates and sets a plurality of services provided by a plurality of business operators respectively for one common user. The recommendation system according to any one of claims 1 to 3.

10. A setting unit that sets a common user who uses both a first service and a second service different from the first service, Based on the usage history of the first similar user who is similar to the common user in the first service, which is specified based on the usage history of the common user in the first service, and based on the usage history of the common user in the second service, for the second similar user who is similar to the common user in the second service, a communication unit that transmits information about the product on the first service to be recommended. A management server comprising:

11. In an information processing apparatus, A step of setting a common user who uses both a first service and a second service different from the first service, Based on the usage history of the first similar user who is similar to the common user in the first service, which is specified based on the usage history of the common user in the first service, and based on the usage history of the common user in the second service, a step of transmitting information about the product on the first service to be recommended to the second similar user who is similar to the common user in the second service. A control method for a management server that causes the above to be executed.

12. In an information processing apparatus, A function of setting a common user who uses both a first service and a second service different from the first service, Based on the usage history of the first similar user who is similar to the common user in the first service, which is specified based on the usage history of the common user in the first service, and based on the usage history of the common user in the second service, a function of transmitting information about the product on the first service to be recommended to the second similar user who is similar to the common user in the second service. A control program for a management server that realizes the above.

Images