JP2008282353A - Recommendation device and recommendation method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide service information depending on a user, and prevent privacy information leakage. <P>SOLUTION: A recommendation device 10 first calculates recommendation basic information in which service attributes is associated with a user attribute model representing the attributes of a user model presumed to desire the service attributes, based on the use history information representing a history of a service used by a plurality of users, which associates a plurality of user attributes representing user attributes with a plurality of service attributes representing the attributes of service selected by the user; and then calculates a service attribute model representing the attributes of a service model presumed to be desired by a requesting user, based on the recommendation basic information calculated by the recommendation basic information calculating means and the user attributes of the requesting user. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a recommendation apparatus and a recommendation method for providing service information corresponding to a user to a user who requests service information.

  Services handled in this type of technology include, for example, a service that provides food in a restaurant, a service that delivers a pot set, and the like. This type of technology is also applied when introducing various recommended products such as books and content.

For example, there is a technology that, when a history of content that the user has viewed in the past is transmitted, analyzes the content and provides content information that matches the user (see Patent Document 1).
JP 2003-173350 A

  By the way, for the user, information related to the user's taste such as a history of contents viewed in the past is privacy information. Therefore, there is a problem because privacy information is leaked when information on the user's preferences leaks in a state where the user can be identified.

  In the technique disclosed in Patent Document 1, the history of content transmitted by the user is information related to the user's preferences. When the user transmits information on the user's preferences, the information on the user's preferences may leak in a state where the user can be specified.

  Therefore, an object of the present invention is to provide a recommendation device and a recommendation method capable of providing service information corresponding to a user and preventing leakage of privacy information.

  The recommendation device according to the present invention is (1) usage history information indicating usage histories of services used by a plurality of users in a recommendation device that provides service information corresponding to a requesting user to a requesting user requesting service information. Service usage history storage means for storing usage history information in which a plurality of user attributes indicating user characteristics and a plurality of service attributes indicating the characteristics of the service selected by the user are associated, and (2) service usage history Based on the usage history information stored in the storage means, the recommended basic information calculation that calculates the recommended basic information in which the service attribute and the user attribute model indicating the feature of the user model estimated to desire the service attribute are associated with each other. Means, (3) recommended basic information calculated by recommended basic information calculating means, and A user attribute of the user, based on the request user, characterized in that it comprises a service attribute estimating means for calculating a service attribute model indicating a feature of a service model that is estimated to desired.

  A recommendation method according to the present invention is a recommendation method for providing service information corresponding to a requesting user to a requesting user requesting service information, wherein the recommended basic information calculating means stores usage histories of services used by a plurality of users. Based on usage history information in which a plurality of user attributes indicating user characteristics and a plurality of service attributes indicating service characteristics selected by the user are associated with each other. A recommended basic information calculating step for calculating recommended basic information in association with a user attribute model indicating the characteristics of the user model estimated to have an attribute, and a recommendation calculated by the service attribute estimating means in the recommended basic information calculating step. Based on the basic information and the user attributes of the requesting user, the requesting user The is characterized in that and a service attribute estimating step of calculating a service attribute model indicating a feature of a service model that is estimated to desired.

  According to the present invention, it is estimated that the requesting user desires based on the user history of a plurality of users who have used the service in the past, the usage history information that associates the service attributes of the service, and the user attribute of the requesting user Service attributes are calculated. In this way, service information indicating service attributes suitable for the requesting user can be provided by using the usage history information of a plurality of users. In addition, even when information on the use of the service by the requesting user is not included in the use history information, service information suitable for the requesting user can be provided. Therefore, service information can be provided even when the requesting user uses the service for the first time. Furthermore, the usage history information, the recommended basic information, and the service attribute model handled in the recommendation device are not privacy information of the requesting user. As described above, since the privacy information directly related to the requesting user is not handled in the recommendation device, leakage of privacy information can be prevented.

  Preferably, the recommendation device includes (4) service listing means for listing services in ascending order of difference between the service attribute model calculated by the service attribute estimation means and the service attribute. In this case, the service can be listed up and presented to the requesting user.

  Preferably, the recommendation device includes (5) a transmission unit that transmits the service list listed by the service list-up unit to a communication terminal used by the requesting user, and (6) a request from the service list transmitted by the transmission unit. Receiving means for receiving service specifying information indicating the service selected by the user, and the transmitting means indicates the service specifying information to a production server that receives a service provision request indicated by the service specifying information received by the receiving means. Request information for requesting service provision is transmitted.

  In this case, the requesting user can order a recommended service via the recommendation device without transmitting privacy information to the producer server. Therefore, since it is not necessary to pass privacy information related to the requesting user's preference to the service provider side, leakage of privacy information can be reliably prevented.

  ADVANTAGE OF THE INVENTION According to this invention, while providing the service information according to a user, the leakage of privacy information can be prevented.

  An embodiment of a recommendation device and a recommendation method of the present invention will be described with reference to the drawings.

  FIG. 1 is a diagram showing an outline of a recommendation system including a recommendation device according to the present embodiment. The recommendation system 1 is a system for introducing a service recommended to the user 2 and placing an order for the service selected by the user 2. The service is, for example, a service for providing food in a restaurant or the like, a service for delivering a pot set, or the like.

  The recommendation system 1 includes a communication terminal 3, a management server 5, a plurality of producer servers 7, and a delivery server 9. The communication terminal 3, the management server 5, the plurality of manufacturer servers 7, and the delivery server 9 are configured to be able to transmit and receive information to and from each other via a wireless network.

  Of the devices provided in the recommendation system 1, the communication terminal 3 is a terminal used by the user 2. The management server 5 is arranged in the management center 4 of the recommendation system 1 and manages information in the recommendation system 1. The producer server 7 is arranged in each producer 6 that provides the service. The producer 6 is, for example, a producer of a product to be delivered, a market, a restaurant, or the like.

  The delivery server 9 is an apparatus used by a delivery company that delivers a home delivery service to the user 2. The recommendation device 10 according to the present embodiment is a device that is mounted on the management server 5 and provides recommended service information suitable for the user in response to a request from the user 2. The recommended service information is, for example, a list of information on restaurant course meals and home delivery hot pot sets.

  Here, an outline of the recommendation system 1 will be described. First, the user 2 (hereinafter referred to as “requesting user”) uses the communication terminal 3 to request recommended service information from the recommendation device 10 of the management center 4. In response to this request, the recommendation device 10 transmits service information listing recommended services to the communication terminal 3 (step S11; transmission step).

  The requesting user 2 selects a service from the service information received by the communication terminal 3 and sends service specifying information for specifying the service to the recommendation device 10 to place an order for the service. In response to the transmission of the service specifying information by the communication terminal 3, the recommendation device 10 receives the service specifying information. (Step S12; reception step).

  The recommendation device 10 orders service from the producer 6 by sending request information for requesting provision of the service indicated by the service identification information to the producer server 7 (step S13; request information sending step). In response to ordering of this service, the manufacturer 6 provides the service to the requesting user 2. For example, the course meal selected by the user 2 is provided at a restaurant that is the producer 6.

  On the other hand, when the service selected by the requesting user 2 is a home delivery service, the recommendation device 10 transmits delivery request information to the requesting user 2 to request delivery of the product to the delivery server 9, thereby Is arranged for delivery (step S14). Then, the delivery company 8 physically delivers the product produced by the producer 6 from the producer 7 to the requesting user 2 (steps S15 and S16).

  As described above, the recommendation system 1 provides service information corresponding to the requesting user 2 and places an order for the service selected by the requesting user 2. Therefore, it is possible to provide a service corresponding to the requesting user 2 without disclosing the privacy information regarding the taste of the requesting user 2 to the manufacturer 6 and the delivery company 8.

  Next, the recommendation device 10 according to the present embodiment will be described in detail. FIG. 2 is a diagram for explaining functional blocks of the recommendation device according to the present embodiment. The recommendation device 10 includes a communication unit (transmission unit, reception unit) 11, a calculation unit 12, and a storage unit 13.

  The communication unit 11 transmits / receives information to / from the communication terminal 3, the manufacturer server 7, and the delivery server 9 via the wireless network N. The calculation unit 12 performs information processing calculation based on the data stored in the storage unit 13.

  The calculation unit 12 includes a user management unit 14, a service management unit 15, a service usage history management unit 16, a recommended basic information calculation unit (recommendation basic information calculation unit) 17, a service attribute estimation unit (service attribute estimation unit) 18, and a service list. An up part (service list up means) 19 and a request information generation part 20 are provided. The storage unit 13 includes a user master database 21, a service master database 22, a service usage history database (service usage history storage unit) 23, and a recommended basic information database 24. Hereafter, each function which the calculating part 12 has and the data memorize | stored in the memory | storage part 13 are demonstrated in detail.

  The user management unit 14 uses the user master database 21 to securely manage information regarding a plurality of registered users. The information about the user is information such as a user identifier ID, user authentication information PW, name, telephone number, and address.

  The user management unit 14 stores a plurality of user attributes in the user master database 21 as information about users. FIG. 3 is a table showing user attributes stored in the user master database of FIG. As shown in FIG. 3, the user master database 21 stores a user attribute identifier c (c = 1, 2,... C), a user name, and a plurality of user attributes in association with each other.

The plurality of user attributes are information indicating user characteristics, and are information indicating V items such as the user's age, the user's annual income, and the user's gender, respectively. Features for one user is indicated by the user attribute vector x _c to V-number of the user attribute elements. That is, the user attribute vector x _c is represented by a row vector R ^{1 × V} as shown in the equation (1). The user attribute identifier c is uniquely identifies the user attribute vectors x _c.

Specifically, as shown in FIG. 3, the characteristic of “user 1” identified by the user attribute identifier 1 is a user attribute having user attributes such as age “25 years old” and annual income “3 million” as elements. It is stored as a vector x ₁ = (25 million,...).

The user attribute matrix X indicating the user attribute vector x _c for each of C users is represented by a C × V matrix as shown in Equation (2). Note that each element of the matrix used in the present embodiment is a real number.

  When the communication unit 11 receives information requesting registration from the user, the user management unit 14 issues a user identifier ID and user authentication information PW to perform user registration. In addition, the user management unit 14 acquires a plurality of user attributes, names, telephone numbers, addresses, and information on the user's family from the communication terminal 3 at the time of user registration, and together with the user identifier ID and the user authentication information PW 21.

  The service management unit 15 uses the service master database 22 to securely manage information regarding the manufacturer 6 and information regarding services provided by the manufacturer 6. The service management unit 15 stores service attributes in the service master database 22 as information about services. FIG. 4 is a table showing service attributes stored in the service master database of FIG.

As shown in FIG. 4, the service master database 22 stores service attribute identifiers s (s = 1, 2,..., S), service names, and service attributes in association with each other. The service attribute is information indicating the characteristics of the service, and is information indicating E items such as “all-you-can-drink”, “night view”, and “fee”. The characteristics of one service are indicated by a service attribute vector y _s whose elements are E service attributes. That is, the service attribute vector y _s is represented by a row vector R ^{1 × E} as shown in Expression (3). The service attribute identifier s are uniquely identifies the service attribute vector y _s.

As shown in FIG. 4, for the service attribute, for example, for the “all-you-can-drink” item, information “1” indicating flag ON when all-you-can-drink is set, and flag OFF when all-you-can-drink is not set. The indicated information “0” is set. For example, for “charge”, information “1” indicating flag ON is set in any of the items “0 to less than 3000 yen”, “3000 to less than 5000 yen”, and “5000 yen or more”. The Specifically, the characteristics of “winter taste gourmet course” identified by service attribute identifier 1 include service attributes such as “all-you-can-drink”, “no night view”, and a fee of 4000 yen. As a service attribute vector y ₁ = (1, 0,..., 0, 10).

サービス管理部１５は、サービスを提供する生産元６からの登録要求に応じて、サービス属性の識別子を払出し、そのサービス属性の識別子と関連付けて、サービス名及びサービス属性ベクトルｙ_ｓをサービスマスタデータベース２２に格納する。 Service attribute matrix Y indicating the service attribute vector y _s for each S-number service, as shown in Equation (4), indicated by a matrix of S × E.

Service management unit 15, in response to the registration request from the production source 6 for providing services, payout the identifier of the service attributes, in association with the identifier of the service attribute, the service name and the service attribute vector y _s service master database 22 To store.

The service usage history management unit 16 uses the service usage history database 23 to manage usage history information indicating usage histories of services used by a plurality of registered users. FIG. 5 is a table showing usage history information stored in the service usage history database of FIG. As shown in FIG. 5, the service usage history database 23, a user attribute vector x _c of the user, it is stored in association with each other and the service attribute vector y _s services that the user has used, the.

Information indicating one use history is represented by a row vector R ^{1 × (V + E)} obtained by combining the user attribute vector x _c and the service attribute vector y _s as shown in the equation (5). For example, in FIG. 5, information x ₃₅ y ₁₅ indicating the first usage history indicates that the user indicated by the user attribute vector x ₃₅ has used the service indicated by the service attribute vector y ₁₅ .

The usage history information H indicating the usage history of N times is represented by an N × (V + E) matrix, as shown in Expression (5).

Service usage history management unit 16, the communication unit 11, the service specifying information sent from the communication terminal 3 is received, it extracts the user attribute vector x _c of the corresponding user from the user master database 21, the service identification information extracting a service attribute vector y _s of service indicated. Then, the service usage history management unit 16 generates information x _c y _s indicating one usage history and updates the usage history information H.

  The recommended basic information calculation unit 17 calculates recommended basic information Θ using the usage history information H stored in the service usage history database 23. The recommended basic information Θ is information in which a service attribute is associated with a user attribute model indicating a feature of the user model estimated to be desired for the service attribute.

Ｅ個のサービス属性それぞれについてのユーザ属性モデルθ_ｅを示すリコメンド基礎情報Θは、式（７）に示すように、Ｖ×Ｅの行列によって示される。

The user attribute model θ _e indicating the characteristics of the user model estimated to desire the e-th service attribute among the E service attributes is represented by a vector having V user attributes as elements. That is, the user attribute model θ _e is represented by a column vector R ^{V × 1} as shown in Expression (6).

Recommendation basic information Θ indicating the user attribute model theta _e for each E-number of service attributes, as shown in equation (7), represented by a matrix of V × E.

なお、式（８）においてＢ_ｅは、ｅ番目のサービス属性にフラグがあるユーザの人数を示す。また、式（８）において、Ｘ|_yｅ=1は、ｅ番目のサービス属性にフラグがあるＢ_ｅ人のユーザ属性ベクトルを含むＢ_ｅ×Ｖの行列である。また、式（８）において、上付き文字「Ｔ」は、転置行列である旨を示す。 The recommended basic information calculation unit 17 first calculates the user attribute model θ _e for each service attribute using the usage history information H, and generates recommended basic information using the user attribute model θ _e . The user attribute model θ _e of the e-th service attribute is represented by a vector whose element is an average value of each user attribute indicating the characteristics of the user with a flag in the e-th service attribute, as shown in Expression (8). It is.

Incidentally, B _e in equation (8) indicates the number of users there is a flag e th service attributes. In Equation (8), X | _{ye = 1} is a B _e × V matrix including the user attribute vector of B _e who has a flag in the e th service attribute. In the equation (8), the superscript “T” indicates that it is a transposed matrix.

  The recommended basic information calculation unit 17 updates the recommended basic information every time the usage history information is updated by the service usage history management unit 16.

Service attribute estimating unit 18 uses the user attribute vector x _c of recommendation basic information Θ and requesting user 2, requesting user 2 calculates a service attribute model y _M indicating a feature of a service model that is estimated to desired. Service attribute model _{y M} is calculated by the equation (9).

Service attribute model y _M is calculated based on the use history of services by features and a plurality of user requests user 2. Service attribute model y _M is indicated by the service attribute vector.

When the service attribute estimation unit 18 receives the user identifier ID of the requesting user 2 together with information that the communication unit 11 requests service information recommended from the requesting user 2, the service attribute estimation unit 18 requests the user master database 21 based on the user identifier ID. It extracts the user attribute vector x _c of the user 2. Then, the service attribute estimation unit 18 performs the calculation represented by the equation (9), and outputs information indicating the service attribute model y _M obtained as a result of the calculation to the service list up unit 19.

As illustrated in FIG. 6, the service list-up unit 19 lists services in the order in which the service attribute model y _M calculated by the service attribute estimation unit 18 and the service attribute vector y _s are relatively similar. FIG. 6 is a diagram showing service information provided in the recommendation device according to the present embodiment.

Services listed unit 19 extracts each service attribute vector y _s stored in the service master database 22, the difference between the respective service attribute vector y _s and service attribute model y _{M |} y _M -y s | calculated To do. Then, the difference _| y _M -y s | is small order, listing the service name and the service identifier in the order in which that is similar.

Specifically, as shown in FIG. 6, the service list-up unit 19 sets a predetermined number of “mochi nabe course”, “fish course”, and “wild vegetable hot pot” in order from the smallest difference | y _M −y _s |. The service name is listed. The service list-up unit 19 outputs the listed service information to the communication unit 11. The communication unit 11 transmits service information to the communication terminal 3 and receives service specifying information returned from the communication terminal 3 in response to the transmission.

  The request information generation unit 20 generates request information for requesting provision of the service indicated by the service specifying information. Then, the request information generation unit 20 outputs the address of the service producer server 7 indicated by the service identification information together with the request information to the communication unit 11. As a result, the request information is transmitted from the communication unit 11 to the service producer server 7 indicated by the service identification information, and an order for the service is performed.

  Each function of the recommendation device 10 described above is realized by the hardware resources shown in FIG. FIG. 7 is a hardware configuration diagram of the recommendation device according to the present embodiment. As shown in FIG. 7, the recommendation device 10 physically includes a CPU 101, a RAM 102 and a ROM 103 that are main storage devices, an auxiliary storage device 105 such as a hard disk, an input device 106 such as a keyboard and mouse that are input devices, and a display. The computer system includes an output device 107 such as a network module and a communication module 104 which is a data transmission / reception device such as a network card.

  The calculation function of the calculation unit 12 of the recommendation device 10 is performed by the CPU 101. The storage function of the storage unit 13 is performed by the RAM 102 and the auxiliary storage device 105. Each function of the recommendation device 10 described with reference to FIG. 2 includes a communication module 104 and an input device 106 under the control of the CPU 101 by reading predetermined computer software on hardware such as the CPU 101 and the RAM 102 shown in FIG. This is realized by operating the output device 107 and reading and writing data in the RAM 102 and the auxiliary storage device 105.

  Returning to FIG. 2, the communication terminal 3 will be described. The communication terminal 3 includes a communication unit 31, a display unit 32, and an input unit 33. The communication unit 31 transmits and receives information to and from the recommendation device 10. For example, the communication unit 31 performs transmission of information for requesting recommended service information, reception of service information, transmission of service specifying information, and the like.

  The display unit 32 is configured with a display or the like, and displays service information received by the communication unit 31. The input unit 33 includes a keyboard, a mouse, a scanner, an OCR, and the like, and accepts a user's selection from service information displayed on the display unit 32.

  Subsequently, the operation of the recommendation device 10 according to the present embodiment will be described with reference to FIGS. 8 to 10, and the recommendation method according to the present embodiment will be described. First, user registration processing in the recommendation device 10 will be described. FIG. 8 is a flowchart showing user registration processing in the recommendation device according to the present embodiment.

First, information about the user including user attribute vector x _c is transmitted from the communication terminal 3 to the recommendation apparatus 10 together with information requesting registration is received by the communication unit 11 of the recommendation apparatus 10 (step S21). Then, information about the user is stored in the user master database 21 by the user management unit 14, and the user attribute identifier c is paid out (step S22).

  Subsequently, the user identifier ID and the user authentication information PW are paid out by the user management unit 14 and transmitted to the communication terminal 3 by the communication unit 11 (step S23). In this way, user registration processing is performed. The user identifier ID and the user authentication information PW are used to protect a user's privacy and authenticate the user and prevent impersonation by a third party.

  Next, service information generation processing in the recommendation device 10 will be described. FIG. 9 is a flowchart showing service information generation processing in the recommendation device according to the present embodiment. First, a user inputs a user identifier ID and user authentication information PW to the communication terminal 3 together with information indicating that an information request is made. This input may be made automatically by opening the foldable communication terminal 3, or may be made automatically when the communication terminal 3 is turned on, or automatically every predetermined time. May be made. In this case, information can be requested without the user being aware of it. When the user is input, the user identifier ID and the user authentication information PW are transmitted from the communication terminal 3 and received by the communication unit 11 of the recommendation device 10, and a request for service information is accepted (step S31).

The user attribute identifier c corresponding to the received user identifier ID is extracted from the user master database 21 by the service attribute estimation unit 18 (step S32). Subsequently, the user attribute vector _xc is extracted from the user master database 21 by the service attribute estimation unit 18 (step S33). The requested service attribute model user 2 is estimated to be desired y _M, based on the user attribute vector x _c and recommendation basic information theta, is calculated by the service attribute estimating unit 18 (step S34; service attribute estimating step) .

Services in the order in which they resemble relative to the service attribute model y _M is listed by the service list 19, the service information is generated (step S35; listed steps). And the service information according to a request | requirement user is transmitted to the communication terminal 3 by the communication part 11 (step S36). In this way, recommended service information corresponding to the requesting user 2 is provided to the requesting user.

  Next, the update process of the recommended basic information Θ in the recommendation device 10 will be described. FIG. 10 is a flowchart showing the update process of the recommended basic information in the recommendation device according to the present embodiment.

  First, the requesting user 2 selects a favorite service from the service list indicated by the service information, inputs the service attribute identifier s to the communication terminal 3, and transmits it to the recommendation device 10. In the recommendation device 10, the service attribute identifier s is received by the communication unit 11 as the service specifying information, and the service selected by the user is accepted (step S41).

Then, the user attribute vector x _c corresponding to the requesting user 2 is extracted from the user master database 21 by the service usage history management unit 16, and the service attribute vector y _s corresponding to the service attribute identifier s is extracted from the service master database 22. It is extracted by the service usage history management unit 16. Then, the history information x _c y _s of service is generated by the service utilization history management unit 16, is registered in the service usage history database 23 (step S42).

  When the service usage history database 23 is updated, the recommended basic information Θ is newly calculated based on the usage history information H, and the recommended basic information Θ is updated (step S43). In this way, the recommended basic information is updated every time the user uses the service, and the latest service information can always be provided to the user.

According to the recommendation device 10 according to the present embodiment described above, the user history of a plurality of users who have used the service in the past, the usage history information H that associates the plurality of service attributes of the service, and the user attributes of the requesting user 2 and the vector x _c, based on the request user 2 service attribute model y _M that is estimated to be desired is calculated. Accordingly, service information indicating recommended service attributes suitable for the requesting user 2 can be provided by using use history information of a plurality of users.

  Further, even when the information on the use of the service by the requesting user 2 is not included in the usage history information H, service information suitable for the requesting user 2 can be provided. That is, even for a new user who has never received a service, as long as a user attribute vector indicating the user's characteristics can be acquired, a service attribute that the user would desire can be estimated. Furthermore, even a new service that has never been used can recommend a service having a service attribute similar to that of a service that is frequently used.

Furthermore, the usage history information H, the recommended basic information Θ, and the service attribute model y _M handled in the recommendation device 10 are not privacy information of the requesting user 2. Thus, since the recommendation apparatus 10 does not handle privacy information that is directly related to the requesting user 2, leakage of privacy information can be prevented.

  Further, the recommendation device 10 transmits request information for requesting the provision of the service indicated by the service identification information to the producer server 7 that receives the service provision request indicated by the service identification information received from the communication terminal 3. Accordingly, the requesting user 2 can order a recommended service via the recommendation device 10 without transmitting privacy information to the producer server 7. Therefore, it is not necessary to pass privacy information related to the preferences of the requesting user 2 to the service provider side, so that leakage of privacy information can be reliably prevented.

  The privacy information stored in the user master database 21, the service master database 22, and the service usage history database 23 is managed safely. In the recommendation apparatus 10, the delicate recommendation which utilized privacy information on it can be implement | achieved.

  The recommendation device 10 of the present invention is not limited to the above embodiment, and various modifications are possible. For example, the user may use the family user attribute using the family information option. A plurality of service attributes that the user's family would like can be estimated. Thereby, the user can acquire the latest recommended service information using his / her information, family information, and other person's information.

  In this case, information about the user's family is stored in the user master database 21 in association with the user ID. And as shown in FIG. 11, you may notify the service which can be recommended to a user's family with service information. FIG. 11 is a diagram illustrating an example of service information displayed on the communication terminal.

  The recommendation system 1 according to the present embodiment may be configured to manage information from provision of recommended service information to service ordering and delivery in combination with an existing escrow service or the like. In addition, according to the user's request, the service attribute estimation unit 18 may use the past recommended basic information calculated using the service usage history information used during the predetermined period, and the service attribute that the user may desire. A model may be estimated.

  In the above embodiment, the recommended service information is transmitted to the communication terminal 3 used by the user. However, the means for notifying the service information may be an e-mail, a phone call from an operator, a postcard You may also use a means of mailing letters or letters. In addition, transmission / reception of information between the communication terminal 3 and the recommendation device 10 may be performed by encryption. Further, the recommended basic information Θ may be encrypted and stored in the recommended basic information database 24. Further, the recommended basic information before the update may be managed / stored as a separate file (secured by encryption or the like) together with the update date and time.

  Further, the service information to be notified to the user may be notified together with advantageous information such as discounts in delivery aggregation, accurate manufacturer display, and the like. A target to be recommended by the recommendation device 10 according to the present embodiment may provide information on a product recommended as a present to be given to a father, for example. In the above embodiment, user registration is performed in advance and user attributes are registered in advance. However, the present invention is not limited to this. When a user requests service information, information about the user including user attributes may be transmitted.

  Moreover, in the said embodiment, although numerical values, such as age "25" and annual income "3 million", were used as a user attribute, it is not restricted to this. For example, numerical values such as age “25” and annual income “3 million” may be normalized so as to take a value of 0 to 1. In addition, for example, when the annual income is less than 3 million, “1” is used, when it is 3 million or more and less than 5 million, “2” is used, and when it is 5 million or more, “3” is used. Also good.

It is a figure which shows the outline | summary of the recommendation system containing the recommendation apparatus which concerns on this embodiment. It is a figure for demonstrating the functional block of the recommendation apparatus which concerns on this embodiment. It is a table | surface which shows the user attribute stored in the user master database of FIG. It is a table | surface which shows the service attribute stored in the service master database of FIG. 3 is a table showing usage history information stored in the service usage history database of FIG. 2. It is a figure which shows the service information provided in the recommendation apparatus which concerns on this embodiment. It is a hardware block diagram of the recommendation apparatus which concerns on this embodiment. It is a flowchart which shows the user registration process in the recommendation apparatus which concerns on this embodiment. It is a flowchart which shows the production | generation process of the service information in the recommendation apparatus which concerns on this embodiment. It is a flowchart which shows the update process of the recommendation basic information in the recommendation apparatus which concerns on this embodiment. It is a figure which shows the example of the service information displayed on a communication terminal.

Explanation of symbols

2 ... requesting user, 3 ... communication terminal, 7 ... producer server, 10 ... recommendation device, 11 ... communication unit (transmission means, reception means), 17 ... recommendation basic information calculation unit (recommendation basic information calculation means), 18 ... Service attribute estimator (service attribute estimator), 19... Service list-up unit (service list-up unit), 23... Service usage history database (usage history storage unit), H... Usage history information, x _c . Plural user attributes), y _M ... service attribute model, y _s ... service attribute vector (plural service attributes), Θ ... recommended basic information, θ _e ... user attribute model.

Claims (4)

In a recommendation device that provides service information corresponding to the requesting user to a requesting user requesting service information,
Usage history information indicating usage history of services used by a plurality of users, wherein a plurality of user attributes indicating the characteristics of the user and a plurality of service attributes indicating the characteristics of the service selected by the user are associated with each other Service usage history storage means for storing usage history information;
Based on the usage history information stored in the service usage history storage means, the recommended basic information in which the service attribute is associated with the user attribute model indicating the characteristics of the user model estimated to be desired for the service attribute is calculated. Recommend basic information calculation means,
A service attribute for calculating a service attribute model indicating a feature of the service model estimated to be desired by the requesting user based on the recommended basic information calculated by the recommended basic information calculating unit and the user attribute of the requesting user An estimation means;
A recommendation device comprising:   The recommendation apparatus according to claim 1, further comprising: a service listing unit that lists a service in ascending order of difference between the service attribute model calculated by the service attribute estimating unit and the service attribute. Transmitting means for transmitting the service list listed by the service list-up means to a communication terminal used by the requesting user;
Receiving means for receiving service specifying information indicating the service selected by the requesting user from the service list transmitted by the transmitting means;
The transmitting means transmits request information for requesting provision of a service indicated by the service specifying information to a production server that receives a service providing request indicated by the service specifying information received by the receiving means. The recommendation device according to claim 2. A recommendation method for providing service information corresponding to a requesting user to a requesting user requesting service information,
The recommended basic information calculation means is usage history information indicating usage histories of services used by a plurality of users, and includes a plurality of user attributes indicating the characteristics of the users and a plurality of characteristics indicating the characteristics of the services selected by the users. Based on usage history information that associates a service attribute, a recommendation basis that calculates recommendation basis information that associates the service attribute with a user attribute model that indicates the characteristics of the user model that is estimated to be desired for the service attribute. An information calculation step;
A service attribute indicating characteristics of a service model estimated to be desired by the requesting user based on the recommended basic information calculated in the recommended basic information calculating step and the user attribute of the requesting user. A service attribute estimation step for calculating a model;
A recommendation method characterized by comprising:

Images