JP2019003594A - Gift items ai offer recommendation device - Google Patents

Abstract

To enable even gift-item recipients and gift-item keeper to optimally deal with gift items with less burdens.SOLUTION: A gift item AI offer recommendation device 10 has: a user database 11 that includes contact information capable of getting in contact with users; a reception item database 12 that changeably registers rights to receive gift items for each user; a weight network parameter 15 that has an updatable weight space including a relationship between individual commercial items; gift item registration means 22 that registers the right to receive the gift item about the commercial item designated by a recipient based on a client designation in a recipient record in the reception item database 12; notification means 23 that, when a right to receive a new gift item is added to the recipient record of the reception item database 12, uses preliminarily registered contact information to notify the recipient of the added right; and commercial item offer means 21 that offers the commercial items to be output by the weight network parameter 15.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a proposal recommendation device and service based on AI analysis of gifts.

  In giving gifts, it is always a problem for gift givers (hereinafter referred to as clients) and gift recipients (hereinafter referred to as recipients) who should receive what products. It is difficult to accurately grasp the other party's actual wishes, and the received goods often do not meet the recipient's wishes. On the other hand, it is desirable that gift gifts be continued in order to facilitate human relations, and it is desirable that gifts be appropriately returned. For this reason, in various systems in the gift gift field, various studies have been conducted to reduce the burden on the client and recipient.

  Until now, the proposal and recommendation (recommendation) of products in gifts provided by product providers such as gift products by computers has been performed based on the past purchase history of the client. In other words, there is no way to know the gift that the recipient wants. In addition, it is easy to forget important event dates such as the date and time of past ceremonial occasions, birthdays of others, wedding anniversaries, and death dates. In addition, when displaying products suggested or recommended by product providers on catalogs such as printed materials or on e-commerce sites on the Internet, HTML or other means like e-mail magazines are used to attach photos or image images and navigate to URLs. The same information is generally distributed in a broadcast manner.

  It is a taboo at a ceremonial occasion that a “client” and a “recipient” interact to decide on a gift. In addition, the “product provider” has specialized knowledge such as customs and ceremonial occasions, but there is no way to collect and propose a dialogue between the three parties. For this reason, at present, the client selects a product and presents it unilaterally. Nowadays, the same gifts and miscellaneous goods are received and sold at thrift stores, etc., and are also sold at online auctions. It is desirable to improve this situation as much as possible.

  Patent Document 1 proposes a delivery management apparatus that has a return candidate database for storing and managing return product candidate groups suitable for return for each price range and presents a return candidate list. It is linked to a database of delivered products, and proposes general products corresponding to the delivered products as contents and price ranges.

  Patent Document 2 proposes a system in which a gift recipient side selects a plurality of products desired by the recipient and notifies the gift recipient via the network of the product information and gives all or part of the gift information. ing.

  In Patent Document 3, if the name of the recipient is registered in the delivery server that has received the notification to deliver the gift, the gift information is transmitted to the recipient's terminal before the delivery is actually delivered to the recipient. However, a system for processing gifts in response to processing requests related to gifts received from user terminals has been proposed. The recipient can confirm what kind of product has been sent before actually receiving it, and if it is judged unnecessary, it will give a gift by any of the methods of delivery, transfer, hold / exchange, auction listing without actual delivery. The product can be processed. Gifts can be used effectively without delivery. The delivery server opens a home page for conducting an auction on the web, and presents gifts at the auction in response to a request from the recipient.

JP 2003-256732 A JP2003-30485 Japanese Patent Laid-Open No. 2002-342428

  However, since the return candidate list proposed in the system of Patent Document 1 simply proposes a general list, it cannot be said that it is moderately optimized for each gift situation.

  In the system of Patent Document 2, there is no problem for the gifting side, and it is useful for a request for sending, but there is a problem that it is too direct in a gift that facilitates human relations such as Chugen and year-end gifts.

  In the system of Patent Document 3, the delivery company goes to the store that manages the store terminal that received the delivery request, collects the gift, and delivers the gift to the recipient, the forwarding destination, or the successful bidder of the auction. Although the recipient's effort can be saved, the actual product itself had to be moved, and the burden on the management side was insufficient.

  In addition, current product proposals and recommendations (recommendations) are mainly analyzed from purchase history, but artificial intelligence can lead to changes in merchandising plans, event planning, gifting, etc. Not done.

  SUMMARY OF THE INVENTION Therefore, an object of the present invention is to make it possible to handle gifts appropriately and with less burden on both the person receiving the gifts and the person managing the gifts. Eventually, personal information (personal attribute information) for each person, attribute information for each product, gift use / event information, contract information, operation history, etc. are linked to AI, in addition to proposals for gifts, etc. It is a further object to be able to produce, manufacture, and propose sales of new products and their assortments, or new products themselves.

This invention
A user database containing contact information that can be contacted to identifiable users;
Received goods database that registers the right to receive gifts for each of the above users,
A weight network parameter that constitutes a weight space set for each product;
A gift registration means for registering the right to receive a gift for the product designated by the recipient designated by the client, designated by the terminal, in the recipient's record in the receipt database, and receiving the receipt database When a right to receive a new gift is added to a person's record, a notification means for notifying the recipient using the contact information registered in advance and output by the weight network parameter The above-mentioned problem has been solved by executing a gift AI proposal recommendation device having a product proposal means for presenting a product, or a product presentation service using them.

  The weight network parameters are used for proposing and recommending gifts, exchanges, return items, and related products. In the present invention, it is analogized that the information on the gift product preferred by the recipient actually needs the gift application using the “rule base” and “machine learning” which are generally referred to by the weight network parameter. The above-mentioned client can be notified in advance, and it is possible to propose a gift desired by the recipient. It also provides a function to record the date and time of the past ceremonial occasion, the birthday of another person, the wedding anniversary date, the date of death, etc. as supplementary information of the client or recipient.

  In the present invention, in order to analyze the information of “client”, “recipient”, and “product provider”, an intelligent process analysis is performed using a rule base and machine learning. Introduce the concept of "weight space" by defining multiple weights for products. The weight space value is used as a feature value in machine learning, and the selected product by the actual client and the processing result of the recipient (reception, gift exchange, gift transfer, hold, etc.) are used as teacher data. Proposal / recommendable functions that can be improved can be generated. As a result, for each product, the “counter information”, “number of purchases, receipts, etc.” of the “requester”, “recipient”, “product provider” or set by the “counter” Analysis based on the value, supplementary information of “client” and “receiver”, and customs related to general gifts, etc., to determine a specific gift application, the order of proposals of gift products for the client and receiver, and the like. Unnecessary items can be deleted, weights can be reduced, ranking can be lowered, or grouping or group ranking can be performed.

  As a result, the client can make proposals and recommendations for analysis based on event information such as ceremonial occasions that are used for gifts and attributes of gifts, and the products that the recipient wants to receive at the appropriate time for each use. It is possible to maximize the probability of receiving a gift.

  As an embodiment of the gift AI suggestion device according to the present invention, the product proposal means inputs gift information including a gift notified when the notification means notifies the recipient. In the case where it is executed, the following function added to the gift can be realized.

  The product presented by the product proposal means presents a gift product proposal form that suggests or recommends a gift to the recipient to the client, and a product suitable as a return from the recipient to the client A return product proposal form, a related product proposal form that presents a suitable product to be received by the recipient in addition to the product, and the one recipient does not receive the gift at the time of the gift notification, and the gift later A right exchange proposal form that presents a gift subject to the right of receipt of another recipient who wishes to exchange the right to receive the goods (hereinafter referred to as a right to receive).

  Further, the notified gift item can be exchanged for gift items such as companies A and B different from the product provider (gift product exchange) as shown in the conceptual diagram of FIG. 9 registered and provided by the product provider. Moreover, you may have a point conversion means which converts this goods into a monetary value and registers as a point which the said recipient can use. This point can be used as a consideration for the use of other services such as purchase of return merchandise, addition of monetary value at the time of exchange.

  Furthermore, in response to the notification by the notification means, the receipt of the gift from the recipient is suspended, and the right to receive at a later date is held and exchanged with the receipt right of other recipients (hereinafter referred to as exchange of receipt rights). Or a processing means for receiving an instruction to transfer the receiving right to a third party (hereinafter referred to as transferring the receiving right) and exchanging or transferring the product while omitting a part of the actual delivery operation.

  Furthermore, a user who has a specific individual desired product database for registering a product for which a user wishes to give a gift by designating a user who can be the client by the user who can be the recipient, and the product proposing means is a user who can be the client When a candidate for a gift to a user who can be a recipient is presented, the product designated by the pair of the client and the recipient in the specific individual desired product database is a designated product. If the request is presented without displaying, the client can select a gift for the recipient without fail. In other words, it is possible to implement a so-called “between” function for individuals with specific connections.

  In the present invention, it is possible to know the customs related to gifts, the past receipt history and preferences of the recipient, and the relationship between the client and the recipient, or the relationship between the three parties including the product provider. It is possible to recommend and propose gift products to “clients”, “recipients”, and “product providers” through detailed information analysis.

  On the other hand, for the recipient, for example, product information is recommended / suggested in the return by the above-mentioned return product proposal form or the ask function using the specific personal desired product database. In addition, for example, product trend information indicating what products are likely to be sold in the future can be provided to a product provider who contracts with a business provider that provides an apparatus and service according to the present invention. In addition, since it is possible to send information about gift products preferred by the recipient to the client who is presumed to actually need the gift application, the burden and effort that the client is worried about Reduce gifts that are inappropriate for the recipient.

  Furthermore, with regard to sales promotion and promotion advertisement of specific products for each person, it is possible to make accurate proposals / recommendations in products and sales promotion, and a contract effect is expected. For example, analyzing past transaction data of online shops, comparing the percentage of people who bought A and B together with the percentage of people who bought A and C together, and suggesting / recommending the higher percentage If so, it can be expected that the effectiveness of proposals and recommendations will increase. Analyzing such relationships and regularity from the data, find out the rules and patterns lurking there, suggest and recommend.

  If gifts can be interactively exchanged between the “client” and the “recipient”, or if the “product provider” can analyze the information, the necessary product is not unilateral. However, it becomes feasible to give and receive gifts when necessary at appropriate events, and it is possible to create necessary gifts that are not unilateral.

The block diagram which shows the structural example of the gift AI proposal recommendation apparatus concerning this invention The conceptual diagram which shows the example of execution form of the gift AI proposal recommendation apparatus concerning 1st embodiment The conceptual diagram which shows the example of an execution form of the gift AI proposal recommendation apparatus concerning 2nd embodiment Diagram showing standardization of weights Diagram showing weight space (A) Conceptual diagram showing group overlap in the third embodiment, (b) Conceptual diagram showing an example of a family relationship constituting the group in the third embodiment, (c) Group in the third embodiment Conceptual diagram showing an example of affiliation with a product Conceptual diagram showing an execution form example of the gift AI proposal recommendation device according to the third embodiment The conceptual diagram which shows the example of an execution form of the gift AI proposal recommendation apparatus concerning 4th embodiment The conceptual diagram which shows the gift exchange in the gift AI proposal recommendation apparatus concerning the application form of 4th embodiment (A) Conceptual diagram showing an execution form example of the gift AI proposal recommendation device according to the fifth embodiment, (b) A figure showing an example of the state of the received goods database before exchange in (a), (c) After exchange Diagram showing an example state (A) Conceptual diagram showing an execution form example of the gift AI proposal recommendation device according to the application form of the fifth embodiment, (b) A figure showing an example state of the received goods database before transfer in (a), (c) ) Diagram showing an example after transfer

  The present invention will be described below together with specific embodiments. The present invention is a gift AI proposal recommendation device capable of presenting and selecting a suitable product for a user who gives or receives a gift, and a service providing method for presenting the gift.

  The gift AI suggestion device according to the present invention may be a physical device configured by a single server or a plurality of servers existing on the network, or a database and the like described later on the cloud. It may be a virtual device realized. In any case, the gift AI proposal recommendation device according to the present invention is connected to the user terminal via the network. The service providing method according to the present invention provides a service including presentation of information related to a product to be a gift to the user's terminal according to the configuration requirements of the gift AI proposal recommendation device.

  The gift AI proposal recommendation device according to the present invention is connected via a network, and a user who provides a service includes both a client who wants to give a gift and a recipient who is given a gift. Since the client for a gift is a recipient in return, the user can be either. Hereinafter, the client and recipient without any particular designation are positions in each transaction, and both refer to users.

  In addition to the above-described users, the gift AI proposal recommendation device according to the present invention may be able to register a product provider. This product provider is a corporation such as a product development company, product sales company, etc., and has a business relationship with the service provider who operates the gift AI proposal recommendation device of the present invention, and provides products. It is possible to provide related information related to weighting, which will be described later, about the product.

  FIG. 1 shows a functional block diagram of an embodiment of the gift AI proposal recommendation device 10 according to the present invention. The gift AI proposal recommendation device 10 has a user database 11 for registering individual users. The information about the user includes an ID for identifying each user, authentication information such as a password, and contact information that can be contacted. The authentication information is not particularly limited as long as it is information that can be used for authentication, such as cooperation information with another SNS. The above contact information includes the address information for actual delivery of the gift or notification by post, and when the right to receive the gift is added mainly in the notification means described later, It is desirable to include address information such as an e-mail address, a mobile phone number, and an SNS ID. The contact information may be an e-mail address that also serves as a user ID.

  In addition, the user database 11 includes the gender, age, date of birth, blood type, current and past addresses, telephone numbers, past gift histories (giving side, receiving side or both), and past items of each user. It is desirable to include page browsing history. Furthermore, it is preferable that registration is possible as optional inputs, including birthdays, wedding dates, and other important dates meaningful to the user, such as anniversary dates, up to the year. In addition, when a user's family becomes a user, it is preferable that extended registration can be performed not only in record units but also in cell units so that shared anniversary and non-shared anniversary can be set. Both are useful information for presenting suitable products. These may be a single table, or may be a format in which a separate table is associated with each application and content.

  Furthermore, it is preferable that the user database 11 has human relation information of the user. Examples of the human relationship information include a family, a relative, a friend, a school relationship, a work relationship, a neighborhood relationship, and other relationships with a human who can assume a gift exchange with the user. It is desirable to register a person who is a related party as the user. This human relationship information includes the relationship between the user and the relationship, including the relationship, the history of the gifts and attributes that were exchanged in the past (including returns), the exchange of money exchanged in the past, and joint participation in the past. Information on events such as ceremonial occasions, etc., are not limited to these.

  Based on the information in the user database 11, a client database 11 a related to a client process when a gift occurs and a recipient database 11 b related to a receiver process when a gift occurs are included in the user database 11. May be associated with each other.

  Gift AI proposal recommendation device 10 is a product that can be actually given as a gift by the operator of the device or the provider of the service, or a contractor with them (hereinafter referred to as “operator”). Product database 13. Each product preferably has information such as price, usage, product genre, color, and other information as well as a product ID for identifying each product so that the relevance can be determined by the weight network parameter 15 described later.

  The gift AI proposal recommendation device 10 has a received product database 12 for registering a gift receiving right so as to be changed for each user. This right to receive is a right that allows the user to actually receive the gift from the operator of the gift AI proposal recommendation device 10, and basically the gift is received by another user (client). Processing on the system to give to the user (recipient) is performed and registered. This right to receive is not queued for immediate processing by the system to deliver gifts, but can be held until processing for which delivery by the recipient is desired, such as setting a certain period of time. Further, as will be described later, this receipt right can be transferred to another user for each gift or converted into points to be extinguished.

  Further, the gift AI proposal recommendation device 10 has a weight network parameter 15 having an updatable weight space including the relationship between individual products for the products registered in the product database 13. When the user tries to select a gift and a product to be handled as a derivation of the gift, the weight network parameter 15 is input with a variable about the situation to be selected, and a product to be presented as a suitable product. It can be output. It is desirable that more appropriate products can be presented in each case including not only the relationship between the products but also various elements described later.

  When the user selects a gift as a client, the gift AI suggestion recommendation device 10 selects a gift (returned gift) for a gift as a recipient, and the user selects a gift as a recipient. Various items related to the exchange of the gifts in the weight network parameter 15 when selecting or receiving a suitable product (related product) to be added at the time of receiving or presenting the product to other users. The product proposal means 21 for inputting the elements and attributes and presenting the product to be output as a suitable product is executed. This embodiment will be described later.

  The gift AI proposal recommendation device 10 is a gift registration means for registering the right to receive a gift for the product designated by the recipient designated by the client, instructed from the terminal, in the recipient's record in the receipt database 12. 22 is executed. Basically, the client terminal 31 who wants to give a gift to the recipient is registered by settlement of the gift of the product, but it is instructed by the terminal managed by the operator who receives the request of the client May be input. Also, in the case of a receiving right exchange function, which will be described later, the receiving rights of both parties are registered in one side and deleted on the other side so that two recipients can exchange each other by designating each other. In the case of a conventional gift system, this right of receipt is actually delivered by the operator as it is, but the gift AI proposal recommendation device 10 according to the present invention receives the final receipt from the recipient terminal 32. Until permission is received, the recipient has the right to receive the gift and will not be delivered until the actual product is finalized.

  The gift AI proposal recommendation device 10 receives the above-mentioned contact information of the recipient previously registered in the user database 11 when the receipt right to receive a new gift is added to the recipient record in the receipt database 12. The notification means 23 for notifying the recipient using the information is executed. With this notification, the recipient receives the notified gift immediately, actually holds it, holds it, converts it to a point described below, exchanges it for another gift, or forwards it to a third party You can decide what to do. Moreover, when receiving, the said product proposal means 21 shows the goods suitable to receive additionally with the goods.

  The gift AI proposal recommendation device 10 receives the gift from the recipient, receives it in exchange for another product, receives it, transfers it to a third party, or executes a delivery, or performs delivery. When there is no instruction for delivery (holding of receipt right), the right hold processing means 24 for recording the receipt right of the product in the recipient record of the received product database is executed. In terms of processing, it is not particularly necessary to change the data in the table, but it is preferable to be able to manage the hold state and receipt right exchange / receipt right transfer. By notifying a product provider who delivers a product that is a gift as an immediate cancellation, occurrence of useless delivery can be suppressed.

  The gift AI proposal recommendation device 10 may be able to execute the point conversion means 25 that converts the notified gift into a monetary value and registers it as a point that can be used by the recipient. This point is registered in the user database 11 or another database as a numerical value that can be used in place of money at the service itself provided by the gift AI proposal recommendation device 10 or at a gift reception site or mail order site operated by the operator. Is.

  The gift AI proposal recommendation device 10 designates a user who can be a client by a user who can be a recipient, a desired product registration means 26 for accepting designation of a product for which a gift is desired, and a specific individual request registered thereby. A product database 17 may be provided. The designated product is used as a recommended product together with the output by the weight network parameter 15 or as an input to the network when the client considers a gift to the recipient. It is possible for the client to make a selection that does not go wrong. In addition, it is desirable that not only products that are desired to be received but also products that are not desired to be received, that is, disliked or unnecessary, can be registered as exclusion data. In other words, it is a wish list and an unnecessary list that specify each individual.

  The gift AI proposal recommendation device 10 may include product provider notification means 28 for notifying the product provider terminal 34 of the product that is actually delivered. The gift AI proposal recommendation device 10 according to the present invention does not immediately request delivery to the product provider by the notification from the client terminal 31, but confirms it to the recipient and, if necessary, the product provider. The notification means 28 is executed.

  The gift AI proposal recommendation device 10 transfers the receipt right of the gift registered in the record of each user in the receipt database 12 to the record of another user for each gift. It is good to have. When this is performed mutually between users, it is exchange of the right to receive. Further, it may be possible for the operator of the gift AI proposal recommendation device 10 to annihilate this receipt right and convert it into a point that can be used as a money substitute in the service. Further, the receipt right may be transferred (transferred) to a third party (receipt right transfer means 29).

  Since the gift AI proposal recommendation device 10 receives instructions necessary for executing these means, the gift AI suggestion device 10 is used for HTTP or similar access from the client terminal 31, the receiver terminal 32, or the operator terminal 33 used by the operator. On the other hand, it is preferable to have web server means 20 that transmits a web page having an input function and receives an input from the web page.

<Embodiment 1: Recommended return merchandise>
Next, a specific service embodiment using the gift AI proposal recommendation device 10 according to the present invention will be described. As a first embodiment, a proposal / recommendation (recommendation) function for return merchandise will be described with reference to FIG. The gift AI proposal recommendation device 10 proposes a gift product using the output by the weight network parameter 15 to the client terminal 31 (S101). Specifically, it is a display in a product selection web page or a notification in e-mail or SNS. The data input to the weight network parameter 15 and referred to at this time include the use of the gift, the date (season), the category of the gift product corresponding to the use and date, the price, the recipient's age, gender, the residence area, and the request. The relationship between the person and the receiver, the product list registered in the specific personal desired product database 17 by the receiver, the recommended product information from the product provider affiliated with the operator of the gift AI proposal recommendation device 10 and the like are used. be able to.

  The client browses the proposed product, designates a gift, and makes a settlement for giving the gift to the recipient (S102). As a result, the gift registration means 22 additionally registers the gift receipt right in the recipient record of the receipt database 12. Next, the notification means 23 notifies the recipient terminal 32 (S103). At this time, the merchandise proposing means 21 proposes a return merchandise from the recipient to the client on the recipient terminal 32 (S104). The recipient examines and designates the return item with the product displayed on the terminal as the first impression (S105).

  As data to be referred to directly by the product proposing means 21 at this time (S104) as the input to the weight network parameter 15, for example, the following are listed. The use of the gift from the original client, date (season), category of the gift product from the original client, price of the original gift, age of the client (ie the recipient of the return gift), gender, residence area , The relationship between the client and recipient of the original gift, the list of desired and unnecessary items in the client's specific personal desired product database 17, customs information on return, recommended product information from the product provider, etc. Can be mentioned.

  These return merchandise proposals can be recommended regardless of how the recipient handles the gift (receipt, exchange, transfer, hold). If the recipient can be informed of the recommended site for the return product in the notification email of the gift, it can be expected that the opportunity for the return will increase. For example, in a system (service) that uses the gift AI proposal recommendation device 10 at the same time when it is received from a client on a wedding or birthday, a service having the following characteristics (1) to (4) is provided as S105. It becomes possible. (1) You can return. (2) Since the return can be made within the range of the value of the gift, the recipient can pay himself or return to the client without paying. (3) Depending on the return rate setting, recipients can return by selecting products such as 3,000 yen, 5,000 yen, etc., and receive a difference product that is the gift product minus the value of the return product. It has the characteristics that can be. Regardless of which of the return product and the difference product is selected first, it can be selected from the selection display of each value product. (4) In addition to being able to return in an instant with this system, the recipient does not have to pay for it, so there is less troublesome work such as searching for other products to return, and the custom of returning is less common. In the meantime, it will be a unique service that can be expected to increase the return opportunities.

  For example, if the client receives a celebration of Buddhist affairs or marriage, the recipient can instantly recommend and select a return product of about 1/3 to half the price. At this time, the gifted product (for example, an amount equivalent to 10,000 yen) is virtually received, the product of 5000 yen is returned, and another product is received by the remaining 5000 yen, or the third Can be transferred to a person.

  Recipients do not receive or exchange the gifts given by the client as they are, but forward (re-gift) the products or exchanges to their close relatives or others, such as by hand or by hand. be able to. It is possible to receive a product equivalent to 5,000 yen, which is half of the gifted goods (for example, an amount equivalent to 10,000 yen), and transfer the remaining 5,000 yen goods at the recipient's will. This transfer function is implemented as “receipt transfer” that transfers the receipt right to a third party due to the system structure. However, this transfer function is exactly the same as “gift transfer” that transfers the gift to the third party as it is. Operates in principle. (This is why the receiving right transferring means in FIG. 1 is connected to the gift transferring means by “=”.)

  Whether or not the original requester (recipient of the return item) receives the specified return item is notified by the above notification means 23 to the original requester. , Gift exchange, gift transfer, hold, or point conversion. However, since it is difficult to think of further repayments for repayments, it is not necessary to propose repayments for notification of replies.

<Embodiment 2: Proposal of related products>
Next, as a second embodiment using the gift AI proposal recommendation device 10 according to the present invention, a related product recommendation function will be described with reference to FIG. The proposal / recommendation (S121) of the gift product to the client terminal 31 and the designation / settlement (S122) of the gift according to the gift product are the same as S101 and S102 of the first embodiment. Next, when the notification unit 23 notifies the recipient terminal 32 (S123), the product proposal unit 21 proposes a related product whose value is increased by receiving the gift product together with the gift product. (S124). The recipient examines and designates additional purchase of related products using the product displayed on the terminal as a first impression (S125). The operator arranges to deliver the designated related product together with the gift product to the recipient (S126).

  Examples of data that the product proposing means 21 inputs at this time (S124) or directly refers to the weight network parameter 15 include the following. Use of gift from original client, date (season), category of gift product from original client, price, recipient's age, gender, residence area, relationship with recipient, specific personal preference of recipient A list of desired and unnecessary items in the product database 17, recommended product information from the product provider, and the like can be given. In other words, such a related product can be recommended based on the same information as when the gift AI proposal recommendation device 10 recommends a gift product to the client at the beginning (S101, S121). In addition, the proposal / recommendation can be proposed based on information selected by the client, and the analysis proposal based on the request and weighting of the recipient.

  The input to the weight network parameter 15 at the time of proposing the related product includes personal information of the recipient, attribute information of the gift, information on the gift event, receipt history of the recipient and receipt history of the gift itself (popularity). It is better to change the weight by adding or subtracting. This weighting of addition and subtraction is based on the rule (rule base) according to the following items, increase / decrease as a statistical counter from the purchase history, etc., and the importance of the purchase history itself is increased or decreased by machine learning, Proposal changes.

1. Weighting of “product” from the client and the same or similar comparison <1> Weighting by the number of entrances (visit, click of interest, etc.) <2> Exit (application, application, request / contract) <3> Requester for browsing operations between entrances and exits, Weight 2 based on results of operations by recipients, Weight 3 for “event use & promotion keywords” from product providers, Weighted by the combination of “season / time” from the client and the demand of the product 4; “Unit price of purchase, receipt, transfer etc. and weight 5 by the total amount of unit price calculated by the system; Weighting <1> Entrance (visit, click of interest, etc.) <2> Exit (application, application, receipt) (3) Weighting by the number of transfers, withdrawals, etc. <3> Requester for the browsing operation between the entrance and exit, weighting 6 by the result of operation by the recipient, purchase and selling price for the product provider to secure profit margin Weighted on the value of profit (rate) after subtracting 7, total of 8 combinations of client, recipient, and product provider, “Father, Mother, Sister, Sister, Brother, Brother, Uncle, Aunt, Lover, etc. The value is assigned to each item of weight 9, which is summed up from the relationship that consists of a specific field of "person", various product information, personal information, and attribute information (gender, age, region, season, hobby, taste, etc.) Consumption forecast by adding weight to the sum of the items 10 and events such as childbirth, growth, graduation, employment, marriage, and the 20th anniversary of marriage. The analysis proposal from these is also a feature in addition to the above adjustment. This corresponds to the “begging” function realized in the specific personal desired product database 17, and can be different for each specified person such as father, mother, sister, sister, brother, brother, uncle, aunt, lover, etc. .
11. As shown in FIGS. 6A, 6B, and 6C, a plurality of groups are defined, and the weights in these divided groups are analyzed. For example, (a) duplication and related weights, and relations (such as the number of clicks, number of contracts, etc.) between relatives such as parents, siblings, grandparents, grandchildren, uncles, aunts (b) on the system Weighting in gender etc. Furthermore, the association with the desired product (c) selected in the past may be added and weighted.

  Each weight is represented by a natural numerical value.

  As described above, after weighting each product from various angles, the average and standard deviation in each weight are obtained and standardized. When standardization of each weight is completed, a “weight space” for each product can be defined. Define a proposal / recommendation function that uses each weight in the weight space as a feature value, and evaluate the calculation result (the actual gift selection result by the client and the gift processing result by the recipient are used as teacher data) By providing feedback to the proposed / recommended functions, machine learning (that is, an AI method) and (semi) automatic improvement can be repeated.

  The weight standardization is performed as shown in FIG. The values of the weights determined for each product (marked with x in the figure) vary greatly depending on the input, so it is difficult to compare each value alone, and how much priority each value has in terms of standard deviation It is desirable to standardize the degree.

The weight space is represented by a radar chart as shown in FIG. Although the number / value of parameters is assumed to be 6, there are actually more parameters. Here, the correspondence between the weight space of the product A and the standardized weight value is shown. Define the proposed / recommended function R with each value in the product weight space as the feature quantity x. R is represented by the following formula (1).
R = Σw [k] · x [k] (1)
Here, w [k] represents a “weight” in R (the value in the “weight space” is input data (that is, a feature value) represented by x [k] in the above function). ). W [k] can be optimized by machine learning. Teacher data used in this optimization includes feedback of evaluation (that is, whether or not the purchase / reception has actually been performed).

  By repeating such optimization, related products that are easy to purchase together with the original gift product become suitable, and at the same time, the purchase rate can be improved.

<Embodiment 3: Asking function>
Next, as a third embodiment using the gift AI proposal recommendation device 10 according to the present invention, an example of the relationship between users is shown in FIG. a) Shown in (b). The user may be divided into one or a plurality of groups and defined as a relationship between the groups (FIG. 6A). In addition, relatives such as parents, siblings, grandparents, grandchildren, uncles and aunts may be registered, and these families may be defined as one group (FIG. 6B). Further, a desired product may be associated with each group (FIG. 6C).

  FIG. 7 shows a conceptual diagram of the asking function utilizing such a relationship. A future recipient who expects a present uses the desired product registration means 26 to specify a related person who is a user and specifies a desired product for each. The solicitation for the person who wants to give it (future client) is performed by the notification means 23 by the system of the gift AI proposal recommendation device 10 to avoid being directly asked by the recipient. Even relatives have different relationships among older brothers, fathers, mothers, sisters, etc. Also, since the products asked for at each gift event are different, it is also characterized by proposing and recommending products that are asked for by identifying the relationship.

  Also in the above, for each product, the “degree of begging” according to the human relationship (= social distance) is set as a numerical value, and the number of times of browsing / searching, selected as begging products / the number of times actually bought, The weighting by the product provider, whether it is a seasonal product, the recipient's desire, and the price (profit amount) are defined and set as the weights in the “Ask”. The content of the weighting is the same as that of the above embodiment, but the elements are different.

That is, <1> standardization of each “between weight”, <2> definition of “between weight” space of each product, <3> “between weight” (y [k]) of each product as a feature value Q) is defined. Q is expressed by the following equation (2).
Q = Σv [k] ・ y [k] (2)
Here, v [k] represents the “weight” in Q (the value of the “onset weight” is input data (feature value) represented by y [k]).

  The evaluation is fed back and v [k] is optimized. Here, the evaluation (teacher data) greatly affects whether or not the product is actually selected as a target product and whether or not the target partner has purchased it. Note that a part of v [k] may have the same value as x [k] in the above equation (1), but Q and R are completely different functions. The initial values of w [k] and v [k] in R and Q are assumed to be set manually by an administrator or a product provider. In addition, feedback by machine learning of R and Q is performed after statistically sufficient data is obtained, so at first, based on the functions R and Q obtained by the above manual setting values, we propose Or the product which is suitable for a request will be calculated.

  A third embodiment of the gift analysis apparatus according to the present invention uses the user database 11 including the above-described human relationship, the specific individual desired product database 17, and the history database 16 obtained as a result thereof as follows. Can be applied.

  As a first application form, data of auxiliary information (gift products, gift amount, gift date / time, gift gifts) performed by a client or recipient using a computer device (hereinafter referred to as PC, server, smart phone, device, etc.) Analysis proposal management system that performs gift event management including application, relationship, memo writing, etc.).

  As a second application, the “client” calls and communicates with “relatives” and “auxiliary information”, “client information”, “relatives information” (relatives and relatives) Friends, acquaintances, bosses, subordinates, etc.) and relationships can be analyzed.

  As the third application form, the result of the second application form, product purchase history, inquiry history, support history, transaction, and the client calls to the close relative and analyzes and recommends the memo information obtained from the close relative. Analytical proposal management system.

  As a fourth application form, there are “monetary celebrations and affairs” such as the product value and points between the “client”, “recipient”, and “product provider” and the amount of money that is wrapped in a celebration bag and given in money such as a celebration bag. Weights from “clients” such as memos for the amount of gifts such as “money” and auxiliary information related to memos written by close relatives (information stored and received between the client and relatives) , Distance creation and analysis suggestion management system by recommending reminders to the upper and lower people by frequency from the number of access and contracts.

  As the fifth application form, in addition to the memo information stored from the dialogue with the close relative in the second application form, sales promotion is promoted by “product providers” and “advertisers” from log data, history information, etc. In order to do this, there is an analysis proposal management system that enables data linkage with computer devices with “product providers, advertisers, participating companies”.

  As a sixth application form, it is possible to send specific products to be analyzed and proposed from the data in the second and third application forms and advertisement contents to product providers, advertisers, participants, clients, and close relatives. Analysis proposal management system.

  As the seventh application form, the analysis method enables posting and recommending specific products based on the analysis results, and commercialization planning (assortment, quality, taste, color, price, shape change, etc.) Proposal management system.

  As an eighth application form, there is an analysis proposal management system that imports data from orders, purchases, sales management systems, etc., and can be set and analyzed to automatically order orders such as remaining inventory from the purchase history. It is done.

  In these application forms, for example, in the case of products, personal information of recipients, attribute information of gifts (gender, age, generation, income, occupation, family structure, region, hobbies, preferences), gift event information, receipt The weighting can be changed by adding points based on the receipt history of the person and the receipt history (popularity) of the gift itself. As a result of this weighting, the combination of the gift product (A) and the related product (1, 2, 3) changes to A and 1, A and 2, A and 3, and so on.

  Specifically, human relationships such as family (father, mother, sister, brother), relatives (grandfather, grandmother, uncle, aunt, nephew, nephew, etc.), friendship (synchronous, junior, senior, etc.), company For each product based on the relationship information between the client and the receiver, the personal information of the client and the receiver, and the product attribute information, which defines all relationships ranging from relationships (supervisors, subordinates, business partners, etc.) This is a method for setting and analyzing the weight space information and optimizing the product proposal / recommendation order by adjusting the weight information.

  Based on the relationship between the client and recipient, etc., and using intelligent processing analysis, the message is issued from the “client”, “recipient”, and “product provider” by two- or three-party dialogue analysis. In the “product analysis” in the relationship weight space, the weight is changed by adding the product, the individual, individual attribute information, and composite information. Furthermore, if the proposal / recommendation order in which weights are more suitable and the weights are changed more automatically or manually due to events or the like, and the contact points for each item are expressed as a space, the contact points become wide and accurate. By changing the display order according to the weight of the relationship, it is possible to construct an analysis proposal management system in which sales promotion recommendations and product contents change according to the event.

  Logic based on event information is, for example, an event such as “Growth, Graduation, Employment, Marriage, 20th Anniversary of Marriage, etc.” predicting the next analogy from the purchase history of childbirth, and an event analogizing in the third, fifth and seventh years after childbirth Since Shichigosan is added, 7553 is added, and products belonging to the Shichigosan group are recommended to be displayed based on the analysis results described above, promoted by advertisements, and encouraged to give gifts or invite congratulatory invitations that match the recipient's needs. To be performed by a person.

  The algorithm for recommending by analogy on the database that is analyzed and proposed from these event predictions is characterized by the recommendation from the purchase history and the confirmed event date and time directly input by the client.

  Receiving function from the recipient is to determine the relationship between the specified “client and recipient” such as father, mother, sister, sister, brother, brother, uncle, aunt, lover, friend, etc. I can ask. For example, as an example of a preference analysis of purchase histories of women in their 20s, a woman who has an adult ceremony searches for general trends in begging products, and watches such as watches and other jewelry items are viewed by the client's parent. It is possible to compare the average festive amount with the amount of money, general genre, and product genre, as well as whether it is too flashy or plain.

  The gift AI proposal recommendation device 10 according to the present invention and the system and service as its application form are provided for both the product provider and the consumer provided by the “product provider” to the “client”. A service system can be realized. Analyzed and proposed based on customer information related to the user database 11, product information related to the product database 13, history information including attribute information recorded in the history database 16 when a product is purchased ("gift purpose event of" client "), etc. The contents of the network parameter 15 are updated, and the relationship is properly maintained. By setting the weight network parameter 15 in combination with these attributes, the notification means 23 implements the sales promotion and measures by the product proposal display and the sales promotion plan by mail delivery or posting of my page. In addition, data is accumulated in the user database 11 or the history database 16 at the time of accessing the EC site or visiting a store in cooperation with the operator, and can be input to the weight network parameter 15 as individual information. In response to such a user's network or physical history, campaigns and recommendation screens are identified and managed for each client. Mainly as an attribute of the “client”, the analysis and recommendation can be performed based on the stored information to promote consumption and sales, and to increase the contract rate of buying and selling. Specifically, “x1 to x6” in FIG. 5 is, for example, that the social distance between the “client” and the “recipient” is close (relatives, friends) or far (company boss, business partner). This value represents the recommended degree of the product, but these values can be changed by the product provider.

  A number of conventional analyzes from simple client information can be seen as general proposal functions. On the other hand, the gift AI proposal recommendation device 10 in this applied embodiment not only dynamically performs the proposal via the weight network parameter 15 but also makes a weight-based analysis proposal based on reverse lookup from the recipient. It also has the asking function described above. In web marketing, general analysis values such as the number of times a product has been selected, the number of purchases, the number of clicks on a site, the number of withdrawals, and the like can be acquired by linking the browser interface.

  Although the initial weighting of the weight network parameter 15 is arbitrarily set, the coefficient in the proposed / recommended function changes depending on the machine learning method, so the same thing is not always proposed and may become obsolete. Since there is no one, it will be delivered separately, so only recommendations related to you and your close relatives will be delivered in a timely manner.

  The training of the weight network parameter 15 is a proposal / recommendation based on a general rule base and “intellectual processing” by machine learning, that is, an analysis proposal method and result based on settings such as grouping and weighting. Proposal / recommendation by adding intelligent processing through feedback is made possible for “client” and “recipient” individually.

  As described above, after weighting each product from various angles, the average and standard deviation in each weight are obtained and standardized. The training and setting of these weight network parameters 15 are the same in the other embodiments except for the presence / absence of the “request” function. When standardization of each weight is completed, a “weight space” for each product can be defined. If a proposal / recommendation function is defined with each weight in the weight space as a feature quantity, and the result of this calculation is evaluated and feedback is provided to the proposal / recommendation function, the machine learning (that is, the method using AI) (semi-) ) You will be able to make improvements automatically.

  In this third embodiment, for each product, the “degree of begging” corresponding to the human relationship (= social distance) is set as a numerical value. The number of times received, the weight by the product provider, whether it is a seasonal product, the recipient's desire, and the price (profit amount) are defined and set as the weights for “Ask”. Thereafter, the flow is the same as above. That is, <1> standardization of each “between weight”, <2> definition of “between weight” space of each product, <3> “between weight” (y [k]) of each product as a feature quantity, (Q) ”is defined. Q is represented by the above formula (2). After that, <4> evaluation (whether it was actually selected as a target product, whether the target customer bought it) was fed back (with teacher data) and a process to optimize v [k] was added. Is done.

  In the above equation, part of v [k] may have the same value as x [k], but Q and R are completely different functions. The initial values of w [k] and v [k] in R and Q are assumed to be set manually by an administrator or a product provider. On the other hand, feedback by machine learning of R and Q is performed after statistically sufficient data is obtained, so at first, based on the function R or Q obtained by the above manual setting value, the proposal Or, a product suitable for asking is calculated.

<Embodiment 4: Rights Hold Processing (Receiving Rights Hold) Function>
Next, as a fourth embodiment using a gift AI proposal recommendation device 10 according to the present invention, a right hold processing (reception right hold) function will be described with reference to the conceptual diagram of FIG. The gift AI proposal recommendation device 10 makes a proposal to the client terminal 31 as in the first and second embodiments (S141), receives a gift designation and settlement (S142), and receives a gift registration. The human terminal 32 is notified (S143). The recipient receives the notice that the recipient chooses to receive the gift, and holds only the right to receive the goods (goods receipt right) with the actual delivery suspended ( S144). Upon receiving this, the gift AI proposal recommendation device 10 executes the right hold processing means 24 for holding the actual delivery. In processing, an order is received, and the gift is registered in the recipient's record in the receipt database 12 and put on hold. At this stage, actual delivery processing by the product provider or delivery company does not occur. However, if the product is actually out of stock, later delivery will not be possible, so the product provider notifying means 28 suspends the delivery of the gifted product by the gift requester although the order is received to the product provider terminal 34. Is executed (S145). As a result, the merchandise provider does not need immediate delivery, but the order is closed and can be counted as received. In addition, upon receiving a notification of cancellation of the hold from the recipient terminal 32, after confirming that the gift is present in the recipient record in the receipt database 12, the merchandise provider terminal 34 is notified. The merchandise provider notification means 28 is executed so as to request actual delivery (S146). Thus, as long as the recipient has the right to receive the product, the recipient can receive the product at his / her will at any time.

<Embodiment 4-2: Gift exchange>
A further application form of the fourth embodiment will be described with the conceptual diagram of FIG. That is, for the received product (S151 to S153), not only the receipt is suspended, but the suspended receipt right can be exchanged with the receipt right of other products provided by the product provider. If the recipient has no intention of receiving the gift in the future, he / she notifies that he / she wants to exchange the gift for exchange with another product (S154). Upon receiving this notification, the gift AI proposal recommendation device 10 notifies the product provider (S155), and the gift exchange means 30 proposes and recommends an exchange product that can be exchanged for a product that currently has the right to receive. (S156). The recipient selects and determines a replacement product from the recipient terminal 32 (S157). Upon receiving this, the gift AI proposal recommendation device 10 requests the merchandise provider terminal 34 for shipping (S158).

  In the gift exchange described above, the exchange product suggested / recommended by the gift AI proposal recommendation device 10 according to the input information from the product provider terminal 34 may not be one company A as shown in FIG. Product listings from multiple product providers (persons) such as B and C are possible. In this case, instead of the product selected by the client, the exchanged product selected by the recipient is delivered, and the sales to the sales (resale) company that provided the product provider terminal 34 from the client's payment price. It is possible to settle the fee and the product price to the product provider actually selected by the recipient.

<Embodiment 5: Receiving Rights Exchange / Transfer Function>
Next, as a fifth embodiment using the gift AI proposal recommendation device 10 according to the present invention, a receiving right exchange / transfer function will be described with reference to the conceptual diagram of FIG. The gift AI proposal recommendation device 10 makes a proposal to the client terminal 31 as in the first and second embodiments (S161), receives a gift designation and settlement (S162), and receives a gift registration. The human terminal 32 is notified (S163). The steps so far are the same as in the other embodiments. When the client who has received the notification (A in FIG. 10) determines that the gift is not required to be received, the right holding processing means 24 registers the receipt right in the receipt database 12 in the fourth embodiment. In this state, a proposal is made to exchange the gift (reception rights exchange) with another user (S164). The gift AI proposal recommendation device 10 performs matching between the receiving rights provided for the exchange and mediates exchange, and exchanges the records in the received goods database 12 when the exchange is agreed between the users. An example of this record exchange is shown in FIGS. User A proposes to have a receipt right for personal computer V and give it in exchange, and User B has the right to receive a gift for bag H and proposes to give it in exchange ing. These are matched by the search or the output by the weight network parameter 15 based on the information of each user in the user database 11, and if established, the state before the replacement in FIG. 10 (b) to the state after the replacement in FIG. 10 (c) The receipt record is exchanged. At this stage, since the actual merchandise is not delivered from the merchandise provider, it is the receiving rights that are provided for the exchange, and the gift that can be received from the merchandise provider after the exchange is new. As a result, the conventional barter service provides a merit that it can be obtained as a new article for the user after the exchange, solving the problem that the conventional barter service has only been exchanged as a new or used article. Further, unlike the case of conversion into points, the exchange is performed based on the value judgment for the person holding the right to receive the merchandise, so the exchange exceeding the monetary value is established.

<Embodiment 5-2: Re-gift>
Further application forms will be described with the conceptual diagram of FIG. Immediately after the recipient has received the right to receive the gift, or after the recipient has suspended the right to receive the gift or after exchanging the right to receive the gift shown in FIG. You can forward (re-gift) to your close relatives and others. When receiving right transfer means 29 is instructed by recipient terminal 32 to transfer the receiving right to specified user C for the receiving right of the specified product (bag H in this case) (S171), FIG. From (b) to FIG. 11 (c), the receipt right record for the bag H is moved. The notification means 23 notifies the terminal of the user C receiving the transfer (re-gift recipient terminal 35) that the transfer has been received (S172). Receiving the notification, the user C notifies the user B of the reception (S173). Actually, S173 is a notification via the gift AI proposal recommendation device 10. In addition, the user C who is the recipient at the time of the re-gift can also exchange the right to receive.

<Application: Hold / Exchange / Transfer>
In the fourth and fifth embodiments described above, the “right” to be held, exchanged, or transferred is the right to receive the product (product ID and number as shown in FIGS. 10, 11 (b) and (c)). ), And may be managed as a gift operation right (which can be sent as well as received) as a point for redeeming the item. This is implemented by moving this right or point. The point may be associated with the received product database 12 or may be associated with the user database 11.

DESCRIPTION OF SYMBOLS 10 Gift AI proposal recommendation apparatus 11 User database 11a Requester database 11b Recipient database 12 Received database 13 Product database 15 Weight network parameter 16 History database 17 Specific personal desired product database 20 Web server means 21 Product proposal means 22 Gift registration Means 23 Notification means 24 Rights hold processing means 25 Point conversion means 26 Desired product registration means 27 Receipt rights exchange means 28 Merchandise provider notification means 29 Receipt rights transfer means 30 Gift gift exchange means 31 Requester terminal 32 Recipient terminal 33 Operator Terminal 34 Product provider terminal 35 Re-gift recipient terminal

Claims (8)

A user database containing contact information that can be contacted to identifiable users;
Received goods database that registers the right to receive gifts for each of the above users,
A weight network parameter that constitutes a weight space set for each product;
A gift registration means for registering the right to receive a gift for the product designated by the recipient designated by the client, designated by the terminal, in the recipient's record in the receipt database, and receiving the receipt database When a right to receive a new gift is added to a person's record, a notification means for notifying the recipient using the contact information registered in advance and output by the weight network parameter A gift AI proposal recommendation device having a product proposal means for presenting a product.   2. The gift AI proposal recommendation device according to claim 1, wherein the product proposal means is executed for input of gift information including a notified gift when the notification means notifies the recipient. .   3. The gift AI proposal recommendation device according to claim 2, wherein the product presented by the product proposal means presents a suitable product as a return from the recipient to the client.   3. The gift AI proposal recommendation device according to claim 2, wherein the product presented by the product proposal means presents a suitable product by being added to the product and received by the recipient.   The product presented by the product proposing means presents a gift that the right to receive by the other recipient is exchangeable for the right to receive the gift of one recipient. Or the gift AI suggestion device described in 2.   6. The gift AI proposal recommendation device according to claim 5, wherein the right to receive the gift of the one recipient is temporarily suspended in accordance with an instruction from the recipient. A specific personal desired product database for registering a product for which a user wishes to give a gift by designating a user who can be the client by the user who can be the recipient,
When the product proposing means presents a candidate for a gift to the user who can be the receiver, the pair of the client and the receiver corresponding to the specific individual desired product database. The gift AI proposal recommendation device according to claim 1, wherein the designated product is presented without displaying that the designated product is designated.   A user database that contains contact information that can be contacted to identifiable users, a receipt database that registers the right to receive gifts for each user, and an update that includes the relationship between individual products Using the weight network parameter having a possible weight space, the right to receive a gift for the product designated by the recipient designated by the client, designated by the terminal, is stored in the recipient database in the receipt database. Register and notify the recipient using the pre-registered contact information when the right to receive a new gift is added to the recipient record in the recipient database, and the weight A method for providing a product presentation service for presenting products output by network parameters.