Recipe presentation device, recipe presentation system, recipe presentation method, and program

Abstract

To provide a recipe presentation apparatus, a recipe presenting system, a recipe presentation method, and a program capable of presenting cooking recipes in consideration of a user's family structure and reflecting chronological cooking recipe selection tendencies.SOLUTION: A cloud server 1 comprises: a family structure estimation unit 114 for estimating a user's family structure using a family structure estimation model for estimating the user's family structure from cooking rice amount information indicating the amount of rice to be cooked in one rice cooking process in a rice cooker 2 used by a user, purchase history information indicating a history of purchases made in electronic commerce in the past by the user, and user information including the user's gender and age; and a rice type recipe estimating unit 115 for estimating a rice type and a cooking recipe desired by the user using a rice type recipe estimating model for estimating the combination of the rice type and cooking recipe desired by the user from the cooking rice amount information, the user information, family structure information indicating the family structure, and recipe identification information identifying the cooking recipes that the user has browsed in the past.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present disclosure relates to a recipe presentation device, a recipe presentation system, a recipe presentation method, and a program. [Background technology]

[0002] There has been proposed an ingredient management system that refers to an inventory list of ingredients stored in a refrigerator and suggests recipes that are in line with a user's desired recipe guidelines (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2021-89105 Summary of the Invention [Problem to be solved by the invention]

[0004] Meanwhile, in systems for proposing recipes such as the food ingredient management system described in Patent Document 1, there is a demand for a system that can propose recipes taking into account the family structure of the user.

[0005] The present disclosure has been made in consideration of the above reasons, and aims to provide a recipe presentation device, recipe presentation system, recipe presentation method, and program that are capable of presenting cooking recipes that take into account the user's family composition and reflect the user's cooking recipe selection trends over time. [Means for solving the problem]

[0006] In order to achieve the above object, the recipe presentation device according to the present disclosure includes: a family structure estimation unit that estimates the user's family structure using a family structure estimation model that estimates the user's family structure from rice cooking amount information that indicates the amount of rice used in one rice cooking process in the rice cooker used by the user, purchase history information that indicates the user's past purchase history through e-commerce, and user information including the user's gender and age; The system includes a rice variety recipe estimation unit that estimates the rice variety and the cooking recipe using a rice variety recipe estimation model that estimates a combination of rice variety and cooking recipe from the cooked rice amount information, the user information, family composition information indicating the family composition, and recipe identification information that identifies cooking recipes that the user has viewed in the past. [Effects of the Invention]

[0007] According to the present disclosure, a family structure estimation unit estimates a user's family structure from rice cooking amount information, purchase history information, and user information, and a rice variety recipe estimation unit estimates combinations of rice varieties and cooking recipes from the rice cooking amount information, purchase history information, and user information, as well as family structure information indicating the estimated family structure and recipe identification information. This makes it possible to present combinations of rice varieties and cooking recipes based on the user's family structure and the user's past cooking recipe browsing history. Therefore, it is possible to present cooking recipes that take the user's family structure into consideration and reflect the user's cooking recipe selection trends over time. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a schematic diagram illustrating a recipe presentation system according to an embodiment of the present disclosure. [Figure 2] FIG. 1 is a block diagram showing a hardware configuration of a recipe presentation system according to an embodiment. [Figure 3] FIG. 1 is a block diagram showing a functional configuration of a terminal device according to an embodiment; [Figure 4] FIG. 10 is a diagram showing an example of information stored in a rice variety recipe storage unit of a terminal device according to an embodiment. [Figure 5] FIG. 1 is a block diagram showing a functional configuration of a cloud server according to an embodiment. [Figure 6]FIG. 1A is a diagram showing an example of information stored in a rice cooking setting storage unit according to an embodiment, and FIG. 1B is a diagram showing an example of information stored in a purchase history storage unit according to an embodiment. [Figure 7] 1A is a diagram showing an example of information stored in a user information storage unit according to an embodiment, and FIG. 1B is a diagram showing an example of information stored in a viewed recipe storage unit according to an embodiment. [Figure 8] 1A is a diagram for explaining a rice variety recipe estimation model according to an embodiment, and FIG. 1B is a diagram for explaining a family structure estimation model according to an embodiment. [Figure 9] FIG. 10 is a diagram showing an example of information stored in a recipe map storage unit according to an embodiment. [Figure 10] FIG. 1 is a sequence diagram illustrating the operation of a recipe presentation system according to an embodiment. [Figure 11] FIG. 1 is a sequence diagram illustrating the operation of a recipe presentation system according to an embodiment. [Figure 12] FIG. 10 is a diagram showing an example of a recipe presentation image displayed on a display unit of a terminal device according to an embodiment. [Figure 13] FIG. 1 is a sequence diagram illustrating the operation of a recipe presentation system according to an embodiment. [Figure 14] 1A is a diagram showing an example of an operation screen image displayed on a display unit of a terminal device according to an embodiment, and FIG. 1B is a diagram showing another example of an operation screen image. [Figure 15] FIG. 10 is a diagram showing an example of a rice variety display image displayed on a display unit of a terminal device according to an embodiment; [Figure 16] 1 is a flowchart showing an example of the flow of a terminal device control process executed by a terminal device according to an embodiment; [Figure 17] 1 is a flowchart showing an example of the flow of a terminal device control process executed by a terminal device according to an embodiment; [Figure 18] A flowchart showing an example of the flow of a rice variety recipe presentation process executed by a cloud server according to an embodiment. [Figure 19] A flowchart showing an example of the flow of a rice variety recipe presentation process executed by a cloud server according to an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] A recipe presentation device according to an embodiment of the present disclosure will be described below with reference to the drawings. The recipe presentation device according to the present embodiment includes a family structure estimation unit and a rice variety recipe estimation unit. The family structure estimation unit estimates a user's family structure using a family structure estimation model that estimates the user's family structure based on rice cooking amount information indicating the amount of rice used in one rice cooking process in the user's rice cooker, purchase history information indicating the user's past purchases through e-commerce, and user information including the user's gender and age. The rice variety recipe estimation unit estimates the rice variety and the cooking recipe using the rice variety recipe estimation model that estimates combinations of rice varieties and cooking recipes based on the rice cooking amount information, purchase history information, user information, family structure information indicating the estimated family structure, and recipe identification information identifying cooking recipes previously viewed by the user.

[0010] The recipe presentation system according to this embodiment includes, for example, a rice cooker 2 placed in a home H, a terminal device 3 carried by a user of the rice cooker 2, and a cloud server 1, as shown in FIG. 1. A broadband router (hereinafter referred to as "BBR") 6 is installed in the home H, which is capable of communicating with the rice cooker 2 via a local network NW2. The BBR 6 is connected to a wide area network NW1 such as the Internet. The local network NW2 is a wireless local area network (LAN) or a wired LAN. The terminal device 3 and the cloud server 1 can communicate with each other via the wide area network NW1. In addition, an EC site management server 91 and a model management server 92 are connected to the wide area network NW1.

[0011] The EC site management server 91 is a server for operating an EC (Electronic Commerce) site where users can purchase products through electronic commerce. The EC site management server 91 includes an order information storage unit that stores order information indicating the contents of an order from a user using the EC site in association with user identification information that identifies the user, and a purchase history storage unit that stores purchase history information indicating the purchase history of the user corresponding to the order information in association with the user identification information each time an electronic commerce transaction corresponding to the order information is concluded. Furthermore, upon receiving registration information transmitted from the terminal device 3, the EC site management server 91 extracts user information from the received registration information and registers it in the order information storage unit and the purchase history storage unit. During the registration process, the EC site management server 91 allocates areas in the order information storage unit and the purchase history storage unit for storing the order information and purchase information of the user corresponding to the registration information. Furthermore, upon receiving order information transmitted from the terminal device 3, the EC site management server 91 stores the received order information in the order information storage unit. Furthermore, each time a user purchases a product through electronic commerce, the EC site management server 91 chronologically stores the purchase information in an area corresponding to that user in the purchase history storage unit. Furthermore, when the EC site management server 91 receives purchase information request information from the cloud server 1 via the wide area network NW1, the EC site management server 91 identifies the purchase information to be sent to the cloud server 1 in response to the request. Here, the purchase information request information includes user identification information, and the EC site management server 91 identifies the purchase information associated with the user identification information included in the purchase information request information. The EC site management server 91 then transmits the identified purchase information to the cloud server 1.

[0012] The model management server 92 manages training information for updating the family structure estimation model, which will be described later. The model management server 92 generates training information and transmits it to the cloud server 1 every time a preset time for updating the family structure estimation model arrives.

[0013] As shown in Fig. 2, the rice cooker 2 includes an inner pot (not shown) into which rice to be cooked is placed, a heating unit 21 that heats the inner pot, a timing unit 22 that measures the time elapsed since heating began, a temperature measurement unit 23 that measures the temperature inside the inner pot, a rice cooker control unit 24 that controls the heating unit 21, timing unit 22, and temperature measurement unit 23, and a communications adapter 29. Each time a user performs a rice cooking start setting operation to schedule rice cooking, the rice cooker control unit 24 generates rice cooking setting information including rice cooking start time information indicating the start time of the rice cooking process, rice variety identification information that identifies the variety of rice to be cooked, rice cooking amount information indicating the amount of rice to be cooked in one rice cooking process, and user identification information that identifies the user. The rice cooker control unit 24 then outputs the generated rice cooker setting information to the communications adapter 29. The communication adapter 29 transmits the rice cooker setting information input from the rice cooker control unit 24 to the cloud server 1 via the local network NW2, BBR6 and wide area network NW1.

[0014] The terminal device 3 is, for example, a smartphone, and includes a CPU (Central Processing Unit) 301, a main memory 302, an auxiliary memory 303, a display 304, an input unit 305, a wide area communication unit 306, and a bus 309 connecting the various units. The terminal device 3 is operated by a user of the rice cooker 2. The main memory 302 is, for example, a volatile memory such as a RAM (Random Access Memory) and is used as a work area for the CPU 301. The main memory 302 also includes an image-only memory (not shown) that temporarily stores image information to be displayed on the display 304. The auxiliary memory 303 is, for example, a non-volatile memory such as a ROM (Read Only Memory) or a semiconductor flash memory, and stores programs for implementing various functions of the terminal device 3. The display 304 is, for example, a liquid crystal display, an organic EL (Electro-Luminescence) display, or the like, and displays various information. The input unit 305 is, for example, a transparent touchpad placed on the display unit 304, and receives various operation information in response to operations by the user of the terminal device 3, and outputs the received operation information to the CPU 301. The wide area communication unit 306 has an interface for connecting to the wide area network NW1.

[0015] The CPU 301 reads out and executes the programs stored in the auxiliary storage unit 303 into the main storage unit 302, thereby functioning as a presentation time acquisition unit 311, a rice variety recipe acquisition unit 312, an image generation unit 313, a display control unit 314, a reception unit 315, a registration request unit 316, a browsed recipe notification unit 317, an order unit 318, and a rice variety recipe identification unit 319, as shown in FIG. 3. The auxiliary storage unit 303 shown in FIG. 2 also has a rice variety recipe storage unit 331 and an image storage unit 332, as shown in FIG. 3. The image-only memory of the main storage unit 302 shown in FIG. 2 has a display image storage unit 321 that temporarily stores images to be displayed on the display unit 304. The rice variety recipe storage unit 331 stores rice variety information and recipe information acquired from the cloud server 1, as well as score information indicating the degree of suitability of the combination of rice variety information and recipe information for the user's preferences, in association with the combination of rice variety identification information and recipe identification information, as shown in FIG. 4, for example. In the example shown in Figure 4, for example, the combination of "AAA BB rice" and "roast beef" shows a compatibility of "91%" with the user's preferences. Returning to Figure 3, the image storage unit 332 stores various types of image information including image information showing basic icon images for constructing a notification screen image.

[0016] When the presentation time acquisition unit 311 acquires presentation time information transmitted from the cloud server 1, it notifies the image generation unit 313 of the acquired presentation time information. When the rice variety recipe acquisition unit 312 acquires rice variety information, recipe information, score information, rice variety identification information, and recipe identification information transmitted from the cloud server 1, it associates the acquired information with each other and stores it in the rice variety recipe storage unit 331. When the image generation unit 313 is notified of the presentation time information from the presentation time acquisition unit 311, it generates a time presentation image based on the notified presentation time information and stores the generated time presentation image in the display image storage unit 321. Furthermore, when the image generation unit 313 is notified of operation information indicating a recipe viewing operation by the user to view a cooking recipe from the reception unit 315 after the presentation time indicated by the presentation time information, it generates a recipe presentation image based on the rice variety information, recipe information, and score information stored in the rice variety recipe storage unit 331 and stores the generated recipe presentation image in the display image storage unit 321. Furthermore, when the image generation unit 313 receives operation information from the reception unit 315 indicating a user's operation to change the viewed recipe to change the cooking recipe being viewed, the image generation unit 313 identifies a cooking recipe based on the operation information, and then generates a recipe presentation image representing the identified cooking recipe based on the rice species information, recipe information, and score information stored in the rice species recipe storage unit 331, and stores the generated recipe presentation image in the display image storage unit 321. When the image generation unit 313 receives operation information from the reception unit 315 indicating a user's operation to view a rice species presentation image presenting rice types, the image generation unit 313 identifies the rice species based on the operation information, and then generates a rice species presentation image representing the identified rice species based on the rice species information, recipe information, and score information stored in the rice species recipe storage unit 331, and stores the generated rice species presentation image in the display image storage unit 321. The display control unit 314 displays an image corresponding to the information stored in the display image storage unit 321 on the display unit 304.

[0017] The reception unit 315 receives the operation content performed by the user on the input unit 305, generates operation information according to the received operation content, and notifies the image generation unit 313, the registration request unit 316, the viewed recipe notification unit 317, or the order unit 318. When the reception unit 315 notifies the registration request unit 316 of the operation information indicating the user registration operation for registering the user as the user of the rice cooker 2 with the cloud server 1, the registration request unit 316 generates registration information including the device identification information of the rice cooker 2, user identification information identifying the user of the terminal device 3, and user information. Here, the user identification information is, for example, terminal identification information identifying the terminal device 3. The user information also includes gender information indicating the user's gender, age information indicating the user's age, and residential area information indicating the user's residential area. The registration request unit 316 then transmits the generated registration information to the cloud server 1. Furthermore, when the registration request unit 316 is notified by the reception unit 315 of operation information indicating an EC site registration operation for registering the user as an EC site user with the EC site management server 91, the registration request unit 316 generates registration information including the user information about the user input by the user in the EC site registration operation and the terminal identification information of the terminal device 3. Here, the user information includes information indicating the user's name, gender, age, postal code of the residential area, etc. Then, the registration request unit 316 transmits the generated registration information to the EC site management server 91.

[0018] When the receiving unit 315 notifies the viewed recipe notification unit 317 of operation information indicating the above-mentioned viewed recipe change operation, the viewed recipe notification unit 317 generates viewed recipe information including recipe identification information of the cooking recipe viewed by the user and transmits it to the cloud server 1. When the receiving unit 315 notifies the ordering unit 318 of operation information indicating an operation by the user to purchase a product through electronic commerce on the EC site, the ordering unit 318 generates order information for ordering the product from a supplier of the product. Then, the ordering unit 318 transmits the generated order information to the EC site management server 91.

[0019] The rice variety recipe identification unit 319 identifies the combination of rice variety and recipe information with the highest score from the recipe information stored in the rice variety recipe storage unit 331, and notifies the image generation unit 313 of the identified recipe information. In addition, the rice variety recipe identification unit 319 identifies the rice variety with the highest score in combination with the cooking recipe being viewed by the user from the rice variety information stored in the rice variety recipe storage unit 331, and notifies the image generation unit 313 of the rice variety information of the identified rice variety.

[0020] Returning to FIG. 2, the cloud server 1 is connected to the rice cooker 2 via the local network NW2, BBR6, and wide area network NW1, and functions as a recipe presentation device that presents rice varieties or cooking recipes to users. The cloud server 1 is, for example, a computer used for server purposes, and includes a CPU 101, a main memory unit 102, an auxiliary memory unit 103, a wide area communication unit 106, a clock unit 108, and a bus 109 connecting the various units. The CPU 101 is, for example, a multi-core processor. The main memory unit 102 is volatile memory used as a work area for the CPU 101. The auxiliary memory unit 103 has a large capacity nonvolatile memory, functions as a ROM (read only memory), and stores programs for implementing the various functions of the cloud server 1. The wide area communication unit 106 communicates with the rice cooker 2 via the wide area networks NW1 and BBR6, and with the terminal device 3 via the wide area network NW1. The timekeeping unit 108 has, for example, a real-time clock, and outputs time information indicating the time kept by the timekeeping unit 108 to the CPU 101 .

[0021] The CPU 101 reads out and executes the programs stored in the auxiliary storage unit 103 into the main storage unit 102, thereby functioning as a rice cooking setting acquisition unit 111, a purchase information acquisition unit 112, a user information acquisition unit 113, a family structure estimation unit 114, a rice variety recipe estimation unit 115, a rice variety recipe selection unit 116, a rice variety recipe notification unit 117, a presentation time identification unit 118, a presentation time notification unit 119, a viewed recipe acquisition unit 120, a rice variety recipe estimation model generation unit 121, a teacher information acquisition unit 122, and a family structure estimation model generation unit 123, as shown in Fig. 5. The auxiliary storage unit 103 shown in Fig. 2 also includes a rice cooking setting storage unit 131, a purchase history storage unit 132, a user information storage unit 133, a viewed recipe storage unit 134, a model storage unit 135, a recipe map storage unit 136, a rice variety recipe storage unit 137, and a teacher information storage unit 138.

[0022] As shown in FIG. 6(A), the rice cooking setting storage unit 131 stores rice cooking start time information indicating the start time of the rice cooking process, rice variety identification information identifying the type of rice to be cooked, and rice cooking amount information indicating the amount of rice to be cooked in one rice cooking process in the rice cooker 2 used by the user, in association with user identification information that identifies the user. Here, the rice variety identification information indicates a rice variety identification number that identifies each rice variety. As shown in FIG. 6(B), the purchase history storage unit 132 stores purchase history information consisting of purchase item information indicating products that the user has previously purchased through e-commerce and purchase time information indicating the time the products were purchased, in association with user identification information that identifies the user who purchased the products. Here, the purchase item information indicates, for example, a product identification number that identifies each purchased item. As shown in FIG. 7(A), the user information storage unit 133 stores gender information indicating the gender, age information indicating the age, and residential area information indicating the residential area of ​​each user of the rice cooker 2, in association with user identification information. Here, the gender information is binary information, for example, with male being "1" and female being "0." Furthermore, the residential area information is information indicating a postal code, for example. As shown in FIG. 7(B), for example, the viewed recipe storage unit 134 stores recipe identification information that identifies cooking recipes previously viewed by the user and viewing time information that indicates when the cooking recipe was viewed, in association with the user identification information. Here, the recipe identification information indicates a recipe identification number unique to each cooking recipe.

[0023] Returning to FIG. 5, the model storage unit 135 stores information indicating a family structure estimation model for estimating a user's family structure from the rice cooking amount information, the purchase history information, and the user information. The model storage unit 135 also stores information indicating a rice variety recipe estimation model for estimating a user's desired combination of rice variety and cooking recipe from the rice cooking amount information, the purchase history information, the user information, the family structure information indicating the family structure, and the recipe identification information. The family structure estimation model is, for example, a forward propagation multilayer neural network. In this case, the model storage unit 135 stores information indicating the structure of the family structure estimation model and information indicating weighting coefficients in the family structure estimation model. Here, the information indicating the structure of the family structure estimation model includes information indicating the number of nodes and layers of the family structure estimation model, as well as weighting coefficients and activation functions corresponding to each node. This family structure estimation model has an input layer L110, a hidden layer L120, and an output layer L130, as shown in FIG. 8(A). The input layer L110 receives the amount of rice cooked indicated by the rice cooking information, the purchased item information and purchase time information indicated by the purchase history information, and the gender information, age information, and residential area information indicated by the user information.

[0024] The hidden layer L120 is composed of N (N is a positive integer) layers containing a preset number M[j] of nodes x[j,i] (1≦i≦M[j], M[j] is a positive integer). Here, the output y[j,k] of each node x[j,i] is expressed by the relational expression (1) below.

number

[0025] The rice variety recipe estimation model is, for example, a recurrent neural network. In this case, the model storage unit 135 stores information indicating the structure of the rice variety recipe estimation model and information indicating weighting coefficients in the rice variety recipe estimation model. Similar to the family structure estimation model, the information indicating the structure of the rice variety recipe estimation model includes information indicating the number of nodes, the number of layers, and weighting coefficients and activation functions corresponding to each node. Here, the weighting coefficients include weighting coefficients used when feeding back the output from the layer to which each node belongs to the input. As shown in FIG. 8(B), this rice variety recipe estimation model has an input layer L210, a hidden layer L220, and an output layer L230. The input layer L210 receives the amount of rice cooked indicated by the rice cooking amount information, the purchased item information and purchase time information indicated by the purchase history information, the gender information, age information, and residential area information indicated by the user information, and recipe identification information identifying the cooking recipe most recently viewed by the user.

[0026] The hidden layer L220 is composed of three layers including a preset number N[j] of nodes y[j,i,t] (1≦i≦N[j], N[j] is a positive integer). Here, the output y[2,j,t] of each node except for the node immediately after the input layer L210 and the node immediately before the output layer L230 is expressed by the relational expression of the following formula (2).

number

[0027] Returning to FIG. 5, the recipe map storage unit 136 stores recipe map information that associates combinations of rice variety identification information and recipe identification information for multiple types of rice within a preset period with the family composition information and residential area information included in the user information. For example, as shown in FIG. 9, the recipe map storage unit 136 stores combinations of rice variety identification information and recipe identification information for one week in association with the family composition information and residential area information. The rice variety recipe storage unit 137 stores multiple types of rice variety information and recipe information that have been preset in association with the rice variety identification information and recipe identification information, respectively. The teacher information storage unit 138 stores teacher information for training a family composition estimation model. Specifically, the teacher information storage unit 138 stores teacher information including cooked rice amount information, age information, gender information, and residential area information included in the user information, purchased item information, and family composition information corresponding to these.

[0028] When the rice cooking setting acquisition unit 111 acquires the rice cooking setting information transmitted from the rice cooker 2, it extracts the rice cooking start time information, rice type information, cooked rice amount information, and user identification information contained in the acquired rice cooking setting information. Then, the rice cooking setting acquisition unit 111 stores the extracted rice cooking start time information, rice type information, cooked rice amount information, and user identification information in the rice cooking setting storage unit 131 in association with one another.

[0029] The presentation time determination unit 118 determines the presentation time at which to present the cooking recipe and rice type to the user. Specifically, the presentation time determination unit 118 first calculates the average past rice cooking start time at which the rice cooker 2 started cooking rice by referencing the rice cooking start time information stored in the rice cooking setting storage unit 131. The presentation time determination unit 118 then determines the presentation time as a time point that is a preset time period prior to the calculated average time. Here, the presentation time determination unit 118 may, for example, determine a time point that is 30 minutes prior to the average time. The presentation time determination unit 118 also classifies the past rice cooking start times indicated by the rice cooking start time information stored in the rice cooking setting storage unit 131 into three time periods and determines the presentation time for each of the three time periods. Here, the presentation time determination unit 118 classifies the time periods into three time periods, for example, between midnight and 9:00, between 9:00 and 15:00, and between 15:00 and 24:00. The presentation time identification unit 118 then calculates the average time for each rice cooking start time for each time period and identifies the presentation time for each time period from the calculated average times. The presentation time identification unit 118 notifies the identified presentation time to the presentation time notification unit 119. The presentation time notification unit 119 generates presentation time information indicating the presentation time identified by the presentation time identification unit 118 and transmits the generated presentation time information to the terminal device 3.

[0030] Each time the purchase information acquisition unit 112 determines, based on the time measured by the timing unit 108, that the presentation time for presenting a cooking recipe and rice type to the user identified by the presentation time identification unit 118 has arrived, it generates purchase information request information that requests the EC site management server 91 to transmit the user's purchase information. This purchase information request information includes user identification information of the user for whom the purchase information is to be obtained. The purchase information acquisition unit 112 transmits the purchase information request information to the EC site management server 91, thereby acquiring the user's purchase information transmitted from the EC site management server 91. The purchase information acquisition unit 112 then extracts purchase time information and purchased item information from the acquired purchase information, and stores purchase history information consisting of the extracted purchase time information and purchased item information in the purchase history storage unit 132 in association with the user identification information.

[0031] When the user information acquisition unit 113 acquires the registration information transmitted from the terminal device 3, it extracts the user information and user identification information contained in the acquired registration information, and stores the extracted user information and user identification information in the user information storage unit 133 in correspondence with each other.

[0032] The family structure estimation unit 114 estimates the user's family structure using the family structure estimation model based on the rice cooking amount information, purchase history information, and user information. Specifically, using the family structure estimation model stored in the model storage unit 135, the family structure estimation unit 114 calculates an expected value for each of multiple predefined family structures based on the rice cooking amount indicated by the rice cooking amount information stored in the rice cooking setting storage unit 131, the product identification number and purchase date indicating the user's past purchases indicated by the purchase history information stored in the purchase history storage unit 132, and the user's gender indicated by the gender information, age indicated by the age information, and postal code indicated by the residential area information stored in the user information storage unit 133. The family structure estimation unit 114 then identifies the family structure that maximizes the calculated expected value. Here, the family structure is expressed, for example, in terms of the number of people. The family structure estimation unit 114 notifies the rice variety recipe estimation unit 115 of the family structure information indicating the estimated family structure.

[0033] The rice variety recipe estimation unit 115 estimates the user's desired rice variety and cooking recipe using the rice variety recipe estimation model based on the rice cooking amount information and rice variety information, user information, family composition information indicating the family composition estimated by the family composition estimation unit 114, and recipe identification information identifying cooking recipes previously viewed by the user. Specifically, using the rice variety recipe estimation model stored in the model storage unit 135, the rice variety recipe estimation unit 115 calculates the expected value of each combination of multiple rice varieties and cooking recipes immediately after a preset period has elapsed, based on the rice cooking amount indicated by the rice cooking amount information and the rice variety identification number indicated by the rice variety information stored in the rice cooking setting storage unit 131, the gender number indicated by the user's gender information stored in the user information storage unit 133, the age indicated by the age information and the postal code indicated by the residential area information, the number of adults and children in the family, and the recipe identification information of cooking recipes previously viewed by the user stored in the viewed recipe storage unit 134. Here, the preset period is set to, for example, one week. When the viewing periods indicated by the viewing period information stored in the viewed recipe storage unit 134 are classified into three time periods, the rice variety recipe estimation unit 115 selects recipe identification information corresponding to the viewing period that belongs to the same time period as the average rice cooking start time corresponding to the presentation time identified by the presentation time identification unit 118. The rice variety recipe estimation unit 115 then identifies all combinations of rice varieties and cooking recipes whose calculated expected values ​​are equal to or greater than a preset reference expected value. Here, the reference expected value is set to, for example, "50" when converted into a score, as described below. Next, the rice variety recipe estimation unit 115 notifies the rice variety recipe selection unit 116 of the rice variety identification numbers and recipe identification numbers of all identified combinations of rice varieties and cooking recipes. In addition, the rice variety recipe estimation unit 115 converts the calculated expected value for each combination of the identified rice variety and cooking recipe into a score indicating the degree of recommendation to the user by multiplying it by a preset coefficient, and notifies the rice variety recipe selection unit 116 of the score information indicating the score in association with the combination of the rice variety identification number and the recipe identification number.

[0034] Furthermore, if the user has not previously viewed a cooking recipe and the viewed recipe storage unit 134 does not store recipe identification information and viewing time information corresponding to the user's user identification information, the rice variety recipe estimation unit 115 identifies a combination of rice variety and cooking recipe based on the recipe map information stored in the recipe map storage unit 136. Specifically, the rice variety recipe estimation unit 115 identifies rice variety identification information and recipe identification information associated with the family composition information and the residential address information included in the user information from among the combinations of rice variety identification information and recipe identification information constituting the recipe map information stored in the recipe map storage unit 136. Here, the rice variety recipe estimation unit 115 may identify a combination of rice variety and cooking recipe based on the recipe map information, for example, within a predetermined period after the user starts viewing a cooking recipe. After the predetermined period has elapsed, the rice variety recipe estimation unit 115 may start estimating a combination of rice variety and recipe using a rice variety recipe estimation model based on the viewing time information and recipe identification information within the predetermined period stored in the viewed recipe storage unit 134. Here, the preset period is set to, for example, one week as described above. In addition, the rice variety recipe estimation unit 115 notifies the rice variety recipe selection unit 116 of the rice variety identification number indicated by the specified rice variety identification information and the recipe identification case indicated by the recipe identification information.

[0035] The rice variety recipe selection unit 116 selects rice variety information and recipe information corresponding to the rice variety identification number and recipe identification number notified by the rice variety recipe estimation unit 115, from the rice variety information and recipe information stored in the rice variety recipe storage unit 137. Then, the rice variety recipe selection unit 116 notifies the rice variety recipe notification unit 117 of the selected rice variety information and recipe information in association with the score information notified by the rice variety recipe estimation unit 115. The rice variety recipe notification unit 117 transmits the rice variety information, recipe information, and score information notified by the rice variety recipe selection unit 116 to the terminal device 3.

[0036] When the viewed recipe acquisition unit 120 acquires the viewed recipe information transmitted from the terminal device 3, it extracts the recipe identification information and viewing time information contained in the acquired viewed recipe information, and stores the extracted recipe identification information and viewing time information in the viewed recipe storage unit 134 in correspondence with each other.

[0037] The rice variety recipe estimation model generation unit 121 generates a rice variety recipe estimation model using the rice cooking amount information and rice variety information, the age information, gender information, and residential address information included in the user information, the family composition information indicating the family composition, and recipe identification information identifying cooking recipes previously viewed by the user. Here, the rice variety recipe estimation model generation unit 121 generates a new rice variety recipe estimation model from the rice variety recipe estimation model already stored in the model storage unit 135, for example, using the BPTT (Back Propagation Through Time) method. Specifically, the rice variety recipe estimation model generation unit 121 first calculates, using the rice variety recipe estimation model stored in the model storage unit 135, the expected value of each combination of rice variety and cooking recipe for each time from a predetermined number of times going back from the most recent estimation to the most recent estimation, based on the rice cooking amount information and rice variety information, the age information, gender information, and residential address information included in the user information, the family composition information indicating the family composition, and recipe identification information identifying cooking recipes previously viewed by the user. Next, the rice variety recipe estimation model generation unit 121 sets an expected value so that the expected value of the combination of rice variety and cooking recipe corresponding to each time from a point going back a preset number of estimations from the most recent estimation time stored in the viewed recipe storage unit 134 to the most recent estimation time is higher than the expected values ​​of other combinations. For example, the rice variety recipe estimation model generation unit 121 sets the expected value of the combination of rice variety and cooking recipe stored in the viewed recipe storage unit 134 to a number greater than 0, and sets the expected values ​​of other combinations to "0". Then, the rice variety recipe estimation model generation unit 121 calculates the error between the expected value calculated using the rice variety recipe estimation model and the expected value set based on the recipe identification information stored in the viewed recipe storage unit 134. The rice variety recipe estimation model generation unit 121 then determines a weighting coefficient for the rice variety recipe estimation model using the BUTT algorithm based on the calculated error, and stores information indicating the determined weighting coefficient in the model storage unit 135.

[0038] When the teacher information acquisition unit 122 acquires the teacher information transmitted from the model management server 92, the teacher information acquisition unit 122 stores the acquired teacher information in the teacher information storage unit 138.

[0039] The family structure estimation model generation unit 123 generates a family structure estimation model using the rice cooking amount information, the age information, gender information, residential information, and purchased item information included in the user information, and the corresponding family structure information. Here, the family structure estimation model generation unit 123 generates a new family structure estimation model from a family structure estimation model already stored in the model storage unit 135, for example, by error propagation. Specifically, the family structure estimation model generation unit 123 first calculates an expected value for each of a plurality of preset family structures using the rice cooking amount information, and the age information, gender information, residential information, and purchased item information included in the user information, using the family structure estimation model stored in the model storage unit 135. Next, the family structure estimation model generation unit 123 sets an expected value so that the expected value of the family structure indicated by the family structure information corresponding to the rice cooking amount information, age information, gender information, residential information, and purchased item information stored in the teacher information storage unit 138 is higher than the expected values ​​of other family structures. For example, the family structure estimation model generation unit 123 sets expected values ​​corresponding to the rice cooking amount information, age information, gender information, residence information, and purchased item information stored in the teacher information storage unit 138 to numbers greater than 0, and sets expected values ​​for other family structures corresponding to the rice cooking amount information, age information, gender information, residence information, and purchased item information to "0." The family structure estimation model generation unit 123 then calculates the error between the expected value calculated using the family structure estimation model and the expected value set based on the correspondence between the rice cooking amount information, age information, gender information, residence information, and purchased item information stored in the teacher information storage unit 138 and the family structure. The family structure estimation model generation unit 123 then determines weight coefficients for the family structure estimation model using an error propagation method based on the calculated error, and stores information indicating the determined weight coefficients in the model storage unit 135.

[0040] Next, the operation of the control system according to this embodiment will be described with reference to Fig. 10 to Fig. 15. First, when a user performs a user registration operation via the input unit 305 of the terminal device 3 to register with the cloud server 1 that they are the user of the rice cooker 2, the terminal device 3 generates registration information including the device identification information, user identification information, and user information of the rice cooker 2, as shown in Fig. 10 (step S1). Next, the generated registration information is transmitted from the terminal device 3 to the cloud server 1 (step S2). Meanwhile, when the cloud server 1 acquires the registration information, it extracts the user information and user identification information included in the acquired registration information, and stores the extracted user information and user identification information in the user information storage unit 133 in association with each other (step S3).

[0041] Next, when the user performs an EC site registration operation to register as a user with the EC site management server 91 via the input unit 305 of the terminal device 3, the terminal device 3 generates registration information including the user information and terminal identification information of the terminal device 3 (step S4). The generated registration information is then transmitted from the terminal device 3 to the EC site management server 91 (step S5). Meanwhile, upon acquiring the registration information, the EC site management server 91 extracts the user information from the acquired registration information and registers it in the order information storage unit and purchase history storage unit described above (step S6).

[0042] Next, when the user performs a rice cooking start setting operation to schedule rice cooking on the rice cooker 2, the rice cooker 2 generates rice cooking setting information including the rice cooking start time information, rice type information, rice cooking amount information, and user identification information (step S7). The generated rice cooking setting information is then sent from the rice cooker 2 to the cloud server 1 (step S8). Meanwhile, when the cloud server 1 acquires the rice cooking setting information, it extracts the rice cooking start time information, rice type information, rice cooking amount information, and user identification information from the acquired rice cooking setting information, and stores the extracted rice cooking start time information, rice type information, rice cooking amount information, and user identification information in the rice cooking setting storage unit 131 in association with each other (step S9). The cloud server 1 then references the rice cooking start time information stored in the rice cooking setting storage unit 131 and calculates the average time of past rice cooking start times at which rice cooking was started in the rice cooker 2 (step S10). Next, the cloud server 1 specifies a time point that is a preset time period before the calculated average time as the presentation time, and generates presentation time information indicating the specified presentation time (step S11). Subsequently, the cloud server 1 transmits the generated presentation time information to the terminal device 3 (step S12). Meanwhile, upon acquiring the presentation time information, the terminal device 3 displays the presentation time indicated by the acquired presentation time information on the display unit 304 (step S13).

[0043] 11, when the aforementioned presentation time arrives, the cloud server 1 generates purchase information request information that requests the EC site management server 91 to transmit the user's purchase information (step S14). Next, the generated purchase information request information is transmitted from the cloud server 1 to the EC site management server 91 (step S15). Meanwhile, upon acquiring the purchase information request information, the EC site management server 91 identifies the purchase information to transmit to the cloud server 1 based on the user identification information included in the acquired purchase information request information (step S16). Next, the identified purchase information is transmitted from the EC site management server 91 to the cloud server 1 (step S17). Meanwhile, upon acquiring the purchase information, the cloud server 1 extracts purchase time information and purchased item information included in the purchase information, and stores purchase history information including the extracted purchase time information and purchased item information in the purchase history storage unit 132 in association with the user identification information (step S18).

[0044] The cloud server 1 then estimates the user's family structure using the family structure estimation model based on the rice cooking amount information, purchase history information, and user information (step S19). Next, the cloud server 1 estimates the user's desired rice species and cooking recipes using the rice species recipe estimation model based on the rice cooking amount information, rice species information, user information, family structure information indicating the estimated family structure, and recipe identification information identifying cooking recipes previously viewed by the user (step S20). Here, the cloud server 1 identifies combinations of rice species and cooking recipes that, for example, have a score of 50 or higher when the expected value is converted into a score. Next, the cloud server 1 calculates a score indicating the degree of recommendation to the user from the expected value calculated for each combination of the estimated rice species and cooking recipe (step S21). The cloud server 1 then selects rice species information and recipe information corresponding to each estimated rice species and cooking recipe from the rice species information and recipe information stored in the rice species recipe storage unit 137 (step S22). Next, the selected rice variety information and recipe information, as well as the score information, are transmitted from the cloud server 1 to the terminal device 3 (step S23).

[0045] Next, when the user performs a recipe browsing operation to browse recipes via the input unit 305 of the terminal device 3, the terminal device 3 identifies the recipe information with the highest score (step S24). After that, the terminal device 3 generates a recipe presentation image including the cooking recipe indicated by the identified recipe information and score information, and displays it on the display unit 304 (step S25). Here, the terminal device 3 displays a recipe presentation image GA1 including score information M11 on the display unit 304, as shown in FIG. 12, for example. The recipe presentation image GA1 also includes an image M12 that displays rice varieties that are suitable for side dishes listed in the cooking recipe list together with score information.

[0046] Thereafter, when the user performs a viewed recipe change operation to change the viewed recipe via the input unit 305 of the terminal device 3, the terminal device 3 generates a recipe presentation image showing the changed cooking recipe and displays it on the display unit 304 (step S26). Next, the terminal device 3 generates viewed recipe information including recipe identification information and viewed time information of the cooking recipe viewed by the user (step S27). Subsequently, as shown in FIG. 13, the generated viewed recipe information is transmitted from the terminal device 3 to the cloud server 1 (step S28). Meanwhile, upon acquiring the viewed recipe information, the cloud server 1 extracts the recipe identification information and viewed time information included in the acquired viewed recipe information, and stores the extracted recipe identification information and viewed time information in the viewed recipe storage unit 134 in association with each other (step S29). Thereafter, the cloud server 1 generates a rice type recipe estimation model using the aforementioned rice cooking amount information and rice type information, the age information, gender information, and residential address information included in the user information, the aforementioned family composition information indicating the family composition, and the recipe identification information identifying cooking recipes previously viewed by the user. Then, the cloud server 1 updates the information indicating the rice variety recipe estimation model stored in the model storage unit 135 with the information indicating the generated rice variety recipe estimation model (step S30).

[0047] Also, assume that the terminal device 3 is displaying a recipe image on the display unit 304, and the user performs an ingredient ordering operation to order ingredients. This corresponds to the case where the terminal device 3 is displaying a recipe presentation image GA21 on the display unit 304, in which a button image BU1 is arranged in a portion of the input unit 305 that corresponds to a portion to be touched when ordering ingredients, as shown in FIG. 14(A), and the user touches a portion of the input unit 305 that corresponds to the button image BU1. In this case, as shown in FIG. 13, the terminal device 3 generates order information for ordering ingredients (step S31). Next, the generated order information is transmitted from the terminal device 3 to the EC site management server 91 (step S32). Meanwhile, upon acquiring the order information, the EC site management server 91 stores the acquired order information in the order information storage unit (step S33). Next, the terminal device 3 identifies the rice variety with the highest score in combination with the cooking recipe being viewed by the user, and displays the identified rice variety in the recipe image (step S34). Here, the terminal device 3 displays on the display unit 304 a recipe image GA22 including a rice variety display image M21 displaying the rice variety with the highest score and a button image BU2 arranged in a portion of the input unit 305 corresponding to a portion touched when ordering rice, as shown in FIG. 14(B), for example. Then, while the terminal device 3 is displaying the recipe image GA22 as shown in FIG. 14(B) on the display unit 304, a user performs a rice ordering operation to order rice of the rice variety with the highest score. This corresponds to a case where the terminal device 3 is displaying the recipe image GA22 as shown in FIG. 14(B) on the display unit 304, and the user touches a portion of the input unit 305 corresponding to the button image BU2. In this case, as shown in FIG. 13, the terminal device 3 generates order information for ordering rice of the rice variety with the highest score (step S35). Next, the generated order information is transmitted from the terminal device 3 to the EC site management server 91 (step S36). On the other hand, upon acquiring the order information, the EC site management server 91 stores the acquired order information in the order information storage unit (step S37).

[0048] Furthermore, when the user performs a rice variety viewing operation to view the rice variety presentation image via the input unit 305 of the terminal device 3, the terminal device 3 generates a rice variety presentation image including score information of cooking recipes that have relatively high scores when combined with the presented rice variety, and displays the image on the display unit 304 (step S38). Here, the terminal device 3 displays a rice variety presentation image GA3 including score information M31 on the display unit 304, as shown in Fig. 15, for example. The rice variety presentation image GA3 also includes an image M32 that displays the score for the combination of the presented rice variety and the cooking recipe for a side dish.

[0049] 13, the model management server 92 further generates training information each time a preset update period for the family structure estimation model arrives (step S39). Subsequently, the generated training information is transmitted from the model management server 92 to the cloud server 1 (step S40). Meanwhile, upon acquiring the training information, the cloud server 1 generates a family structure estimation model using the training information. Then, the cloud server 1 updates the information indicating the rice variety recipe estimation model stored in the model storage unit 135 with information indicating the generated rice variety recipe estimation model (step S41).

[0050] Next, the terminal device control process executed by the terminal device 3 according to this embodiment will be described with reference to FIGS. 16 and 17. This terminal device control process is initiated, for example, when a program for executing the terminal device control process is launched in the terminal device 3. First, as shown in FIG. 16, the reception unit 315 determines whether the above-described registration operation performed by the user via the input unit 305 has been received (step S201). Here, the registration operation corresponds to the above-described user registration operation or EC site registration operation. If the reception unit 315 determines that the registration operation has not been received (step S201: No), the process of step S203, which will be described later, is executed. On the other hand, if the reception unit 315 determines that the registration operation has been received (step S201: Yes), the reception unit 315 generates registration information corresponding to the registration operation and transmits it to the cloud server 1 or the EC site management server 91 (step S202). Next, the presentation time acquisition unit 311 determines whether presentation time information transmitted from the cloud server 1 has been received (step S203). Here, if the presentation time acquisition unit 311 determines that the presentation time information has not been acquired (step S203: No), the process of step S205 described below is executed. On the other hand, if the presentation time acquisition unit 311 determines that the presentation time information has been acquired (step S203: Yes), the image generation unit 313 generates a time presenting image based on the presentation time information, and the display control unit 314 displays the generated time presenting image on the display unit 304 (step S204). Subsequently, the rice variety recipe acquisition unit 312 determines whether or not it has acquired rice variety information, recipe information, and score information transmitted from the cloud server 1 (step S205). Here, if the rice variety recipe acquisition unit 312 determines that it has not acquired rice variety information, recipe information, and score information (step S205: No), the process of step S207 described below is executed. On the other hand, when the rice variety recipe acquisition unit 312 determines that it has acquired rice variety information, recipe information, and score information (step S205: Yes), it associates these pieces of acquired information with each other and stores them in the rice variety recipe storage unit 331 (step S206).

[0051] Thereafter, the receiving unit 315 determines whether or not the above-mentioned recipe viewing operation has been received (step S207). If the receiving unit 315 determines that the recipe viewing operation has not been received (step S207: No), the process of step S201 is executed again. On the other hand, if the receiving unit 315 determines that the recipe viewing operation has been received (step S207: Yes), the rice variety recipe identification unit 319 identifies the recipe information with the highest score from the recipe information stored in the rice variety recipe storage unit 331 (step S208). Next, the image generation unit 313 generates a recipe presentation image based on the recipe information identified by the rice variety recipe identification unit 319 and the corresponding rice variety information and score information stored in the rice variety recipe storage unit 331, and the display control unit 314 displays the generated recipe presentation image on the display unit 304 (step S209). Next, the receiving unit 315 determines whether or not the above-mentioned viewed recipe change operation has been received (step S210). Here, if the receiving unit 315 determines that it has not received a viewed recipe change operation (step S210: No), the process of step S212 described below is executed. On the other hand, if the receiving unit 315 determines that it has received a viewed recipe change operation (step S210: Yes), the image generating unit 313 identifies the cooking recipe after the change due to the viewed recipe change operation, and generates a recipe presentation image representing the identified cooking recipe based on the rice variety information, recipe information, and score information stored in the rice variety recipe storage unit 331. Then, the display control unit 314 displays the generated recipe presentation image on the display unit 304 (step S211). Next, the viewed recipe notification unit 317 generates viewed recipe information including recipe identification information of the identified cooking recipe and transmits it to the cloud server 1 (step S212).

[0052] Then, as shown in FIG. 17, the receiving unit 315 determines whether the user has performed an ingredient ordering operation to order ingredients corresponding to the cooking recipe via the input unit 305 (step S213). If the receiving unit 315 determines that an ingredient ordering operation has not been received (step S213: No), it determines whether the aforementioned rice type browsing operation has been received (step S214). If the receiving unit 315 determines that a rice type browsing operation has not been received (step S214: No), the process of step S209 is executed again. On the other hand, if the receiving unit 315 determines that a rice type browsing operation has been received (step S214: Yes), the image generating unit 313 identifies the rice type corresponding to the rice type browsing operation and then generates a rice type display image representing the identified rice type based on the rice type information, recipe information, and score information stored in the rice type recipe storage unit 331. Then, the display control unit 314 displays the generated rice type display image on the display unit 304 (step S215). Next, the receiving unit 315 determines whether or not the above-mentioned recipe viewing operation has been received (step S216). If the receiving unit 315 determines that the recipe viewing operation has been received (step S216: Yes), the processing of step S209 described above is executed again. On the other hand, if the receiving unit 315 determines that the recipe viewing operation has not been received (step S216: No), the processing of step S217 is executed again. If the receiving unit 315 determines that the display termination operation has not been received (step S217: No), the processing of step S215 is executed again. On the other hand, if the receiving unit 315 determines that the display termination operation has been received (step S217: Yes), the processing of step S201 is executed again.

[0053] Also, assume that the receiving unit 315 determines in step S213 that it has received an ingredient ordering operation (step S213: Yes). In this case, the ordering unit 318 generates order information for ordering ingredients from an ingredient supplier and transmits the generated order information to the EC site management server 91 (step S218). Next, the rice variety recipe identification unit 319 identifies the rice variety with the highest score in combination with the cooking recipe being viewed by the user from the rice variety information stored in the rice variety recipe storage unit 331 (step S219). Thereafter, the image generation unit 313 generates the aforementioned rice variety display image based on the rice variety information of the rice variety identified by the rice variety recipe identification unit 319, and the display control unit 314 displays the generated rice variety display image on the display unit 304 (step S220). Next, the receiving unit 315 determines whether the user has performed a rice ordering operation to order rice of the rice variety represented by the rice variety display image via the input unit 305 (step S221). Here, if the receiving unit 315 determines that a rice ordering operation has not been received (step S221: No), the processing of step S223 described below is executed. On the other hand, it is assumed that the receiving unit 315 determines that a rice ordering operation has been received (step S221: Yes). In this case, the ordering unit 318 generates order information for ordering rice from a rice supplier and transmits the generated order information to the EC site management server 91 (step S222). Next, the receiving unit 315 determines whether a display end operation for ending the display of the recipe presentation image has been received (step S223). Here, if the receiving unit 315 determines that a display end operation has not been received (step S223: No), the processing of step S209 is executed again. On the other hand, if the receiving unit 315 determines that a display end operation has been received (step S223: Yes), the processing of step S201 is executed again.

[0054] Next, the rice variety recipe presentation process executed by the cloud server 1 according to this embodiment will be described with reference to FIGS. 18 and 19. This rice variety recipe presentation process is initiated, for example, when a program for executing the rice variety recipe presentation process is started in the cloud server 1. First, as shown in FIG. 18, the user information acquisition unit 113 determines whether or not registration information transmitted from the terminal device 3 has been acquired (Step S101). If the user information acquisition unit 113 determines that registration information has not been acquired (Step S101: No), the process of Step S103, described below, is executed. On the other hand, when the user information acquisition unit 113 acquires the registration information transmitted from the terminal device 3, it extracts the user information and user identification information contained in the acquired registration information, associates the extracted user information and user identification information with each other, and stores them in the user information storage unit 133 (Step S102). Next, the rice cooking setting acquisition unit 111 determines whether or not rice cooking setting information transmitted from the rice cooker 2 has been acquired (Step S103). Here, if the rice cooking setting acquisition unit 111 determines that rice cooking setting information has not been acquired (step S103: No), the process of step S108, which will be described later, is executed.

[0055] On the other hand, if the rice cooking setting acquisition unit 111 determines that it has acquired rice cooking setting information (step S103: Yes), it extracts the rice cooking start time information, rice type information, cooked rice amount information, and user identification information from the acquired rice cooking setting information.Then, the rice cooking setting acquisition unit 111 associates the extracted rice cooking start time information, rice type information, cooked rice amount information, and user identification information with each other and stores them in the rice cooking setting storage unit 131 (step S104).

[0056] Next, the presentation time identification unit 118 refers to the rice cooking start time information stored in the rice cooking setting storage unit 131 and calculates the average time of past rice cooking start times at which rice cooking processing was started in the rice cooker 2 (step S105). Next, the presentation time identification unit 118 identifies a point in time that is a preset time period prior to the calculated average time as the presentation time (step S106). Here, the presentation time identification unit 118 classifies the past rice cooking start times indicated by the rice cooking start time information stored in the rice cooking setting storage unit 131 into three time periods and identifies a presentation time for each of the three time periods. The presentation time identification unit 118 also notifies the presentation time notification unit 119 of the identified presentation times. The presentation time notification unit 119 then generates presentation time information indicating the presentation times identified by the presentation time identification unit 118 and transmits the generated presentation time information to the terminal device 3 (step S107).

[0057] Next, the purchase information acquisition unit 112 determines whether the presentation time identified by the presentation time identification unit 118 has arrived, based on the time measured by the clock unit 108 (step S108). If the purchase information acquisition unit 112 determines that the presentation time has not yet arrived (step S108: No), the process of step S119, which will be described later, is executed. On the other hand, if the purchase information acquisition unit 112 determines that the presentation time has arrived, it generates the above-mentioned purchase information request information and transmits the generated purchase information request information to the EC site management server 91 (step S109). As a result, the purchase information acquisition unit 112 acquires the user's purchase information transmitted from the EC site management server 91 (step S110). Next, the purchase information acquisition unit 112 extracts purchase time information and purchased item information from the acquired purchase information, and stores purchase history information consisting of the extracted purchase time information and purchased item information in the purchase history storage unit 132 in association with the user identification information (step S111).

[0058] Thereafter, the family structure estimation unit 114 estimates the user's family structure using the family structure estimation model based on the cooked rice amount information, purchase history information, and user information (step S112). Using the family structure estimation model stored in the model storage unit 135, the family structure estimation unit 114 calculates an expected value for each of a plurality of preset family structures based on the cooked rice amount indicated by the cooked rice amount information stored in the cooked rice setting storage unit 131, the product identification numbers and purchase dates indicating the user's past purchases indicated by the purchase history information stored in the purchase history storage unit 132, and the gender number indicated by the gender information, age indicated by the age information, and postal code indicated by the residential area information stored in the user information storage unit 133. The family structure estimation unit 114 then identifies the family structure that results in the highest calculated expected value.

[0059] Next, the rice variety recipe estimation unit 115 determines whether or not there is a cooking recipe that the user has previously viewed (step S113). Here, if the viewed recipe storage unit 134 stores recipe identification information and viewing time information corresponding to the user identification information of the user for whom a combination of rice variety and cooking recipe is to be estimated, the rice variety recipe estimation unit 115 determines that there is a cooking recipe that the user has previously viewed. Here, it is assumed that the viewed recipe storage unit 134 does not store recipe identification information and viewing time information corresponding to the user identification information of the target user, and the rice variety recipe estimation unit 115 determines that there is no cooking recipe that the user has previously viewed (step S113: No). In this case, the rice variety recipe estimation unit 115 identifies a combination of rice variety and cooking recipe based on the recipe map information stored in the recipe map storage unit 136 (step S114). Specifically, the rice variety recipe estimation unit 115 identifies rice variety identification information and recipe identification information associated with the above-mentioned family structure information and residential area information included in the user information from among the combinations of rice variety identification information and recipe identification information that constitute the recipe map information stored in the recipe map storage unit 136. Next, the processing of step S116, which will be described later, is executed.

[0060] On the other hand, suppose that the rice variety recipe estimation unit 115 determines that there is a cooking recipe that the user has previously viewed (step S113: Yes). In this case, the rice variety recipe estimation unit 115 estimates the user's desired rice variety and cooking recipe using the rice variety recipe estimation model described above, from the aforementioned rice cooking amount information and rice variety information, user information, family composition information indicating the family composition estimated by the family composition estimation unit 114, and recipe identification information that identifies cooking recipes that the user has previously viewed (step S115). Specifically, the rice variety recipe estimation unit 115 uses the rice variety recipe estimation model stored in the model storage unit 135 to calculate the expected value for each combination of multiple preset rice varieties and cooking recipes based on the rice cooking amount indicated by the rice cooking amount information stored in the rice cooking setting storage unit 131, the rice variety identification number indicated by the rice variety information, the gender value indicated by the user's gender information stored in the user information storage unit 133, the age indicated by the age information, the postal code indicated by the residential area information, the number of people in the family, and the recipe identification information of cooking recipes previously viewed by the user stored in the viewed recipe storage unit 134. The rice variety recipe estimation unit 115 then identifies all combinations of rice varieties and cooking recipes for which the calculated expected value is equal to or exceeds a preset standard expected value. Here, the standard expected value is set to, for example, "50" when converted into a score. Next, the rice variety recipe estimation unit 115 calculates a score indicating the degree of recommendation to the user by multiplying the expected value calculated for each combination of the estimated rice variety and cooking recipe by a preset coefficient (step S116), as shown in Fig. 19. Then, the rice variety recipe estimation unit 115 notifies the rice variety recipe selection unit 116 of score information indicating the score in association with the combination of the rice variety identification number and the recipe identification number.

[0061] Thereafter, the rice variety recipe selection unit 116 selects rice variety information and recipe information corresponding to the rice variety identification number and recipe identification number notified by the rice variety recipe estimation unit 115 from the rice variety information and recipe information stored in the rice variety recipe storage unit 137 (step S117). Here, the rice variety recipe selection unit 116 notifies the rice variety recipe notification unit 117 of the selected rice variety information and recipe information in association with the score information notified by the rice variety recipe estimation unit 115. Next, the rice variety recipe notification unit 117 transmits the rice variety information, recipe information, and score information notified by the rice variety recipe selection unit 116 to the terminal device 3 (step S118).

[0062] Next, the viewed recipe acquisition unit 120 determines whether or not the viewed recipe information transmitted from the terminal device 3 has been acquired (step S119). If the viewed recipe acquisition unit 120 determines that the viewed recipe information has not yet been acquired (step S119: No), the process of step S122, which will be described later, is executed. On the other hand, if the viewed recipe acquisition unit 120 determines that the viewed recipe information has been acquired (step S119: Yes), it extracts the recipe identification information and the viewed time information included in the acquired viewed recipe information, and stores the extracted recipe identification information and the viewed time information in the viewed recipe storage unit 134 in association with each other (step S120).

[0063] Thereafter, the rice variety recipe estimation model generation unit 121 generates a rice variety recipe estimation model using the aforementioned rice cooking amount information and rice variety information, the age information, gender information, and residential area information included in the user information, the family composition information indicating the aforementioned family composition, and recipe identification information identifying cooking recipes previously viewed by the user.The rice variety recipe estimation model generation unit 121 then updates the rice variety recipe estimation model stored in the model storage unit 135 with the generated rice variety recipe estimation model (step S121).Here, the rice variety recipe estimation model generation unit 121 generates a new rice variety recipe estimation model from the rice variety recipe estimation model already stored in the model storage unit 135 using the BPTT method.

[0064] Next, the teacher information acquisition unit 122 determines whether or not the teacher information transmitted from the model management server 92 has been acquired (step S122). If the teacher information acquisition unit 122 determines that the teacher information has not yet been acquired (step S122: No), the processing of step S101 is executed again. On the other hand, if the teacher information acquisition unit 122 determines that the teacher information has been acquired (step S122: Yes), the teacher information acquisition unit 122 stores the acquired teacher information in the teacher information storage unit 138 (step S123).

[0065] Next, the family structure estimation model generation unit 123 generates a family structure estimation model using the aforementioned cooked rice amount information, the age information, gender information, residential address information, and purchased item information included in the user information, and the family structure information corresponding thereto.The family structure estimation model generation unit 123 then updates the family structure estimation model stored in the model storage unit 135 with the generated family structure estimation model (step S124).Here, the family structure estimation model generation unit 123 generates a new family structure estimation model from the family structure estimation model already stored in the model storage unit 135, for example, by error propagation.After that, the processing of step S101 is executed again.

[0066] As described above, according to the cloud server 1 of this embodiment, the family structure estimation unit 114 estimates the user's family structure from the rice cooking amount information, purchase history information, and user information, and the rice type recipe estimation unit 115 estimates the user's desired combination of rice type and cooking recipe from the rice cooking amount information, purchase history information, and user information, as well as family structure information indicating the estimated family structure and recipe identification information. This makes it possible to present combinations of rice types and cooking recipes based on the user's family structure and the user's past cooking recipe browsing history. Therefore, it is possible to present cooking recipes that take the user's family structure into consideration and reflect the user's tendency to select cooking recipes over time.

[0067] Furthermore, the rice variety recipe estimation unit 115 according to this embodiment identifies the rice variety and recipe based on the recipe map information when there is no recipe identification information stored in the viewed recipe storage unit 134. This makes it possible to present an appropriate cooking recipe to the user even if the user has not viewed the cooking recipe before.

[0068] Furthermore, the user information according to this embodiment includes residential area information indicating the area in which the user resides, which allows the provision of cooking recipes suited to the area in which the user resides.

[0069] Furthermore, the presentation time determination unit 118 according to this embodiment calculates the average time of past rice cooking start times when rice cooker 2 started cooking rice, and determines the aforementioned presentation time as a point in time a preset time before the calculated average time. The presentation time determination unit 118 also classifies the past rice cooking start times into three time periods and determines the presentation time for each of the three time periods. This allows the cooking recipe to be presented to the user at the appropriate time.

[0070] 12, the terminal device 3 displays on the display unit 304 a recipe presentation image GA3 showing the rice variety combined with the cooking recipe indicated by the recipe information. This allows the user to understand the rice variety that is suitable for the cooking recipe along with the cooking recipe, thereby improving convenience for the user when purchasing rice. Furthermore, by presenting the rice variety, it is possible to encourage contributions to the region where rice cultivation of the presented rice variety is a local industry.

[0071] Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the above-described embodiments. For example, the rice variety recipe identification unit 319 of the terminal device 3 may identify multiple recipe information items having a predetermined standard score or higher from the recipe information items stored in the rice variety recipe storage unit 331. In this case, the image generation unit 313 may generate a recipe presentation image showing a list of cooking recipes for each of the identified multiple recipe information items.

[0072] In the embodiment, the family structure estimation unit 114 may estimate the family structure using only the gender information and age information included in the user information, without using the residential area information.

[0073] In the embodiment, the rice variety recipe estimation unit 115 may estimate the combination of rice variety and cooking recipe using a rice variety recipe estimation model represented by a Long Short Term Memory (LSTM) structure or a Hidden Markov Model (HMM). Alternatively, the rice variety recipe estimation unit 115 may estimate the combination of rice variety and cooking recipe using, for example, a multi-class classification type support vector machine.

[0074] Furthermore, the various functions of the cloud server 1 and terminal device 3 according to the present disclosure may be realized by software, firmware, or a combination of software and firmware. In this case, the software or firmware may be written as a program, and the program may be stored and distributed on a computer-readable recording medium such as a flexible disk, a CD-ROM (Compact Disc Read Only Memory), a DVD (Digital Versatile Disc), or an MO (Magneto-Optical Disc). The program may then be read and installed on a computer to configure a computer capable of realizing each of the aforementioned functions. In addition, when each function is realized by sharing the work between an operating system (OS) and an application, or by cooperation between the OS and an application, only the parts other than the OS may be stored on the recording medium.

[0075] Furthermore, each program can be superimposed on a carrier wave and distributed over a network. For example, the program can be posted on a bulletin board system (BBS) on the network and distributed over the network. These programs can then be started and run under the control of the OS in the same way as other application programs, thereby enabling the above-mentioned processing to be performed. [Industrial Applicability]

[0076] The present disclosure is suitable as a system that suggests to a user a cooking recipe as well as a type of rice that is suitable for a side dish corresponding to the cooking recipe. [Explanation of symbols]

[0077] 1 Cloud server, 2 Rice cooker, 3 Terminal device, 6 BBR, 21 Heating unit, 22 Timekeeping unit, 23 Temperature measurement unit, 24 Rice cooker control unit, 29 Communication adapter, 91 EC site management server, 92 Model management server, 101, 301 CPU, 102, 302 Main memory unit, 103, 303 Auxiliary memory unit, 106, 306 Wide area communication unit, 108 Timekeeping unit, 109, 309 Bus, 111 Rice cooking setting acquisition unit, 112 Purchase information acquisition unit, 113 User information acquisition unit, 114 Family composition estimation unit, 115 Rice variety recipe estimation unit, 116 Rice variety recipe selection unit, 117 Rice variety recipe notification unit, 118 Presentation time identification unit, 119 Presentation time notification unit, 120 Viewed recipe acquisition unit, 121 Rice variety recipe estimation model generation unit, 122 Teacher information acquisition unit, 123 family structure estimation model generation unit, 131 rice cooking setting storage unit, 132 purchase history storage unit, 133 user information storage unit, 134 viewed recipe storage unit, 135 model storage unit, 136 recipe map storage unit, 137 rice variety recipe storage unit, 138 teacher information storage unit, 304 display unit, 305 input unit, 311 presentation time acquisition unit, 312 rice variety recipe acquisition unit, 313 image generation unit, 314 display control unit, 315 reception unit, 316 registration request unit, 317 viewed recipe notification unit, 318 ordering unit, 319 rice variety recipe identification unit, 321 display image storage unit, 331 rice variety recipe storage unit, 332 image storage unit, BU1, BU2 button images, GA1, GA21, GA22 recipe presentation image, GA3 rice variety presentation image, H House, L110, L210 input layer, L120, L220 hidden layer, L130, L230 output layer, M21 display image, NW1 wide area network, NW2 local area network

Claims

1. a family structure estimation unit that estimates the family structure of a user using a family structure estimation model that estimates the family structure of the user from rice cooking amount information that indicates the amount of rice used in one rice cooking process in a rice cooker used by the user, purchase history information that indicates the user's past purchase history through e-commerce, and user information including the user's gender and age; and a rice variety recipe estimation unit that estimates the rice variety and the cooking recipe using a rice variety recipe estimation model that estimates a combination of rice variety and cooking recipe from the cooked rice amount information, the user information, family composition information indicating the family composition, and recipe identification information that identifies cooking recipes that the user has previously viewed. Recipe presentation device.

2. The method further includes a viewed recipe storage unit that stores the recipe identification information viewed by the user in chronological order, When the recipe identification information stored in the browsed recipe storage unit does not exist, the rice variety recipe estimation unit identifies the rice variety and the cooking recipe based on preset recipe map information. The recipe presentation device according to claim 1 .

3. The user information includes residential area information indicating an area in which the user resides. The recipe presentation device according to claim 1 or 2.

4. The rice cooker further includes a presentation time specification unit that calculates an average time of past rice cooking start times at which the rice cooking process was started in the rice cooker, and specifies a time point that is a predetermined time before the calculated average time as a presentation time at which to present recipe information indicating at least the cooking recipe. The recipe presentation device according to claim 1 .

5. the presentation time specification unit classifies the past rice cooking start times into three time periods and specifies the presentation time for each of the three time periods; The recipe presentation device according to claim 4 .

6. a terminal device having a display unit; a family structure estimation unit that estimates the family structure of a user using a family structure estimation model that estimates the family structure of the user from rice cooking amount information that indicates the amount of rice used in one rice cooking process in a rice cooker used by the user, purchase history information that indicates the user's past purchase history through e-commerce, and user information including the user's gender and age; a rice variety recipe estimation unit that estimates the rice variety and the cooking recipe using a rice variety recipe estimation model that estimates a combination of rice variety and cooking recipe from the cooked rice amount information, the user information, family composition information indicating the family composition, and recipe identification information that identifies cooking recipes that the user has previously viewed; and a rice variety recipe notification unit that transmits rice variety information indicating the estimated rice variety and recipe information indicating the estimated cooking recipe to the terminal device. Recipe presentation system.

7. The terminal device displays, on the display unit, a recipe presentation image representing the rice variety combined with the cooking recipe indicated by the recipe information. The recipe presentation system according to claim 6 .

8. A recipe presentation system includes a step of estimating the user's family structure using a family structure estimation model that estimates the user's family structure from rice cooking amount information indicating the amount of rice to be cooked in one cooking process in the rice cooker used by the user, purchase history information indicating the user's past purchase history through e-commerce, and user information including the user's gender and age; the recipe presentation system includes a step of estimating the rice variety and the cooking recipe using a rice variety recipe estimation model that estimates a combination of rice variety and cooking recipe from the cooked rice amount information, the purchase history information, the user information, family composition information indicating the family composition, and recipe identification information that identifies cooking recipes that the user has viewed in the past, How to present a recipe.

9. Computer, a family structure estimation unit that estimates the user's family structure using a family structure estimation model that estimates the user's family structure from rice cooking amount information that indicates the amount of rice used in one rice cooking process using a rice cooker used by the user, purchase history information that indicates the user's past purchase history through e-commerce, and user information including the user's gender and age; a rice variety recipe estimation unit that estimates the rice variety and the cooking recipe using a rice variety recipe estimation model that estimates a combination of rice variety and cooking recipe from the cooked rice amount information, the user information, family composition information indicating the family composition, and recipe identification information that identifies cooking recipes that the user has viewed in the past; A program to function as a

Images