Information processing systems, information processing methods, and programs

Abstract

This invention provides an information processing system, information processing method, and program that can meet diverse needs in product recommendations displayed on checkout screens. [Solution] In the first extraction step, the information processing device extracts a first recommended product group from the product group of the population handled by the online supermarket using a first algorithm. In the second extraction step, it extracts a second recommended product group from the product group of the population using a second algorithm different from the first algorithm. In the display control step, it displays the first recommended product group with priority over the second recommended product group on the recommendation screen that is displayed to the target user before transitioning from the cart confirmation screen to the payment confirmation screen. The first algorithm is an algorithm based on the user information of the target user using the online supermarket.

Description

Technical Field

[0001] The present disclosure relates to an information processing system, an information processing method, and a program.

Background Art

[0002] In online shopping, a function of recommending products that a consumer is likely to purchase is known. For example, Patent Document 1 discloses an information processing apparatus that performs a recommendation process based on a user's behavior history for each processing reference period, a product category, and products for which the user has performed a predetermined action during the same period.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, in an online supermarket where it is assumed that a user regularly purchases various types of products, it is desired to recommend products that meet various user needs such as preventing forgetting to buy and impulse buying. In particular, when recommending products on the so-called pre-checkout screen, which is the screen before proceeding to the screen for payment after the user has placed products in the cart, it is desired to recommend products that meet various user needs. However, the information processing apparatus described in Patent Document 1 had room for improvement in order to sufficiently meet such various needs specific to online supermarkets.

[0005] An object of the present disclosure is to provide an information processing system, an information processing method, and a program that can meet various needs in product recommendation on a pre-checkout screen in view of the above problems.

Means for Solving the Problems

[0006] An information processing system according to one aspect of this disclosure comprises one or more processors and memory. The one or more processors are capable of performing the following steps: in a first extraction step, extract a first recommended product group from a population of products handled by the online supermarket using a first algorithm; in a second extraction step, extract a second recommended product group from the population of products using a second algorithm different from the first algorithm; and in a display control step, display the first recommended product group with priority over the second recommended product group on a recommendation screen that is displayed to the target user before transitioning from the cart confirmation screen to the payment confirmation screen. Here, the first algorithm is an algorithm based on user information of the target user using the online supermarket.

[0007] An information processing method according to one aspect of this disclosure is a method by which an information processing system performs each of the steps described above.

[0008] A program according to one aspect of this disclosure is a program that causes a computer having one or more processors and memory to perform each of the steps described above. [Effects of the Invention]

[0009] This disclosure provides an information processing system, information processing method, and program that can meet diverse needs in product recommendations on checkout screens. [Brief explanation of the drawing]

[0010] [Figure 1] This is a block diagram showing the overall structure of the information processing system. [Figure 2] This is a block diagram showing the hardware configuration of an information processing device. [Figure 3] Block diagram showing the hardware configuration of headquarters terminals and user terminals. [Figure 4] This is a block diagram showing the functional configuration of the control unit of the information processing device according to Embodiment 1. [Figure 5]It is a sequence diagram showing the registration process flow. [Figure 6] It is a diagram showing an example of the data structure of the product database. [Figure 7] It is a diagram showing an example of the data structure of the member database. [Figure 8] It is a diagram showing an example of the data structure of the action history database. [Figure 9] It is a flowchart showing the recommendation process flow. [Figure 10] It is a diagram showing an example of the cart confirmation screen. [Figure 11] It is a diagram showing an example of the recommendation screen. [Figure 12] It is a block diagram showing the functional configuration of the control unit of the information processing apparatus according to Embodiment 2. [Figure 13] It is a diagram showing an example of the recommendation screen.

Mode for Carrying Out the Invention

[0011] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In each drawing, the same or corresponding elements are denoted by the same reference numerals, and redundant descriptions are omitted as necessary for clarity of explanation.

[0012] <Embodiment 1> First, Embodiment 1 of the present disclosure will be described. 1. Hardware Configuration In this section, the hardware configuration of the information processing system 1 according to this embodiment will be described. FIG. 1 is a block diagram showing the overall configuration of the information processing system 1 according to Embodiment 1. In the following description, the processor is synonymous with a circuit (Circuitry) composed of a central processing unit or the like.

[0013] 1.1 Information Processing System 1 The information processing system 1 is a computer system that recommends products in an online supermarket. The information processing system 1 includes an information processing device 2, a headquarters terminal 3, and one or more user terminals 4. The information processing device 2, the headquarters terminal 3, and the user terminal 4 are communicably connected to each other via a network NW such as the Internet or a LAN (Local Area Network).

[0014] 1.2 Information Processing Device 2 The information processing device 2 is a computer such as a server that controls communication between the application of the online supermarket installed on the user terminal 4 and communication with the headquarters terminal 3, and executes processing in response to requests from the user terminal 4 and the headquarters terminal 3. Hereinafter, the application of the online supermarket is referred to as the online supermarket application. For example, the information processing device 2 registers products in response to a request from the headquarters terminal 3. Further, the information processing device 2 executes membership registration in response to a request from the user terminal 4 and records the behavior history of the user operating the user terminal 4 on the online supermarket. Further, the information processing device 2 recommends products to the user on a recommendation screen before displaying a payment confirmation screen for confirming payment information after transitioning from a cart confirmation screen for confirming the contents of the cart before payment in the online supermarket. Note that the recommendation screen is also referred to as a pre-checkout screen.

[0015] The information processing device 2 FIG. 2 is a block diagram showing the hardware configuration of the information processing device 2 according to Embodiment 1. The information processing device 2 includes a control unit 21, a storage unit 22, and a communication unit 23, and these components are electrically connected to each other inside the information processing device 2 via a communication bus 26.

[0016] The control unit 21 performs processing and control of the overall operation related to the information processing device 2. The control unit 21 is a processor, such as a central processing unit (CPU). The control unit 21 realizes various functions related to the information processing device 2 by reading predetermined programs stored in the memory unit 22. That is, information processing instructions stored in the memory unit 22 can be executed as various functional units included in the control unit 21 by being specifically realized by the control unit 21, which is an example of hardware. The functional configuration will be described in more detail in the next section. Note that the control unit 21 is not limited to being a single unit, and may be implemented with multiple control units 21 for each function, or a combination thereof.

[0017] The storage unit 22 is a storage device including a non-temporary computer-readable medium or a physical storage medium, and stores various types of information as defined above. The storage unit 22 may be implemented, for example, as a storage device such as a solid-state drive (SSD) that stores various programs related to the information processing device 2 executed by the control unit 21, or as a memory such as random access memory (RAM) that stores temporarily necessary information (arguments, arrays, etc.) related to program calculations. The storage unit 22 stores various programs, variables, etc. related to the information processing device 2 executed by the control unit 21.

[0018] The communication unit 23 is a communication interface for connecting to the network NW.

[0019] 1.3 Headquarters Terminal 3 Headquarters Terminal 3 is a computer terminal operated by a headquarters user. A headquarters user is a person in charge of the headquarters of a supermarket that operates an online supermarket, particularly a person in charge of product procurement or inventory management. Headquarters Terminal 3 is a personal computer, smartphone, tablet, or other input / output communication terminal.

[0020] Figure 3(A) is a block diagram showing the hardware configuration of the headquarters terminal 3. The headquarters terminal 3 comprises a control unit 31, a storage unit 32, a communication unit 33, an input unit 34, and a display unit 35, and these components are electrically connected within the headquarters terminal 3 via a communication bus 36. The control unit 31 is a processor such as a CPU. The hardware configuration of the control unit 31, storage unit 32, and communication unit 33 is substantially the same as that of the control unit 21, storage unit 22, and communication unit 23 in the information processing device 2 described above, so a detailed explanation is omitted.

[0021] The input unit 34 may be included in the housing of the headquarters terminal 3, or it may be an external unit. For example, the input unit 34 may be integrated with the display unit 35 and implemented as a touch panel. Of course, instead of a touch panel, switch buttons, a mouse, a QWERTY keyboard, etc., may be used. In other words, the input unit 34 receives operation input made by the headquarters user. This input is transmitted as a command signal to the control unit 31 via the communication bus 36, and the control unit 31 can perform predetermined controls and calculations as needed.

[0022] The display unit 35 may be included in the housing of the headquarters terminal 3, for example, or it may be an external unit. The display unit 35 displays a graphical user interface (GUI) screen that can be operated by the headquarters user. Preferably, the display unit 35 uses a display device such as a CRT display, liquid crystal display, organic EL display, and plasma display, depending on the type of headquarters terminal 3.

[0023] 1.4 User Terminal 4 User terminal 4 is a computer terminal operated by the online supermarket user (end user). User terminal 4 is a smartphone running the online supermarket app, or a personal computer that uses the online supermarket service via a web browser.

[0024] Figure 3(B) is a block diagram showing the hardware configuration of the user terminal 4. The user terminal 4 comprises a control unit 41, a storage unit 42, a communication unit 43, an input unit 44, and a display unit 45, and these components are electrically connected within the user terminal 4 via a communication bus 46. The hardware configuration of the control unit 41, storage unit 42, communication unit 43, input unit 44, and display unit 45 is substantially the same as that of the control unit 31, storage unit 32, communication unit 33, input unit 34, and display unit 35 in the headquarters terminal 3 described above, so a detailed explanation is omitted.

[0025] 2. Functional Configuration This section describes the functional configuration of Embodiment 1, particularly the functional configuration of the information processing device 2. The information processing method described later is executed by the information processing device 2. Figure 4 is a block diagram showing the functional configuration of the control unit 21 of the information processing device 2 according to Embodiment 1. The control unit 21 implements the functions of the registration unit 211, data processing unit 212, first extraction unit 213, second extraction unit 214, and display control unit 216 by executing a program.

[0026] The registration unit 211, as part of the registration step, retrieves product data from the headquarters terminal 3 and registers it in the product database T1, which will be described later. The registration unit 211 also receives member registration requests from the user terminal 4 and registers the user's attribute information (hereinafter sometimes referred to as user attributes) in the member database T2, which will be described later. Furthermore, the registration unit 211 registers the user's activity history in the online supermarket application run by the user terminal 4 in the activity history database T3, which will be described later.

[0027] The data processing unit 212 determines the group of products in the population to be recommended as a data processing step. The data processing unit 212 also excludes products to be excluded from display as an exclusion step, either from the group of products in the population or from the recommended product groups extracted by the first and second extraction steps described later. The excluded products include at least one of the first, second, and third excluded products described later. Furthermore, the data processing unit 212 transmits payment information based on the confirmed cart information to a payment server (not shown) in response to the user confirming the cart information on the cart confirmation screen and then receiving a confirmation instruction from the user on the payment confirmation screen via the recommendation screen. The cart information is information about the products intended for purchase selected by the user and held in the cart, i.e., the products in the cart. For example, the cart information includes the identification information, unit price, and quantity of the products in the cart. The payment information is the information necessary for payment, such as the total amount obtained by summing the amounts of the products in the cart.

[0028] The first extraction unit 213, as the first extraction step, extracts a first recommended product group from the product group of the population using the first algorithm. The first algorithm is one of the algorithms for extracting recommended products to present to users of the online supermarket (hereinafter sometimes referred to as target users), and is a personalized model based on the user information of the target users. Specifically, the first algorithm is a model that extracts products from the product group of the population based on user information. User information includes user attributes and at least one of the following: online supermarket purchase history or display history on the app.

[0029] The second extraction unit 214, as the second extraction step, extracts a second set of recommended products from the product group of the population using the second algorithm. The second algorithm is one of the algorithms for extracting recommended products to present to the target user, and unlike the first algorithm, it is not based solely on the user information of the target user.

[0030] The display control unit 216, as a display control step, causes the user terminal 4 to display an input area for information necessary for information processing by the control unit 21 or the result of information processing. For example, the display control unit 216 displays a recommendation screen on the user terminal 4 before transitioning from the cart confirmation screen to the payment confirmation screen for the target user. The recommendation screen is an example of a pre-checkout screen. Here, on the recommendation screen, the display control unit 216 displays information about the products included in each recommended product group extracted by the first extraction unit 213 and the second extraction unit 214, based on a predetermined display priority for each recommended product group. If the exclusion step is executed by the data processing unit 212 at this time, the display control unit 216 displays information about each product included in the first recommended product group and the second recommended product group, excluding the excluded products, on the recommendation screen. In this embodiment 1, the first priority, which is the display priority of the first recommended product group, is higher than the second priority, which is the display priority of the second recommended product group. In other words, the display control unit 216 causes the first recommended product group to be displayed on the recommendation screen with priority over the second recommended product group.

[0031] Furthermore, the display control unit 216 causing information to be displayed on the user terminal 4 may mean that the display control unit 216 transmits screen data showing that information to the user terminal 4, or it may mean that the display control unit 216 transmits data necessary to generate the screen showing that information to the user terminal 4. Also, displaying information is one form of outputting information.

[0032] 3. Operation of Information Processing Device 2 Section 3 will explain the flow of information processing performed by the information processing device 2, with reference to the diagram. The information processing includes registration processing and recommendation processing.

[0033] 3.1 Registration Process First, we will explain the overview of the registration process performed by Information Processing System 1. Figure 5 is a sequence diagram showing the flow of the registration process. First, the headquarters terminal 3 sends a registration request for product data to the information processing device 2 (S101). The registration request includes product data such as product ID, product name, product attributes, price, and inventory quantity. Upon receiving the registration request, the registration unit 211 of the information processing device 2 registers the product data in the product database T1, which will be described later (S102).

[0034] Next, the user terminal 4 sends a member registration request to the information processing device 2 (S103). The member registration request includes the user ID and user attributes. Upon receiving the registration request, the registration unit 211 registers the user attributes in the member database T2, which will be described later (S104).

[0035] When a user takes action on the online supermarket using the user terminal 4, the user terminal 4 transmits information indicating that action to the information processing device 2 (S105). Actions on the online supermarket refer to actions within the online supermarket application, and include, for example, adding items to the cart, purchasing items, removing items from the cart, as well as requests for and actual display of product information to the user terminal 4. Upon receiving the information, the registration unit 211 registers that action as an action history in the user's action history database T3, which will be described later (S106).

[0036] 3.2 Data Structures Next, the structure of the database used in this embodiment 1 will be described. [Product Database (Product DB)] Figure 6 shows an example of the data structure of the product database T1. The product database T1 is stored in the storage unit 22. The product database T1 has a record for each product that associates the product ID T10, product name T11, product attributes T12, price T13, and inventory quantity T14. Product ID T10 is information that identifies the product. Product name T11 is the name of the product. Product attribute T12 is information that represents attributes such as the type or nature of the product. Product attribute T12 may have a hierarchical structure and may include, for example, major and minor categories. For example, the major category indicates the items handled at the sales floor level and is a departmental classification related to the division of sales floor staff or ordering responsibilities. As an example, the major category may be the departmental code of the POS code and may indicate produce, meat, fresh fish, prepared foods, or daily necessities. On the other hand, the minor category is a subdivision of the departmental classification and is a classification related to shelf allocation. As an example, the minor category may be the category code of the POS code and may indicate pork, tofu, bread, or lactic acid bacteria beverages. Price T13 is the selling price of the product. Inventory count T14 is the number of items remaining in stock on the day the transaction is taking place.

[0037] [Member Database (Member DB)] Figure 7 shows an example of the data structure of the member database T2. The member database T2 is stored in the storage unit 22. The member database T2 has a record for each user that associates a user ID T20 with user attributes T21. The user ID T20 is information that identifies the user. For example, the user ID T20 may be the name, telephone number, email address, or the user's membership number. The user attributes T21 are attribute information of the user. User attributes T21 include information such as age, gender, address, and family structure.

[0038] [Behavioral History Database (Behavioral History DB)] Figure 8 shows an example of the data structure of the activity history database T3. The activity history database T3 is stored in the storage unit 22 for each user. The activity history database T3 has a record that associates each user action with the date and time T30, session ID T31, action type T32, and the product / page T33 targeted by the action. The date and time T30 is the date and time when the user action occurred. The session ID T31 is information used to identify a series of actions, i.e., a session, when a user uses the online supermarket app. Known techniques can be used for assigning session IDs. For example, the information processing device 2 assigns the same session ID to each action from the time it detects the user action until a certain period of time has passed without user operation. The action type T32 indicates the type of action the user took on the online supermarket, such as viewing a product page, adding a product to the cart, removing a product from the cart, and purchasing a product. The product / page T33 targeted by the action is information that identifies the product or page targeted by the action.

[0039] 3.3 Overview of the Recommendation Process Next, we will explain the overview of the recommendation process performed by Information Processing System 1. Figure 9 is a flowchart illustrating an example of the recommendation process flow. The recommendation process is performed for each user. The information processing device 2 first executes the data processing step, the first extraction step, and the second extraction step for each user. For example, the information processing device 2 may execute the above steps periodically (for example, once a day). Specifically, the data processing unit 212 of the information processing device 2 obtains the user ID of the target user using the member database T2 (S110). Then, the data processing unit 212 determines the group of products that will be the population for the recommendation process (S111).

[0040] Next, the first extraction unit 213 executes the first extraction step. Specifically, the first extraction unit 213 uses the first algorithm to extract a first recommended product group from the product group of the population based on the user ID of the target user (S112), and determines the display order of the products within the first recommended product group (S113).

[0041] Next, the second extraction unit 214 executes the second extraction step. Specifically, the second extraction unit 214 uses the second algorithm to extract a second recommended product group from the product group of the population (S114) and determines the display order of products within the second recommended product group (S115). When the target user starts an action on the online supermarket app, checks the products in the cart, and proceeds to payment, the user terminal 4 operated by the target user sends a payment confirmation request to the information processing device 2. When the information processing device 2 receives the payment confirmation request (YES in S116), the data processing unit 212 executes the exclusion step (S117). Specifically, the data processing unit 212 performs exclusion processing, such as removing products to be excluded from display from the first recommended product group and the second recommended product group. The information processing device 2 then proceeds to processing in S118. The information processing device 2 may repeat the reception confirmation process shown in S116 while retaining the contents of the cart until it receives a payment confirmation request. The information processing device 2 may terminate processing if it does not receive a payment confirmation request for a certain period of time while the product remains in the shopping cart.

[0042] Finally, the display control unit 216 executes the display control steps shown in S118 to S119. Based on the predetermined priority for each algorithm, the display control unit 216 determines the overall display order of each recommended product (S118), and displays information about the recommended products on the recommendation screen based on the determined display order (S119).

[0043] [effect] In this embodiment 1, the information processing device 2 uses a first algorithm personalized for the target user and other algorithms to recommend products to the target user on the checkout screen. At this time, the information processing device 2 displays the recommended products extracted from the personalized first algorithm with priority over other recommended products. Therefore, the information processing device 2 can respond to a variety of needs in product recommendations on the checkout screen, including the "prevention of forgetting to buy" specific to online supermarkets. Furthermore, because the information processing device 2 displays recommended products according to the priority of the algorithms, it can appropriately recommend products that the target user is likely to purchase at the checkout.

[0044] Furthermore, for users who have just registered or who use the service infrequently, the number of products that can be recommended by the first algorithm may be insufficient because there is little or no accumulated user information. However, according to the information processing device 2, in addition to the first algorithm, a second algorithm, which is an algorithm other than the personalized algorithm, is used to extract recommended products, so even in the above cases, a certain number of recommended products can be secured. As a result, the information processing device 2 can effectively encourage users to make additional purchases at the checkout counter.

[0045] Furthermore, the products recommended by the second algorithm may include products that the target user has never purchased or products that are not very relevant to the target user. Therefore, by using the second algorithm to recommend products, the information processing device 2 makes it possible for the target user to accidentally discover and purchase products. In other words, the information processing device 2 can address the need for "preventing forgotten purchases" with the first algorithm and the need for "impulse purchases" with the second algorithm.

[0046] 3.4 Details of the recommendation process Next, we will explain the details of each step in the recommendation process. [Data processing step (S111)] In S111, the data processing unit 212 determines the product group of the population. Specifically, the data processing unit 212 may determine the product group of the population to be all products that are handled by the online supermarket, that is, all products registered in the product database T1. Alternatively, the data processing unit 212 may determine the population by excluding predetermined exclusion products from the products that are handled by the online supermarket (first exclusion step).

[0047] The products excluded from display in this case may be first-class products that have predetermined attributes. For example, first-class products are products such as medicines or alcohol that require user action such as age verification at the time of purchase, or products that are not appropriate to display to certain users for religious reasons. The products excluded from display may also be second-class products that are currently or will not be available due to discontinuation of sales or stock shortages, i.e., products that are out of stock or will be out of stock soon. By excluding these products from the population in advance, unnecessary recommendations that impair the user experience can be avoided, and the burden of subsequent processing can be reduced.

[0048] [First extraction step (S112, S113)] In S112, the first extraction unit 213 extracts the first recommended product group using the first algorithm. (First Algorithm) The first algorithm is a personalized model that determines recommended products based on the user information of the target user, with the primary purpose of preventing customers from forgetting to buy items. The user information entered into this algorithm includes at least one of the following: the target user's user attributes, purchase history, and other behavioral history on the online supermarket (view history, cart addition history, etc.). For example, the first algorithm is a model that calculates the degree of relevance between the target user's user information and each product, and extracts products in descending order of relevance. The degree of relevance here can be measured by metrics such as the number of purchases, the rate at which products are added to the cart, the probability of purchase, the purchase experience, the frequency of purchases, the number of days since the last purchase, the purchase amount, the time spent on the page, the number of impressions or the frequency of impressions, or a combination of these. As an example, the first algorithm extracts products from the target user's purchase history that have been purchased many times, in descending order of purchase frequency. As another example, the first algorithm extracts products related to pages displayed to the target user that have an impression rate, number of purchases, or purchase frequency above a predetermined value. As yet another example, the first algorithm extracts products related to pages where the target user has spent a predetermined amount of time or more, or products related to pages where the number of impressions or the frequency of impressions above a predetermined value. The above-mentioned relevance index is calculated from behavioral history such as purchase history. The above-mentioned relevance index may be calculated by giving more weight to recent behavioral history than to older behavioral history. For example, the first algorithm may calculate the relevance of a target user's behavioral history within a predetermined period with a higher weight than behavioral history from other periods, and then extract products based on the relevance.

[0049] Here, the first algorithm may be a rule-based model that extracts products based on the above relevance, or it may be a machine learning model that learns relevance features for each user. For example, the first algorithm may be a machine learning model that learns the number of purchases each user has made. Any machine learning model can be used, such as a neural network model or a regression model.

[0050] The number of products extracted by the first algorithm may be limited (for example, 100). In this case, the first extraction unit 213 may rank the products by relevance and select a predetermined number of products from the top of the list for extraction. Setting a limit helps to avoid unnecessary recommendations, such as recommending products that are not highly relevant to the target user.

[0051] As for the specific processing in S112, the first extraction unit 213 first retrieves the user attributes of the target user from the member database T2 using the target user's user ID as the key, and retrieves the target user's behavioral history, including purchase history, from the behavioral history database T3 corresponding to the target user. Next, the first extraction unit 213 uses the first algorithm to extract products that are predicted to have a high purchase probability for the target user, based on the acquired user information, as the first recommended product group.

[0052] In S113, the first extraction unit 213 determines the display order within the first recommended product group. At this time, the first extraction unit 213 determines the display order for each product included in the first recommended product group based on the degree of relevance between the product and the user information. For example, by arranging products in order of purchase probability or cart addition rate based on the target user's behavior history, products that the target user is likely to be interested in can be recommended in a more visible position. The relevance used for extraction and the relevance used for determining the display order may be based on the same or different underlying indicators. In the latter case, for example, the number of purchases may be used for extraction, and the cart addition rate may be used for determining the display order.

[0053] [Second extraction step (S114, S115)] In S114, the second extraction unit 214 extracts a second group of recommended products using the second algorithm. The first algorithm alone has the problem of not being able to make sufficient recommendations when the target user has little purchase history, known as the cold start problem, or it is difficult to meet the need to satisfy the need for users to encounter products they have never purchased before, known as serendipity. Therefore, by using the second algorithm, a variety of needs such as impulse purchases can be addressed.

[0054] (Second algorithm) For example, the second algorithm is an algorithm that extracts a second set of recommended products based on at least one of the following for each product in the population set during a predetermined period: sales information, profit information, and pre-payment purchase probability. Sales information may be, for example, sales amount or number of units sold. Profit information may be, for example, gross profit amount or gross profit margin. The pre-payment purchase probability indicates the probability that a product whose information is displayed on the recommendation screen is subsequently added to the cart and purchased. The second algorithm is a machine learning model and may be reinforced through repeated extraction of second recommended products in order to maximize the rate of adding items to the cart before checkout. For example, the second extraction unit 214 uses a second algorithm to extract the top predetermined number of best-selling products (e.g., the top 200) from among all users, in order of sales volume. This allows for the recommendation of products that are not personalized but are likely to interest many users, thereby encouraging impulse purchases.

[0055] Alternatively, the second algorithm could be one that identifies similar users to the target user and extracts a second set of recommended products based on the user information of those similar users. One example of a second algorithm is a machine learning model that utilizes collaborative filtering. The second algorithm identifies similar users with similar user attributes to the target user and extracts a second group of recommended products based on the user information of these similar users. The user information used as the basis for extraction may be user information excluding the user attributes used to identify similar users, and may particularly include purchase history. This allows for the suggestion of products that are not directly related to the target user's past behavior but are likely to be of interest, thereby enabling the meeting of a wider range of needs.

[0056] Alternatively, the second algorithm may be one that extracts a second set of recommended products based on the items held in the target user's shopping cart. For example, the second algorithm extracts products that are highly related to the items in the shopping cart. Products with a high degree of relevance are, for example, products that are likely to be purchased together. Products with a high degree of relevance to the items in the shopping cart may be products whose relevance to the items in the shopping cart is above a predetermined value, or the top predetermined number of products in terms of relevance. This allows for the appropriate recommendation of products to buy together and further promotes impulse purchases.

[0057] The degree of relevance may be calculated based on product attributes or the purchase history of all users. The information processing device 2 may have a pre-stored list of highly related product combinations in the product database T1, and the second algorithm may refer to the product database T1 to identify highly related products to the products in the cart. Alternatively, the second algorithm may calculate the degree of relevance between the products in the cart and each product in the population each time it is executed, thereby identifying highly related products. The combination of highly related products may be determined from a cooking recipe. For example, carrots, onions, potatoes, and curry roux may be defined as a highly related combination of ingredients for curry. In this case, if the products in the cart include carrots, onions, and curry roux, the second algorithm may output potatoes as a recommended purchase item.

[0058] It is also possible to perform the second exclusion step, described later, within the second algorithm instead of the data processing unit 212. For example, the second algorithm may exclude the third display exclusion products, which are products already held in the target user's cart, from the second recommended product group extracted by any of the methods described above or by any other method, and determine this as a new second recommended product group. In such a case as well, the second algorithm can be said to be an algorithm that extracts the second recommended product group based on the products held in the target user's cart.

[0059] In S115, the second extraction unit 214 determines the display order within the second recommended product group. This display order may be determined randomly. This allows for serendipitous discoveries for the target user and fulfills the expectation of serendipity. Alternatively, the second extraction unit 214 may divide the extracted second recommended product group into multiple groups based on relevance, randomly determine the display order within each group, and determine the display order between groups based on relevance. This makes it possible to place relatively popular products in positions easily visible to the target user while still fulfilling the expectation of serendipity to some extent.

[0060] [Exclusion step (S117)] The data processing unit 212 further removes excluded products from each of the extracted first and second recommended product groups (second exclusion step). The excluded products removed at this stage are third excluded products, which are products already held in the target user's shopping cart. This avoids recommending products that the user has already decided to purchase, thereby improving the user experience. The data processing unit 212 may also remove second excluded products from each recommended product group at this stage. This allows for the removal of out-of-stock or soon-to-be-out-of-stock products in real time, further improving the user experience. Alternatively, the data processing unit 212 may remove first excluded products from each recommended product group in the second exclusion step instead of the first exclusion step.

[0061] Furthermore, if there are duplicate products in the first recommended product group and the second recommended product group, the data processing unit 212 may remove the duplicate product from the recommended product group with the lower priority. In this embodiment 1, since the first priority is higher than the second priority, the data processing unit 212 removes the duplicate product from the second recommended product group. Therefore, in the display control step described later, the display control unit 216 displays information about the first recommended product group and information about the second recommended product group from which the duplicate product has been removed on the recommendation screen. This avoids unnecessary duplicate recommendations and makes effective use of the limited display space. Also, by recommending the duplicate product within the framework of the first recommended product group, it can be recommended in a position that is easily noticeable to the user.

[0062] [Display control step (S118, S119)] Figure 10 shows an example of the cart confirmation screen 61 before the recommendation screen is displayed. The cart confirmation screen 61 displays a list of items 612 and 613 in the cart, and cart summary information 614. The cart summary information 614 includes information such as the total amount, total number of items, and shipping fee for the items in the cart. The cart confirmation screen 61 may also further include an object 611 that shows the steps to payment completion and the current position within those steps, and a button 616 that allows the user to return to various product pages and continue shopping at the online supermarket. When the user operates the "Proceed to Checkout" button 615, the user terminal 4 sends a payment confirmation request to the information processing device 2. If the information processing device 2 receives the payment confirmation request (YES in S116), the data processing unit 212 executes an exclusion step (S117), and the display control unit 216 transitions the screen to the recommendation screen (S118-S119).

[0063] Figure 11 shows an example of the recommendation screen 71. In S118, the display control unit 216 determines the overall display order of each recommended product. This order is determined based on the priority set in advance for each algorithm. In this embodiment 1, the first priority of the first recommended product group is set higher than the second priority of the second recommended product group. Therefore, as shown in Figure 11, the display control unit 216 displays the information of the second recommended product group P2 after the information of the first recommended product group P1. As a result, personalized recommended products (P1) that are expected to have a higher purchase probability are displayed higher than general recommended products (P2). The display order of each product in the first recommended product group P1 and the display order of each product in the second recommended product group P2 are as determined in S113 and S115.

[0064] The display control unit 216 may also display the recommendation screen 71 including the reward information 712. The reward information 712 indicates the rewards that the target user may receive and the conditions required to receive those rewards, based on the status of the target user's shopping cart, for example, the items held in the cart. For example, the reward information 712 may indicate a reward such as "free delivery if you purchase an additional 1,083 yen" or "double points if you purchase 3 more items." This can effectively encourage the user to make additional purchases.

[0065] If the recommended products on the recommendation screen include products that are currently out of stock, the display control unit 216 may display information about similar alternative products on the recommendation screen instead of the product in question. In this case, if the data processing unit 212 includes a product in the recommended product group that is currently out of stock in the product database T1, it may identify products with a high degree of similarity to that product as alternative products. Products with a high degree of similarity may be products with a similarity score of a predetermined value or the top predetermined number of products in terms of similarity. The similarity score may be calculated based on product attributes or the purchase history of all users. This helps to avoid the loss of sales opportunities. The above process can be applied not only to products that are currently out of stock, but also to products that are scheduled to become out of stock.

[0066] The total number of recommended products displayed on the recommendation screen may be predetermined. For example, if the number of products that can be extracted by the first algorithm is less than a predetermined number, the display control unit 216 may adjust the number of products in the second recommended product group to ensure that the total number of displayed products is equal to the number of products that can be displayed. This ensures that the target user is always recommended a certain number of products, thus avoiding the cold start problem and improving the user experience. In this case as well, the display control unit 216 prioritizes displaying the first recommended product group on the recommendation screen over the second recommended product group. However, if there is no first recommended product group, the display control unit 216 may display a recommendation screen based on the second recommended product group. The above recommendation screen includes information about the second recommended product group but does not include information about the first recommended products. The absence of a first recommended product group may occur when the first extraction step using the first algorithm is executed but there is no first recommended product group to extract, or when the first extraction step is not executed because the user information does not meet the extraction conditions. For example, the extraction criteria include the target user's purchase history not being registered in the behavioral history database T3, or the target user having a purchase history but the last purchase date being more than three years ago, or other conditions outside the specified period.

[0067] The display order within the first recommended product group may be updatable in response to user actions. For example, the first extraction unit 213 may update the display order within the first recommended product group in response to a change in the session, such as when the target user logs in again, or in response to the target user performing an update operation using the update button placed on the recommendation screen. In this case, the first extraction unit 213 may rearrange the display order of at least some of the products within the first recommended product group so that it differs from the previous display order. For example, the first extraction unit 213 may update the display order so that the display order within the first recommended product group is random. Similarly, the display order of the second recommended product group may also be updateable in response to user interaction. In particular, the second extraction unit 214 may update the display order so that the display order within the second recommended product group is random. Here, while the order of products in the first recommended product group, which is related to features such as purchase frequency, tends to change from day to day, the order of the second recommended product group does not usually fluctuate significantly from day to day. Therefore, by randomly rearranging the display order, the target user is more likely to discover products that could be "impulse buys" sooner than if the display order were not rearranged. This can create a sense of anticipation for the target user and increase their motivation to browse the second recommended product group.

[0068] <Embodiment 2> Next, Embodiment 2 of this disclosure will be described. The information processing system 1 of Embodiment 2 is basically the same as the information processing system 1 of Embodiment 1 in terms of configuration and function, but differs in that the control unit of the information processing device 2 uses a third algorithm in addition to the first and second algorithms in the recommendation process.

[0069] Figure 12 is a block diagram showing the functional configuration of the control unit 21a of the information processing device 2 according to Embodiment 2. The control unit 21a is basically the same as the control unit 21 in terms of function, but differs from the control unit 21 in that it includes a third extraction unit 215.

[0070] The third extraction unit 215, as the third extraction step, extracts a third recommended product group from the product group of the population using a third algorithm. (Third Algorithm) The third algorithm is different from both the first and second algorithms. For example, the third algorithm identifies similar users to the target user and extracts a third group of recommended products based on the user information of those similar users. As an example, the third algorithm is a machine learning model that utilizes collaborative filtering. In the specific processing of the third extraction step, the third extraction unit 215 identifies similar users whose user attributes are similar to those of the target user, and extracts a third group of recommended products based on the user information of those similar users. The user information that forms the basis of the extraction may be user information excluding the user attributes used to identify similar users, and may particularly be purchase history. This makes it possible to suggest products that are not directly related to the target user's past behavior but are likely to be of interest, thereby meeting a wider range of needs.

[0071] Then, in the display control step, the display control unit 216 displays information about each product included in the third recommended product group on the recommendation screen based on the first priority, the second priority, and the third priority, which is the display priority for the third recommended product group. When the third algorithm is used, the third priority may be set lower than the first priority and higher than the second priority. In other words, the display control unit 216 displays the first recommended product group with priority over the third recommended product group, and displays the third recommended product group with priority over the second recommended product group.

[0072] Figure 13 shows an example of the recommendation screen 81 according to Embodiment 2. As shown in Figure 13, the first recommended product group P1 consists of personalized products, the third recommended product group P3 consists of products recommended based on similar users, and the second recommended product group P2 consists of randomly selected products.

[0073] Furthermore, among the third group of recommended products, low-relevance products, such as those that the target user has never purchased, may be displayed at a lower ranking than other products within the same third group. Low-relevance products may include not only products that the target user has never purchased, but also products whose relevance (such as purchase frequency) is below a predetermined value, products that fall outside the top predetermined relevance rankings, or products for which a predetermined period has passed since the target user last purchased. This allows us to provide new discoveries to highly motivated users who view the recommendation screen to the end.

[0074] While the present disclosure has been described above with reference to embodiments, it is not limited thereto. Various modifications to the structure and details of the present disclosure are possible, which can be understood by those skilled in the art within the scope of the invention.

[0075] For example, in embodiments 1 and 2 described above, the control units 21 and 21a used two to three algorithms to extract recommended products. However, the number of algorithms is not limited to this, and there may be four or more. At least some of these algorithms may include machine learning models, particularly language learning models such as large language models (LLMs). In this case as well, the display control unit 216 can make appropriate recommendations while responding to the diverse needs of users by displaying information about each product on the recommendation screen based on the display priority of the algorithms.

[0076] Furthermore, in the embodiments 1 and 2 described above, the product database T1, member database T2, and behavioral history database T3 were assumed to be stored in the storage unit 22, but they may also be stored in an external device accessible by the information processing device 2.

[0077] Furthermore, in embodiments 1 and 2 described above, the user's actions on the online supermarket were defined as the user's actions within the online supermarket app, but they may also refer to the user's actions when using the online supermarket via a browser. Specifically, actions on the online supermarket may refer to adding items to the cart, purchasing items, removing items from the cart, as well as requesting and displaying information about products on the user terminal 4.

[0078] Furthermore, in embodiments 1 and 2 described above, user information was assumed to include purchase history, but the purchase history may include not only purchase history on the online supermarket but also purchase history at physical stores. For example, the registration unit 211 of the information processing device 2 may acquire information from a store terminal (not shown) that associates the products purchased by a user who made a payment at a physical store with the user ID read from the point card presented by the user at the time of payment, and register this information in the user's behavior history database T3. In embodiments 1 and 2 described above, the information processing device 2 acquired product data from the headquarters terminal 3, but instead, at least a portion of the product data, such as the number of items in stock, may be acquired from the store terminal.

[0079] Furthermore, in embodiments 1 and 2 described above, the control units 21 and 21a determined the display order among the recommended products extracted using the same algorithm, then performed an exclusion process, and finally determined the overall display order based on the algorithm's priority. However, the timing for determining the display order among the recommended products extracted using the same algorithm is not limited to this, and may be after the exclusion process. Also, in embodiments 1 and 2 described above, since the control units 21 and 21a display information for the second recommended product group after the first recommended product group, a boundary existed between the display area for the first recommended product group and the display area for the second recommended product group. However, there does not need to be a clear boundary between the display area for the first recommended product group and the display area for the second recommended product group, and the information for the products in the second recommended product group may be placed within the display area for the first recommended product group. In this case, the control units 21 and 21a may determine the priority among the entire recommended products for each product in the first and second recommended product groups based on the relevance used in the algorithm corresponding to that product and the algorithm's priority, and determine this as the display order.

[0080] While Embodiments 1 and 2 describe a case where the information processing device 2 is a single device, some or all of the components of the information processing device 2 may be realized by multiple information processing devices or circuits. In this case, the multiple information processing devices or circuits may be centrally located or distributed. For example, the information processing devices or circuits may be realized in a form where each is connected via a communication network, such as a client-server system or a cloud computing system. Furthermore, the functions of the information processing device 2 may be provided in SaaS (Software as a Service) format.

[0081] Some or all of the above embodiments may also be described as follows, but are not limited to the following: (Note 1) An information processing system comprising one or more processors and memory, The one or more processors mentioned above are capable of performing the following steps: In the first extraction step, the first recommended product set is extracted from the product set of the population handled by the online supermarket using the first algorithm. The first algorithm is an algorithm based on user information of the target users who use the online supermarket, In the second extraction step, a second set of recommended products is extracted from the product group of the population using a second algorithm different from the first algorithm. In the display control step, the first recommended product group is displayed with priority over the second recommended product group on the recommendation screen that is displayed to the target user before they transition from the cart confirmation screen to the payment confirmation screen. Information processing system. (Note 2) The second algorithm is an algorithm that extracts the second recommended product group based on at least one of the following: sales information, profit information, and pre-purchase probability for each product included in the product group of the population during a predetermined period. The information processing system described in Appendix 1. (Note 3) The second algorithm identifies similar users to the target user and extracts the second recommended product group based on the user information of the similar users. The information processing system described in Appendix 1. (Note 4) The second algorithm extracts a second set of recommended products from the product group of the population based on the products the target user intends to purchase and which are currently in their shopping cart. An information processing system as described in any one of the items 1 to 3 of the appendix. (Note 5) The one or more processors, in the display control step, display information about each product included in the first recommended product group and the second recommended product group based on the first priority, which is the display priority of the first recommended product group, and the second priority, which is the display priority of the second recommended product group. The aforementioned first priority is higher than the aforementioned second priority. An information processing system as described in any one of the items 1 to 4 of the appendix. (Note 6) The aforementioned processor 1 or more In the display control step, if there are duplicate products in the first recommended product group and the second recommended product group, the duplicate products are removed from the second recommended product group whose information is displayed on the recommendation screen. An information processing system as described in any one of the items 1 to 5 of the appendix. (Note 7) The one or more processors, in the display control step, display information on the recommendation screen about each product included in the first recommended product group and the second recommended product group, excluding products excluded from display. The aforementioned excluded products include at least one of the following: a first excluded product having predetermined attributes, a second excluded product that is out of stock or expected to be out of stock, and a third excluded product that is a product intended for purchase held in the target user's shopping cart. An information processing system as described in any one of the items 1 to 6 of the appendix. (Note 8) The aforementioned one or more processors can further perform the exclusion step, In the exclusion step, the second or third product to be excluded from labeling is excluded from the extracted first and second recommended product groups. The information processing system described in Appendix 7. (Note 9) The user information input to the first algorithm includes at least one of the following: the attributes of the target user, purchase history, and behavioral history at the online supermarket. An information processing system as described in any one of the items 1 to 8 of the appendix. (Note 10) The one or more processors, in the display control step, The display order for each product included in the first recommended product group is determined based on the degree of relevance between that product and the user information. An information processing system as described in any one of the items 1 through 9 of the appendix. (Note 11) The one or more processors, in the display control step, The display order for each product included in the second recommended product group is determined randomly. An information processing system as described in any one of the items 1 through 10 of the appendix. (Note 12) The one or more processors, in the display control step, If the first recommended product group does not exist, or if the user information does not meet the extraction criteria, the recommendation screen will be displayed, which includes information about the second recommended product group but does not include information about the first recommended product group. An information processing system as described in any one of the items from Appendix 1 to 11. (Note 13) The aforementioned one or more processors are capable of performing the third extraction step, In the third extraction step, A third set of recommended products is extracted from the product group of the aforementioned population using a third algorithm that is different from both the first and second algorithms. In the display control step, the third recommended product group is displayed on the recommendation screen. The third algorithm is an algorithm that identifies similar users to the target user and extracts the third recommended product group based on the user information of the similar users. An information processing system as described in any one of the items 1 to 12 of the appendix. (Note 14) The one or more processors, in the display control step, cause the first recommended product group to be displayed with priority over the third recommended product group, and the third recommended product group to be displayed with priority over the second recommended product group. The information processing system described in Appendix 13. (Note 15) The one or more processors, in the display control step, Among the third group of recommended products, low-relevance products are given a lower display ranking than products other than the low-relevance products. The aforementioned low-relevance products are products that the target user has never purchased, products that the target user purchases less than a predetermined frequency, or products for which a predetermined period has elapsed since the last time the target user purchased them. The information processing system described in Appendix 13 or 14. (Note 16) The one or more processors, in the display control step, The recommendation screen displays the benefits that the target user may receive based on the products held in the target user's shopping cart, and the conditions required to receive those benefits. An information processing system as described in any one of the items 1 through 15 of the appendix. (Note 17) The one or more processors, in the display control step, If the first and second recommended product groups include products that are out of stock or scheduled to be out of stock, information about similar products will be placed on the recommendation screen in their place. An information processing system as described in any one of the items from Appendix 1 to 16. (Note 18) The information processing system performs each of the steps described in any one of the appendices 1 to 17. Information processing methods. (Note 19) A program for causing a computer having one or more processors and memory to perform each of the steps described in any one of the appendices 1 to 17. [Explanation of Symbols]

[0082] 1. Information Processing System 2. Information Processing Device 3. Headquarters terminal 4. User terminals 21, 21a Control Unit 22 Memory section 23 Communications Department 31, 41 Control Unit 32, 42 Storage section 33, 43 Communications Department 34, 44 Input section 35, 45 Display section 211 Registration Department 212 Data Processing Unit 213 1st extraction part 214 2nd extraction part 215 Third extraction part 216 Display Control Unit 61 Cart Confirmation Screen 71, 81 Recommendation screen 712 Bonus Information T1 Product Database T2 Member Database T3 Behavioral History Database P1 1st recommended product group P2 2nd recommended product group P3 3rd recommended product group

Claims

1. An information processing system comprising one or more processors and memory, The one or more processors mentioned above are capable of performing the following steps: In the first extraction step, the first recommended product group is extracted from the product group of the population handled by the online supermarket using the first algorithm. The first algorithm is an algorithm based on user information of the target users who use the online supermarket, The user information includes any of the following of the target user: user attributes, purchase history, and behavioral history at the online supermarket. In the second extraction step, a second set of recommended products is extracted from the product set of the population using a second algorithm different from the first algorithm. In the display control step, the first recommended product group is displayed with priority over the second recommended product group on the recommendation screen that is displayed to the target user before transitioning from the cart confirmation screen to the payment confirmation screen. The second algorithm is one of the following: (i) an algorithm that extracts the second recommended product group based on at least one of the following: sales information, profit information, and purchase probability of each product included in the product group of the population during a predetermined period; (ii) an algorithm that identifies similar users with similar user attributes to the target user and extracts the second recommended product group based on the purchase history of the similar users; and (iii) an algorithm that extracts the second recommended product group from the product group of the population based on the products the target user plans to purchase that are held in their shopping cart. Information processing system.

2. The one or more processors, in the display control step, display information about each product included in the first recommended product group and the second recommended product group based on the first priority, which is the display priority of the first recommended product group, and the second priority, which is the display priority of the second recommended product group. The first priority is higher than the second priority. The information processing system according to claim 1.

3. The one or more processors mentioned above are: In the display control step, if there are duplicate products in the first recommended product group and the second recommended product group, the duplicate products are removed from the second recommended product group whose information is displayed on the recommendation screen. The information processing system according to claim 1.

4. In the display control step, the one or more processors display information on the recommendation screen about each product included in the first recommended product group and the second recommended product group, excluding products excluded from display. The excluded products include at least one of the following: a first excluded product having predetermined attributes, a second excluded product that is out of stock or expected to be out of stock, and a third excluded product that is a product intended for purchase held in the target user's shopping cart. The information processing system according to claim 1.

5. The one or more processors mentioned above are: If the product group handled by the online supermarket includes the first product excluded from labeling, the product group of the population is determined by excluding the first product excluded from labeling before the first extraction step. If the extracted first recommended product group includes the second product to be excluded from labeling or the third product to be excluded from labeling, after the first extraction step, the second product to be excluded from labeling or the third product to be excluded from labeling is removed. If the extracted second recommended product group includes the second product to be excluded from labeling or the third product to be excluded from labeling, after the second extraction step, the second product to be excluded from labeling or the third product to be excluded from labeling is removed. The information processing system according to claim 4.

6. In the display control step, the one or more processors The display order for each product included in the first recommended product group is determined based on the degree of relevance between that product and the user information. The information processing system according to claim 1.

7. In the display control step, the one or more processors The display order for each product included in the second recommended product group is determined randomly. The information processing system according to claim 1.

8. In the display control step, the one or more processors If the first recommended product group does not exist, or if the user information does not meet the extraction criteria, the recommendation screen will be displayed, which includes information about the second recommended product group but does not include information about the first recommended product group. The information processing system according to claim 1.

9. The one or more processors mentioned above are capable of performing the third extraction step, In the third extraction step described above, A third set of recommended products is extracted from the product group of the aforementioned population using a third algorithm that is different from both the first and second algorithms. In the display control step, the third recommended product group is displayed on the recommendation screen. The second algorithm is either (i) an algorithm for extracting the second recommended product group based on at least one of the sales information, profit information, and purchase probability of each product included in the product group of the population during a predetermined period, or (iii) an algorithm for extracting the second recommended product group from the product group of the population based on the products intended for purchase held in the shopping cart of the target user. The third algorithm is an algorithm that (ii) identifies similar users whose user attributes are similar to the target user and extracts the third recommended product group based on the purchase history of the similar users, In the display control step, the first recommended product group is displayed with priority over the third recommended product group, and the third recommended product group is displayed with priority over the second recommended product group. The information processing system according to claim 1.

10. In the display control step, the one or more processors Among the third group of recommended products, low-relevance products are given a lower display ranking than products other than the low-relevance products. The aforementioned low-relevance products are products that the target user has never purchased, products that the target user purchases less than a predetermined frequency, or products for which a predetermined period has elapsed since the last time the target user purchased them. The information processing system according to claim 9.

11. In the display control step, the one or more processors The recommendation screen displays the benefits that the target user may receive based on the products held in the target user's shopping cart, and the conditions required to receive those benefits. The information processing system according to claim 1.

12. In the display control step, the one or more processors If the first and second recommended product groups include products that are out of stock or scheduled to be out of stock, information about similar products will be placed on the recommendation screen in their place. The information processing system according to claim 1.

13. The information processing system performs each step described in any one of claims 1 to 12. Information processing methods.

14. A program for causing a computer, which comprises one or more processors and memory, to perform each of the steps described in any one of claims 1 to 12.

Images