Information processing device, information processing method, and program

Abstract

To provide information for generating commodity information which arouses commodity purchase motivation of a user.SOLUTION: Commodity information of an object commodity is acquired, having a feature regarding the object commodity in a commodity page of the object commodity being sold in an electronic commerce (EC) site. Among one or more other commodities which are sold in the EC site and satisfy a prescribed condition, one or more similar commodities being similar to the object commodity are specified. Commodity information of the one or more similar commodities having features of the one or more similar commodities is acquired in the commodity pages of the one or more similar commodities. The commodity information of the object commodity and the commodity information of the one or more similar commodity are compared. An improvement point for the commodity page of the object commodity is determine on the basis of the comparison result, and the improvement point is presented to a user.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to an information processing device, an information processing method, and a program, and in particular to a technique for generating product information. [Background technology]

[0002] In recent years, electronic commerce (EC) has become widespread. On EC sites, users tend to decide whether or not to purchase a product by referring to information about the product being sold or exhibited (product information). Therefore, it is important to present effective product information to users.

[0003] Patent Document 1 discloses a method for searching image data of one or more similar products similar to the product from image data of the product, acquiring access information associated with the image data of the similar products, and generating product information of the product using the access information. Here, the access information refers to information used to access media data (image data, audio data, etc.) representing the product from multiple terminals, and includes search log information such as search keywords. In this document, for example, the product information can be generated by including the search keywords obtained from the access information in the product name and category of the product information of the product. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2019-75019 A Summary of the Invention [Problem to be solved by the invention]

[0005] The above-mentioned document has a certain effect in that it is possible to generate product information for a product that is comprehensively determined by multiple terminals by using access information associated with image data of similar products similar to the product. However, in the past, when creating product information, it was not possible to grasp what kind of product information would be effective, and it could not be said that product information that stimulates users' desire to purchase the product was generated.

[0006] The present invention has been made in consideration of the above-mentioned problems, and has an object to provide information for generating product information that stimulates a user's desire to purchase a product. [Means for solving the problem]

[0007] In order to solve the above problems, one aspect of an information processing device according to the present invention has a first acquisition means for acquiring product information of a target product, including features related to the target product, from a product page of the target product sold on an electronic commerce (EC) site; an identification means for identifying one or more similar products that are similar to the target product from among one or more other products sold on the EC site that satisfy a predetermined condition; a second acquisition means for acquiring product information of one or more similar products, including features related to the one or more similar products, from the product pages of the one or more similar products; a determination means for comparing the product information of the target product with the product information of the one or more similar products and determining improvement points in the product page of the target product based on a result of the comparison; and a presentation means for presenting the improvement points to a user.

[0008] In order to solve the above problems, one aspect of an information processing method according to the present invention includes a first acquisition step of acquiring product information for a target product, including features related to the target product, from a product page of the target product sold on an electronic commerce (EC) site; an identification step of identifying one or more similar products that are similar to the target product from among one or more other products sold on the EC site that satisfy predetermined conditions; a second acquisition step of acquiring product information for one or more similar products, including features related to the one or more similar products, from the product pages of the one or more similar products; a determination step of comparing the product information of the target product with the product information of the one or more similar products and determining improvement points in the product page of the target product based on a result of the comparison; and a presentation step of presenting the improvement points to a user.

[0009] In order to solve the above problem, one aspect of the program according to the present invention is an information processing program for causing a computer to execute information processing, the program including a first acquisition process for acquiring product information of a target product, including features related to the target product, from a product page of the target product sold on an electronic commerce (EC) site; an identification process for identifying one or more similar products that are similar to the target product from among one or more other products sold on the EC site that satisfy certain conditions; a second acquisition process for acquiring product information of one or more similar products, including features related to the one or more similar products, from the product pages of the one or more similar products; a determination process for comparing the product information of the target product with the product information of the one or more similar products and determining improvement points in the product page of the target product based on the result of the comparison; and a presentation process for presenting the improvement points to a user. Effect of the Invention

[0010] According to the present invention, it is possible to provide information for generating product information that stimulates a user's desire to purchase a product. The above-mentioned objects, aspects, and advantages of the present invention, as well as objects, aspects, and advantages of the present invention not described above, will be understood by those skilled in the art from the following detailed description of the invention by referring to the accompanying drawings and the claims. [Brief description of the drawings]

[0011] [Figure 1] FIG. 1 shows an example of the configuration of an information processing system. [Diagram 2] FIG. 2 shows an example of a product page. [Diagram 3] FIG. 3 shows an example of the functional configuration of the information processing device 10 according to the first embodiment. [Figure 4] FIG. 4 shows an example of the hardware configuration of the information processing device 10 and the user device 11. As shown in FIG. [Diagram 5] FIG. 5 shows a conceptual diagram of the similarity prediction procedure. [Figure 6A] FIG. 6A is a diagram for explaining an example of presentation of improvement point information. [Figure 6B] FIG. 6B is a diagram for explaining an example of presentation of improvement point information. [Figure 7] Figure 7 shows an example of a product page that has been improved based on the recommended improvements. [Figure 8] FIG. 8 shows a flowchart of the process executed by the information processing device according to the first embodiment. [Figure 9] FIG. 9 shows an example of the functional configuration of an information processing device 10 according to the second embodiment. [Figure 10] FIG. 10 shows an example image including weight information including a reference weight and an achievement rate for a target product. [Figure 11] FIG. 11 shows a conceptual diagram of the quality evaluation model. [Figure 12] FIG. 12 shows an example of improvement point information and weight information for a product page. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, with reference to the attached drawings, an embodiment for carrying out the present invention will be described in detail. Among the components disclosed below, those having the same functions are given the same reference numerals, and their description will be omitted. Note that the embodiment disclosed below is an example of a means for realizing the present invention, and should be appropriately modified or changed depending on the configuration of the device to which the present invention is applied and various conditions, and the present invention is not limited to the following embodiment. Furthermore, not all of the combinations of features described in the present embodiment are necessarily essential to the solution of the present invention.

[0013] First Embodiment [Example of information processing system configuration] Fig. 1 shows an example of the configuration of an information processing system according to this embodiment. As an example, as shown in Fig. 1, this information processing system includes an information processing device 10 and multiple user devices 11-1 to 11-N (N>1) used by multiple arbitrary users 1 to N. In the following description, unless otherwise specified, user devices 11-1 to 11-N may be collectively referred to as user device 11. In the following description, the terms user device and user may be used synonymously.

[0014] The user device 11 is, for example, a device such as a smartphone or a tablet, and is configured to be able to communicate with the information processing device 10 via a public network such as LTE (Long Term Evolution) or a wireless communication network such as a wireless LAN (Local Area Network). The user device 11 has a display unit (display surface) such as a liquid crystal display, and each user can perform various operations using a GUI (Graphic User Interface) equipped on the liquid crystal display. The operations include various operations on content such as images displayed on the screen, such as tapping, sliding, and scrolling using a finger or a stylus.

[0015] Note that user device 11 is not limited to the device shown in Fig. 1, and may be a device such as a desktop PC (Personal Computer) or a notebook PC. In this case, each user may perform operations using input devices such as a mouse and a keyboard. User device 11 may also be provided with a separate display unit.

[0016] The information processing device 10 may be a server device that provides an electronic commerce platform such as a marketplace, and the user device 11 can use web services (internet-related services) provided by the information processing device 10. Note that the information processing device 10 is not limited to being the above-mentioned server device, and the user device 11 may be configured to use web services provided by a server device (not shown) separate from the information processing device 10 via the information processing device 10. The web services may include, for example, electronic commerce (EC) services such as online malls, online supermarkets, flea markets, etc. The web services may provide a wide variety of products, such as fashion, books, food, daily necessities, furniture, and home appliances.

[0017] In order to use the web service, the user device 11 may register information such as the user's name, the user's address, product delivery address information, credit card information held by the user, and user demographic information (demographic user attributes such as gender, age, residential area, occupation, and family structure). When registering to use the web service, each user is associated with a user ID that identifies the user, and the above registration information is linked to the user ID.

[0018] The user device 11 can access any electronic commerce site (EC site) according to user operations, and transmit any search keyword or any product category information designated by the user, or information for specifying a product, to the information processing device 10. The information processing device 10 provides the user device 11 with product page information corresponding to the information transmitted from the user device 11, and the user device 11 can receive the product page information provided by the information processing device 10 and display the product page on the display unit. Furthermore, the user device 11 can carry out a product purchase procedure on the product page, i.e., transmit a purchase instruction including information on the payment method and delivery destination of the specified product to the information processing device 10, so that the user can have the specified product delivered to the delivery destination.

[0019] FIG. 2 is an example of a product page 20 displayed on the display unit of the user device 11. The product page 20 may display product-related features such as an image 202 representing a product 201, a title (e.g., product name) 203, a price 204, and delivery information (delivery schedule) 205. In this embodiment, the overall product-related features on the product page are referred to as "product information", and each feature is referred to as an "entity". In the example of FIG. 2, the image 202, the title 203, the price 204, and the delivery information 205 correspond to entities, and the entire entity corresponds to product information. Note that in the product page 20, the product information includes four entities (the image 202, the title 203, the price 204, and the delivery information 205), but may include other entities. The other entities are, for example, a product description, a classification (a category or a subcategory), a review, and a rating (e.g., a score out of a maximum of 5). When a user wishes to purchase the product 201, the user can step by step proceed to the purchase procedure for the product 201 by selecting the "Add to Cart" button 211. Alternatively, the user may directly proceed to the purchase procedure for the product 201 by selecting the "Purchase" button 212. In the following description, an action for adding a product to the cart on a product page, such as selecting the "Add to Cart" button 211 in the example of the product page 20, is referred to as a "cart action."

[0020] The product page may be created by a user such as a merchant (business operator) who sells the product for sale included in the product page (hereinafter referred to as the "target product") or a website operator. In this embodiment, the user who creates the product page is also referred to as the "creator." Note that the creator does not necessarily have to be a user (human being) and may be any information processing device. In the example of Fig. 2, a product 201 corresponds to the target product. In general, a creator may determine each entity of product information to be displayed (listed) on a product page based on guesswork or experience. However, the product information determined based on the guesswork may not necessarily be effective information that stimulates a user's desire to purchase the product. The information processing device 10 according to the present embodiment is configured to identify similar products similar to the target product from past popular products (hit products) and compare the product information of the target product with the product information of the similar products. Furthermore, the information processing device 10 is configured to determine improvement points for the product information of the target product based on the comparison and recommend (present) the improvement points to the creator of the product information of the target product. In the present embodiment, a popular product refers to a product whose sales performance indicators such as sales amount, sales volume, and CVR (conversion rate) are at or above a predetermined level (for example, at or above a predetermined value). Note that the improvement points correspond to parts (correction parts, change parts, improvement parts, etc.) that are determined to require correction, change, etc., in order to stimulate the user's desire to purchase the product, for at least a part of the product information, target (correction target, change target, improvement target, etc.), and entity (correction entity, change entity, improvement entity, etc.).

[0021] [Functional configuration of the information processing device 10 according to the first embodiment] FIG. 3 shows an example of a functional configuration of the information processing device 10 according to the present embodiment. The information processing device 10 shown in Fig. 3 includes a user feature acquisition unit 101, a popular product information acquisition unit 102, a target product information acquisition unit 103, a similar product identification unit 104, an improvement point determination unit 105, a recommendation unit 106, a learning model storage unit 110, and a data storage unit 120. The learning model storage unit 110 is configured to be able to store a user feature prediction model 111 and a similarity prediction algorithm 112. As described later, the similarity prediction algorithm 112 is configured to include a plurality of learning models. In addition, the data storage unit 120 is configured to be able to store product information 121 and user features 122.

[0022] The product information 121 is product information included in a product page of a product that the information processing device 10 can provide to the user device 11 through a web service, and is registered (stored) in advance in the data storage unit 120. A product ID for identifying each product is assigned to the product that can be provided through the web service, and the product information of each product page is registered in the product information 121 in association with the product ID. As described above with reference to FIG. 2, the product information includes one or more entities. Note that, although the product information 121 is stored in the information processing device 10 in FIG. 3, the product information 121 may be stored in an external device, and the information processing device 10 may be configured to be able to acquire the product information 121.

[0023] The user feature acquisition unit 101 acquires fact features (fact information, fact attributes) (hereinafter, user features) about each of the user devices 11-1 to 11-N. The user features are fact-based features (information) that are actually or objectively obtained from the user device or user. The user feature acquisition unit 101 can acquire the user features directly from the user device 11, for example. The user feature acquisition unit 101 can also acquire the user features as information registered in a web service by the user of the user device 11.

[0024] The user characteristics include the IP address of the user device, the user's address, the user's name, the number of the credit card held by the user, the user's demographic information (demographic user attributes such as gender, age, residential area, occupation, family structure, etc.), etc. The user characteristics may also include a registration number and a registered name when using a web service. The user characteristics may also include information related to a call history, a usage status of the web service, and points that can be accumulated by using the service. In this embodiment, the usage status of the web service includes the usage status of the product on the EC site. Specifically, the usage status of the product on the EC site includes a click history for viewing the product, a cart action history for the product, a search history for the product, and a purchase history for the product (including the purchase results). These usage status data are linked to the product ID of the product on the EC site. Thus, user characteristics may include any information, including information related to the user device or the user himself / herself, and information related to the use of a given service via communication.

[0025] The user feature acquisition unit 101 is also configured to apply the acquired user features to a trained user feature prediction model 111 to acquire user features (estimated user features (attributes)) estimated for the user features. The user feature prediction model 111 is configured, for example, to a Variational Autoencoder (VAE). The user feature prediction model 111 is configured to receive user features (i.e., fact features) of an arbitrary target user as input, and to output the probability (correspondence probability) that each of a plurality of user features corresponds (matches) the target user. The user feature acquisition unit 101 ultimately determines the estimated user features of the target user from the correspondence probability.

[0026] For example, the user feature acquisition unit 101 inputs demographic information of an arbitrary target user and data indicating a purchase history and preferences in past web services of the target user as the user features of the target user to the user feature prediction model 111. The user feature prediction model 111 outputs probabilities for each of a plurality of products that the target user is estimated to purchase and a plurality of preferences that the target user may have as corresponding probabilities. Then, the user feature acquisition unit 101 can acquire products and preferences that have probabilities equal to or greater than a predetermined value as estimated user features of the target user.

[0027] The user feature acquiring unit 101 associates the acquired user features (actual features and estimated features) with a user ID for identifying the user, and stores them in the data storage unit 120 as user features 122. As described above, data on the usage status of a product on the EC site, which is included in the user features, is associated with the product ID of the product.

[0028] The popular product information acquisition unit 102 identifies one or more popular products (which may also be called hit products or successful products) based on the purchase histories of users 1 to N included in the user characteristics 122, and acquires product information on the identified popular products. For example, first, the popular product information acquisition unit 102 identifies one or more popular products based on purchase history over a certain period of time, and acquires product IDs of the identified popular products (hereinafter referred to as "popular product IDs"). In this embodiment, a popular product refers to a product whose indicators, such as sales amount, sales volume, or CVR, are at or above a predetermined level (e.g., at or above a predetermined value). The predetermined level for determining a popular product may be set by an operator through an input operation using the input unit (input unit 45 in FIG. 4), may be set in advance in the system, or may be set by any program stored in the storage unit (ROM 42 or RAM 43 in FIG. 4). The popular product information acquiring unit 102 may identify popular products based on information or data other than the purchase history of the users 1 to N, and acquire the product IDs of the identified popular products. The popular product information acquiring unit 102 acquires product information of the popular product from the product information 121 stored in the data storage unit 120 using the acquired popular product ID.

[0029] The target product information acquisition unit 103 acquires the target product page from the creator of the product page of the target product. The creator may be any one of users 1 to N shown in Fig. 1. The product page of the target product may be acquired by a communication unit (communication I / F 47 in Fig. 4) or an input unit, may be set in advance in the system, or may be stored in a storage unit (ROM 42 or RAM 43 in Fig. 4). Furthermore, the target product information acquisition unit 103 acquires (extracts) product information in the product page of the target product. The product information includes one or more entities. The product page of the target product may be a released product page (i.e., a product page in an uploaded state that can be viewed by users 1 to N) or a product page before release. Referring to FIG. 2, the target product information acquisition unit 103 acquires product information including four entities (image 202, title 203, price 204, and delivery information 205) in the product page 20.

[0030] The similar product identification unit 104 uses the product information of the popular product acquired by the popular product information acquisition unit 102 and the product information of the target product acquired by the target product information acquisition unit 103 to identify one or more similar products that are similar to the target product among the popular products. In this embodiment, the similar product identification unit 104 is configured to identify the similar products using a similarity prediction algorithm 112 stored in the learning model storage unit 110. The procedure for identifying similar products will be described later.

[0031] The improvement point determination unit 105 compares the product information of one or more similar products identified by the similar product identification unit 104 with the product information of the target product acquired by the target product information acquisition unit 103. Then, the improvement point determination unit determines improvement points for the product page of the target product for the creator of the product page based on the comparison. The determined improvement points are transmitted to the recommendation unit 106. The procedure for determining improvement points will be described later. Furthermore, the improvement point determination unit 105 may generate information indicating the improvement points (hereinafter, referred to as "improvement point information"). The generated improvement point information is transmitted to the recommendation unit 106. Note that the improvement point information may be generated by the recommendation unit 106 that receives the improvement points.

[0032] The recommendation unit 106 recommends (presents) the improvement points determined by the improvement point determination unit 105 to the creator of the product page of the target product. For example, the recommendation unit 106 transmits the improvement point information via the communication unit so that it is displayed on the display unit of the user device 11 of the creator (or the user involved in the creation of the product page). A display example of the improvement point information will be described later. Note that, in this embodiment, an example of recommending improvement points by displaying improvement point information will be described, but the improvement points may be recommended by other means such as voice.

[0033] [Hardware configuration of information processing device 10] FIG. 4 is a block diagram showing an example of a hardware configuration of the information processing device 10 according to the present embodiment. The information processing apparatus 10 according to the present embodiment can be implemented on any single or multiple computers, mobile devices, or any other processing platform. 4, the information processing device 10 is illustrated as being implemented in a single computer, but the information processing device 10 according to the present embodiment may be implemented in a computer system including multiple computers. The multiple computers may be connected to each other via a wired or wireless network so as to be able to communicate with each other.

[0034] 4, the information processing device 10 may include a CPU 41, a ROM 42, a RAM 43, a HDD 44, an input unit 45, a display unit 46, a communication I / F 47, and a system bus 48. The information processing device 10 may also include an external memory. A CPU (Central Processing Unit) 41 generally controls the operation of the information processing device 10, and controls each of the components (42 to 47) via a system bus 48, which is a data transmission path.

[0035] The ROM (Read Only Memory) 42 is a non-volatile memory that stores control programs and the like necessary for the CPU 41 to execute processing. Note that the programs may be stored in a non-volatile memory such as an HDD (Hard Disk Drive) 44 or an SSD (Solid State Drive) or an external memory such as a removable storage medium (not shown). The RAM (Random Access Memory) 43 is a volatile memory and functions as a main memory, a work area, etc. of the CPU 41. That is, the CPU 41 loads necessary programs, etc. from the ROM 42 to the RAM 43 when executing processing, and realizes various functional operations by executing the programs, etc.

[0036] The HDD 44 stores, for example, various data and various information required when the CPU 41 performs processing using a program. The HDD 44 also stores, for example, various data and various information obtained when the CPU 41 performs processing using a program. The input unit 45 is composed of a keyboard and a pointing device such as a mouse. The display unit 46 is configured with a monitor such as a liquid crystal display (LCD), etc. The display unit 46 may be configured in combination with the input unit 45 to function as a GUI (Graphical User Interface).

[0037] The communication I / F 47 is an interface that controls communication between the information processing device 10 and an external device. The communication I / F 47 provides an interface with a network and executes communication with an external device via the network. Various data, various parameters, etc. are transmitted and received between the external device and the communication I / F 47. In this embodiment, the communication I / F 47 may execute communication via a wired LAN (Local Area Network) or a dedicated line conforming to a communication standard such as Ethernet (registered trademark). However, the network available in this embodiment is not limited to this, and may be configured as a wireless network. This wireless network includes wireless PANs (Personal Area Networks) such as Bluetooth (registered trademark), ZigBee (registered trademark), and UWB (Ultra Wide Band). It also includes wireless LANs (Local Area Networks) such as Wi-Fi (Wireless Fidelity) (registered trademark), and wireless MANs (Metropolitan Area Networks) such as WiMAX (registered trademark). It also includes wireless WANs (Wide Area Networks) such as LTE / 3G, 4G, and 5G. Note that the network only needs to connect each device so that they can communicate with each other, and the communication standard, scale, and configuration are not limited to those described above.

[0038] At least some of the functions of each element of the information processing device 10 shown in Fig. 3 can be realized by the CPU 41 executing a program. However, at least some of the functions of each element of the information processing device 10 shown in Fig. 3 may be operated as dedicated hardware. In this case, the dedicated hardware operates under the control of the CPU 41.

[0039] [Hardware configuration of user device 11] The hardware configuration of the user device 11 shown in Fig. 1 may be the same as that of Fig. 4. That is, the user device 11 may include a CPU 41, a ROM 42, a RAM 43, a HDD 44, an input unit 45, a display unit 46, a communication I / F 47, and a system bus 48. The user device 11 can display various information provided by the information processing device 10 on the display unit 46, and perform processing corresponding to an input operation received from a user via a GUI (configured by the input unit 45 and the display unit 46).

[0040] [Procedure for identifying similar products] Next, a procedure for identifying similar products by the similar product identifying unit 104 will be described. In this embodiment, the similar product identifying unit 104 predicts (determines) the similarity between a popular product and a target product from multiple perspectives using the similarity prediction algorithm 112, and ultimately identifies one or more similar products that are similar to the target product based on the result of the prediction. The prediction of the similarity corresponds to a product matching process.

[0041] In this embodiment, the similar product identification unit 104 predicts the similarity between the popular product and the target product from four perspectives: (1) title similarity, (2) image similarity, (3) price comparison, and (4) attribute similarity. Note that, in this embodiment, the similarity between the popular product and the target product is predicted from these four perspectives, but the perspectives for predicting the similarity are not limited to four. Fig. 5 shows a conceptual diagram of a similarity prediction procedure according to this embodiment. In Fig. 5, the similarity prediction algorithm 112 includes a first learning model 501, a second learning model 502, a third learning model 503, a fourth learning model 504, and a fifth learning model 505 of machine learning, as will be referred to below. Note that, in this embodiment, a procedure for predicting similarity using the learning models will be described, but similarity may be predicted by a method other than the learning models with respect to the following viewpoints (1) to (4).

[0042] (1) Similarity of titles The similar product identification unit 104 acquires the title (hereinafter referred to as the "popular product title") from the product information of the popular product acquired by the popular product information acquisition unit 102. The similar product identification unit 104 also acquires the title (hereinafter referred to as the "target product title") from the product information of the target product acquired by the target product information acquisition unit 103. The similar product identification unit 104 then derives the similarity between the popular product title and the target product title using a first learning model 501. The first learning model 501 is a learning model for natural language processing (NLP), such as spaCy or BERT. The popular product title and the target product title are applied to the first learning model 501 to output the similarity (hereinafter referred to as the "title similarity"). The title similarity is, for example, cosine similarity. The cosine similarity takes a value (scale) ranging from 0 to 1, and the closer it is to 1, the more similar the popular product title and the target product title are.

[0043] (2) Image similarity The similar product identification unit 104 acquires an image (hereinafter referred to as a "popular product image") from the product information of the popular product acquired by the popular product information acquisition unit 102. The similar product identification unit 104 also acquires an image (hereinafter referred to as a "target product image") from the product information of the target product acquired by the target product information acquisition unit 103. The similar product identification unit 104 then derives the similarity between the popular product image and the target product image using the second learning model 502. The second learning model 502 is a learning model for image recognition, for example, a CNN (Convolutional Neural Network). The similar product identification unit 104 applies the popular product image to the second learning model 502 to extract the feature amount of the popular product image. Similarly, the similar product identification unit 104 applies the target product image to the second learning model 502 to extract the feature amount of the target product image. The similar product identification unit 104 then derives the difference (e.g., Euclidean distance) between the two feature amounts in a common feature vector space. The smaller the difference, the more similar the popular product image and the target product image are. For this reason, the inverse of the difference is hereinafter referred to as "image similarity." The closer the image similarity is to 1, the more similar the popular product image and the target product image are.

[0044] (3) Price comparison The similar product identification unit 104 acquires the price (hereinafter referred to as the "popular product price") from the product information of the popular product acquired by the popular product information acquisition unit 102. The similar product identification unit 104 also acquires the price (hereinafter referred to as the "target product price") from the product information of the target product acquired by the target product information acquisition unit 103. The similar product identification unit 104 then compares the popular product price with the target product price using the third learning model 503. The third learning model 503 is a learning model for outlier detection (abnormal value detection), such as a chi-square test, a Smirnoff-Grubbs test, or a Dixon's Q test. By applying the popular product price and the target product price to the third learning model 503, it is detected whether the target product price is extremely out of line compared to the popular product price (i.e., the presence or absence of an outlier is detected).

[0045] (4) Similarity of attributes The similar product identification unit 104 extracts the attributes of a popular product (hereinafter referred to as "popular product attributes") and the attributes of a target product (hereinafter referred to as "target product attributes"), and derives the similarity between the two attributes. The attributes are, for example, the brand, size, model number, and color of the product. The similar product identification unit 104 extracts (predicts) the popular product attributes by applying character information such as the title and product description included in the product information of the popular product to the fourth learning model 504. The fourth learning model 504 is a learning model for natural language processing (NLP), and is, for example, a long short-term memory (LSTM). Similarly, for the target product attributes, character information such as the title and product description included in the product information of the target product is applied to the fourth learning model 504 to extract the target product attributes. Alternatively or in addition to this, the popular product attributes and the target product attributes may be extracted from images included in the product information of the popular product and the target product, respectively.

[0046] The similar product identification unit 104 then derives the similarity between the extracted popular product attribute and the target product attribute using the fifth learning model 505. The fifth learning model is a learning model for natural language processing (NLP) like the first learning model, such as spaCy or BERT. The popular product attribute and the target product attribute are applied to the fifth learning model 505 to output a similarity (hereinafter, "attribute similarity"). The attribute similarity is, for example, cosine similarity. The cosine similarity takes a value from 0 to 1, and the closer to 1 the value is, the more similar the popular product attribute and the target product attribute are.

[0047] The similar product identification unit 104 predicts the similarity from the four viewpoints as described above, and then finally identifies one or more similar products that are similar to the target product based on the result of the prediction. For example, the similar product identification unit 104 predicts the similarity with the target product for all popular products with popular product IDs acquired by the popular product information acquisition unit 102. Then, the similar product identification unit 104 identifies, as a similar product, a popular product that satisfies the following conditions: the title similarity is equal to or greater than a first threshold, the image similarity is equal to or greater than a second threshold, the attribute similarity is equal to or greater than a third threshold, and the target product price is not extremely different from the popular product price. The conditions for determining a similar product are not limited to this, and for example, a popular product that satisfies the condition that the image similarity is equal to or greater than a second threshold value may be specified as a similar product.

[0048] [Procedure for determining improvement points] Next, a procedure for determining improvement points by the improvement point determining unit 105 will be described. In this embodiment, the improvement point determination unit 105 compares the product information of one or more similar products identified by the similar product identification unit 104 with the product information of the target product, recognizes different entities, and determines improvement points based on the different entities. In this way, in this embodiment, the improvement point determination unit 105 determines improvement points from entities where the product information of the target product is insufficient (different) from the product information of the similar products.

[0049] For example, if the product information of the similar product includes multiple images, but the product information of the target product includes only one image, the improvement point determination unit 105 recognizes that there is a difference in the number of images among the entities of the product information.Then, the improvement point determination unit 105 determines "image" as the improvement point.Also, for example, if the product information of the similar product includes a product description, but the product information of the target product does not include a product description, the improvement point determination unit 105 determines "product description" as the improvement point. When there are a plurality of similar products, the improvement point determination unit 105 may determine, as an improvement point, an entity that is more different between the product information of the plurality of similar products and the product information of the target product.

[0050] Furthermore, the improvement point determination unit 105 may determine improvement points using the results obtained by the similarity prediction procedure by the similar product identification unit 104 described above. For example, the similarity between the titles of the similar product and the target product may be derived using the first learning model 501, and if the similarity is equal to or less than a predetermined threshold, the "title" may be determined as an improvement point. Similarly, the similarity between the attributes of the similar product and the target product may be derived using the fifth learning model 505, and if the similarity is equal to or less than a predetermined threshold, the "product description" may be determined as an improvement point.

[0051] In addition to this, or instead of this, the improvement point determination unit 105 may determine more specific improvement points for the entity of the product information. For example, when the product information of the target product includes only one image while the product information of the similar product includes multiple images, the improvement point determination unit 105 may determine "the number of images" or "increase the number of images" as the improvement point. For example, the improvement point determination unit 105 may determine, as the information related to the improvement point, correction of parameters of the image (one or more parameters selected from a group of parameters that affect the appearance such as hue, saturation, brightness, dimensions, aspect ratio, etc.). For example, the improvement point determination unit 105 may determine, as the improvement point, correction of some character string equivalent to the product description in the image. For example, the improvement point determination unit 105 may determine, as the information related to the improvement point, the upper limit, lower limit, range, etc. of the number of characters related to the product description. For example, the improvement point determination unit 105 may determine, as the information related to the improvement point, whether the price of the product is close to a predetermined price or price range. The improvement point determination unit 105 may determine, for example, whether or not the delivery information indicates a predetermined delivery condition as information related to an improvement point.

[0052] Furthermore, the improvement point determination unit 105 may generate improvement point information. For example, the improvement point determination unit 105 generates text information of the determined improvement points as the improvement point information. In addition to this, or instead of this, the improvement point determination unit 105 may generate other types of improvement point information such as image information representing the improvement points. Note that the improvement point information may be new product information (corresponding to a correction proposal, a change proposal, an improvement proposal, etc.) generated according to the determined improvement points.

[0053] [Example of improvement point information display] An example of presentation of improvement point information by the recommendation unit 106 will be described with reference to Figures 6A and 6B. Figures 6A and 6B are diagrams for explaining an example of presentation of improvement point information. FIG. 6A shows an example of an image 60 including improvement point information expressed by text information such as "number of images." The image 60 includes "number of images," "title," and "product description" as improvement point information. The recommendation unit 106 presents the image 60 to the creator of the product page of the target product. For example, the recommendation unit 106 transmits data of the image 60 to the user device 11 so that the image 60 is displayed on the display unit of the user device 11 of the creator (or a user involved in the creation of the product page). The improvement point information may be presented by being superimposed on the created product page (for example, by a pop-up). Fig. 6B shows an example in which improvement point information "increase image information" is displayed as a pop-up when the creator of the product page 20 shown in Fig. 2 moves the cursor 61 close to the image 202. The recommendation unit 106 may transmit data of the improvement point information to the user device 11 so that the improvement point information is superimposed on the display unit of the user device 11 of the creator (or the user involved in the creation of the product page).

[0054] FIG. 7 shows an example of a product page improved by the recommended improvement points. Specifically, FIG. 7 shows an example of a product page 70 improved on the product page 20 shown in FIG. 2 in accordance with the improvement point information shown in FIG. 6A. Images 701 corresponding to image 202 on product page 20 have an increased number of images. Furthermore, title 702 corresponding to title 203 on product page 20 has been changed to a more detailed description. Furthermore, product information 703 that did not exist on product page 20 has been added. In this way, by improving (correcting) the product information of the target product based on the product information of popular products similar to the target product, it is possible to generate product information that is expected to further stimulate users' desire to purchase the product.

[0055] [Processing flow] Fig. 8 shows a flowchart of processing executed by the information processing device 10 according to this embodiment. The processing shown in Fig. 8 can be realized by the CPU 41 of the information processing device 10 loading a program stored in the ROM 42 or the like into the RAM 43 and executing the program. The user feature prediction model 111 and the similarity prediction algorithm 112 are stored in the learning model storage unit 110. Also, the product information 121 is stored in the data storage unit 120.

[0056] In S11, user feature acquisition unit 101 acquires user features of users 1 to N from user devices 11-1 to 11-N. User feature acquisition unit 101 may also acquire user features (estimated user features (attributes)) estimated for users 1 to N based on the user features using user feature prediction model 111. User feature acquisition unit 101 stores the acquired user features in data storage unit 120 as user features 122.

[0057] In S12, the popular product information acquiring unit 102 identifies one or more popular products based on the purchase histories of users 1 to N included in the user characteristics acquired in S11. The popular product information acquiring unit 102 may identify popular products based on information or data other than the purchase histories of users 1 to N. Furthermore, the popular product information acquiring unit 102 acquires product information of the identified one or more popular products from product information 121 stored in the data storage unit 120.

[0058] In S13, the target product information acquisition unit 103 acquires a product page of the target product created by the creator, and acquires (extracts) product information from the product page. The product page includes one or more entities (one or more features related to the product). An example of a product page is the product page 20 shown in FIG. 2.

[0059] In S14, the similar product identification unit 104 predicts the similarity between the popular product and the target product using the product information of the popular product acquired in S12 and the product information of the target product acquired in S23. Next, in S15, the similar product identification unit 104 uses the result of the prediction in S14 to identify one or more similar products that are similar to the target product among the popular products. The procedures for predicting the similarity (S14) and identifying similar products (S15) are as described above.

[0060] In S16, the improvement point determination unit 105 compares the product information of the target product with the product information of similar products to determine improvement points for the product page of the target product. The improvement point determination unit 105 (or the recommendation unit 106) also generates improvement point information indicating the improvement points. Then, in S17, the recommendation unit 106 recommends (suggests) the improvement points determined in S16 to the creator of the product page of the target product. The recommendation can be made by displaying the improvement point information on the display unit of the user device 11 of the creator.

[0061] In this way, the information processing device 10 according to the present embodiment identifies similar products similar to the target product from one or more popular products identified from past data, and determines improvement points for the product page of the target product based on the product information of the similar products. The information processing device 10 then recommends (suggests) the improvement points to the creator of the product page of the target product. The creator can improve the product information of the product page of the target product in accordance with the recommendation, and users who view the improved product page may be motivated to purchase the product.

[0062] <Second embodiment> In the first embodiment, similar products similar to the target product are identified from popular products, and improvement points for the product page of the target product are determined by comparing the product information of the similar products and the target product, and are recommended to the creator. In this embodiment, a method of evaluating (assessing) the product management of the target product on an EC site will be described. Below, a description of the same configuration and features as in the first embodiment will be omitted.

[0063] [Functional configuration of the information processing device 10 according to the second embodiment] Fig. 9 shows an example of the functional configuration of the information processing device 10 according to this embodiment. The information processing device 10 shown in Fig. 9 is obtained by adding a feature element derivation unit 107 and a quality evaluation unit 108 to the information processing device 10 (Fig. 3) according to the first embodiment. Furthermore, a quality evaluation model 113 is stored in a learning model storage unit 110, and a reference weight 123 is stored in a data storage unit 120.

[0064] The feature element derivation unit 107 derives (acquires) four feature elements (Factors) for a target product whose product page has been released. In this embodiment, the following four feature elements are derived. (1) Content elements A content element is content included in the product information on a product page, such as a product description and a product image. (2)Offer element The provision elements are elements that can be incentives for maintaining the competitiveness of the product, such as the price and delivery information of the product included in the product information on the product page. (3) Positive evaluation (rating) elements The positive evaluation elements are positive evaluation elements among evaluations (reactions) of users (shoppers) such as evaluations, reviews, and comments on a product that can be obtained after the sale (shipment) of the product. For example, the feature element derivation unit 107 can obtain (extract) the positive evaluation elements by applying a learning model for natural language processing (NLP) to character information indicating the user's evaluation. (4) Post-Purchase Quality The post-purchase quality element is quality information that can be obtained after the sale of a product, such as a return rate, a probability of a merchant shipping the product within a predetermined time period, etc. For example, the feature element derivation unit 107 can obtain the post-purchase quality element by tallying up the return rate after the sale of the product or feedback from the merchant.

[0065] The importance (hereinafter also referred to as "weight") of the four feature elements for a product in stimulating the user's desire to purchase may differ depending on the category and subcategory. For example, if the product is toothpaste, the price of the product included in the offer element may be an important element in stimulating the user's desire to purchase. On the other hand, if the product is furniture, the positive evaluation element and the product image included in the content element may be important elements in stimulating the user's desire to purchase.

[0066] In this embodiment, taking such circumstances into consideration, the quality of product management for the target product is evaluated by referring to the ideal weights (importance) of each of the four feature elements determined in advance for products in the same category or subcategory as the target product. The reference weights 123 stored in the data storage unit 120 are ideal weights of each of the four feature elements determined in advance for each of a plurality of categories and subcategories. The reference weights 123 can be obtained by machine learning, for example, for each category or subcategory of popular products identified by the popular product information acquisition unit 102. The weights included in the reference weights 123 are linked to category IDs that identify the categories or subcategories.

[0067] The quality evaluation unit 108 derives an evaluation weight (weight for evaluation) for each of the four feature elements derived for the target product. The evaluation weight is a weight in which a maximum value (e.g., maximum amount of information, maximum probability) is set for each feature element, the probability of each feature element relative to the maximum value is calculated, and the weight is normalized so that the sum of the four evaluation indexes becomes 1. The quality evaluation unit 108 then compares the evaluation weight with a reference weight of a product in the same category as the target product among the reference weights 123, and calculates the achievement rate of the evaluation weight, which is the degree to which the evaluation weight achieves the reference weight. The category of the target product can be identified using a category included in the product information. Alternatively, the category of the target product may be identified using a learning model from information included in the product information.

[0068] The quality evaluation unit 108 may provide weight information including the reference weight and the achievement rate for the target product to the creator of the product page of the target product. For example, the quality evaluation unit 108 may cause the weight information to be displayed on the display unit of the user device 11 of the creator. 10 shows an example image 1000 including weight information. In the example image 1000, the reference weights of the content elements, the offering elements, the positive evaluation elements, and the post-purchase quality elements are 40% (0.4), 10% (0.1), 20% (0.2), and 30% (0.3), respectively. The achievement rate is 90%. By displaying such weight information, the creator can understand which characteristic elements are more effective for the product and how much the evaluation weight has been achieved against the reference weight, and can make improvements in the creation of the product page and the management after the product is sold.

[0069] The quality evaluation unit 108 may also evaluate the product management of the target product using a learning model, taking into account other features of the target product. In this embodiment, the quality assessment model 113 stored in the learning model storage unit 110 is used. The quality assessment model 113 is a learning model based on CatBoost. Alternatively, the quality assessment model 113 may be another boosting-based learning model such as XGBoost or LightGBM. Therefore, it is possible to clearly analyze which of the input data is responsible for the result.

[0070] A conceptual diagram of the quality evaluation model 113 is shown in FIG. 11. The quality evaluation model 113 receives input data 1100 including evaluation weights of the above four feature elements (content element, delivery element, positive evaluation element, and post-purchase quality element) derived for the target product and other features. In this embodiment, the other features may include time features and / or action features. The time features may include features indicating the seasonality of the target product and features of the target product's ranking in a ranking that is updated at regular intervals. In addition, the action features may include features of the number of clicks, number of views, number of cart actions, and number of impressions (the number of times an advertisement for the target product is displayed on any page of a website) for the target product.

[0071] The quality evaluation model 113 is configured to output a probability 1101 that the sales number of the target product will reach a predetermined sales number for the input data 1100. The quality evaluation model 113 also outputs the degree to which the input data will bring about an output result. Thus, when a result is obtained that a predetermined sales number will be reached, the input data resulting from the result is also derived. This makes it possible to confirm which input data was effective in achieving the predetermined sales number. The quality evaluation unit 108 may provide information on the input data resulting from the output result from the quality evaluation model 113 to a user. The user may be any user including the creator of the product page. This allows the user to understand which feature element of the product affects the sales number, and to make improvements in the creation of the product page and the management after the product is sold. The quality evaluation model 113 may be configured to output the probability that an index related to sales performance, such as sales amount or CVR, will reach a predetermined level, without being limited to the number of sales.

[0072] In this manner, in this embodiment, information evaluating product management after the release of a product page can be provided to the creator of the product page. In addition to this, the improvement point information described in the first embodiment may also be provided to the creator at the same time. FIG. 12 shows an example of improvement point information and weight information for a product page. FIG. 12 shows improvement point information 1201 determined for a product page 1200 according to the first embodiment and weight information 1202 acquired according to this embodiment. By displaying such improvement point information 1201 and weight information 1202, the creator can understand which point (entity) among the multiple improvement points should be improved with priority from not only the multiple improvement point information 1201 for the product page but also the weight information 1202, and can attempt to create a product page that will further stimulate the user's desire to purchase.

[0073] Although specific embodiments have been described above, the embodiments are merely illustrative and are not intended to limit the scope of the present invention. The apparatus and method described in this specification can be embodied in forms other than those described above. Furthermore, the above-described embodiments can be appropriately omitted, substituted, and modified without departing from the scope of the present invention. Such omitted, substituted, and modified forms are included in the scope of the claims and their equivalents, and belong to the technical scope of the present invention.

[0074] The disclosure of this embodiment includes the following configuration. [1] An information processing device having: a first acquisition means for acquiring product information of a target product, including features related to the target product, from a product page of the target product sold on an electronic commerce (EC) site; an identification means for identifying one or more similar products that are similar to the target product from among one or more other products sold on the EC site that satisfy predetermined conditions; a second acquisition means for acquiring product information of one or more similar products, including features related to the one or more similar products, from the product pages of the one or more similar products; a determination means for comparing the product information of the target product with the product information of the one or more similar products and determining improvement points in the product page of the target product based on a result of the comparison; and a presentation means for presenting the improvement points to a user.

[0075] [2] The information processing device described in [1], wherein the identification means acquires product information of the other one or more products included in the product pages of the other one or more products, predicts a similarity between the product information of the other one or more products and the product information of the target product, and identifies the one or more similar products based on the result of the prediction.

[0076] [3] The information processing device described in [2], wherein the identification means predicts the similarity based on at least one of the features contained in the product information of the other one or more products and the product information of the target product, including the product title, the product image, the product price, and the product attributes.

[0077] [4] An information processing device described in any of [1] to [3], wherein the determination means extracts different characteristics by comparing product information of the target product with product information of the one or more similar products, and determines the improvement points based on the different characteristics.

[0078] [5] The information processing device according to any one of [1] to [4], wherein the specified condition is that an index related to product sales performance is at or above a specified level.

[0079] [6] The information processing device described in any one of [1] to [5], wherein the presentation means presents the improvement points to the user by displaying improvement point information indicating the improvement points on a user device owned by the user.

[0080] [7] An information processing device described in any of [1] to [6], further comprising a third acquisition means for acquiring a plurality of characteristic elements related to the sale of the target product after the target product is sold on the EC site, and an evaluation means for evaluating the quality of product management of the target product on the EC site based on the plurality of characteristic elements.

[0081] [8] An information processing method comprising: a first acquisition step of acquiring product information of a target product, the product information including features of the target product, from a product page of the target product sold on an electronic commerce (EC) site; an identification step of identifying one or more similar products that are similar to the target product from among one or more other products sold on the EC site that satisfy predetermined conditions; a second acquisition step of acquiring product information of one or more similar products, the product information including features of the one or more similar products, from the product pages of the one or more similar products; a determination step of comparing the product information of the target product with the product information of the one or more similar products and determining improvement points in the product page of the target product based on a result of the comparison; and a presentation step of presenting the improvement points to a user.

[0082] [9] An information processing program for causing a computer to execute information processing, the program causing the computer to execute processes including a first acquisition process for acquiring product information of a target product, including features related to the target product, from a product page of the target product sold on an electronic commerce (EC) site; an identification process for identifying one or more similar products that are similar to the target product from among one or more other products sold on the EC site that satisfy predetermined conditions; a second acquisition process for acquiring product information of one or more similar products, including features related to the one or more similar products, from the product pages of the one or more similar products; a determination process for comparing the product information of the target product with the product information of the one or more similar products and determining improvement points in the product page of the target product based on the result of the comparison; and a presentation process for presenting the improvement points to a user. [Explanation of symbols]

[0083] 1 to N: user, 10: information processing device, 11-1 to 11-N: user device, 101: user feature acquisition unit, 102: popular product information acquisition unit, 103: target product information acquisition unit, 104: similar product identification unit, 105: improvement point determination unit, 106: recommendation unit, 107: feature element derivation unit, 108: quality evaluation unit, 110: learning model storage unit, 111: user feature prediction model, 112: similarity prediction algorithm, 113: quality evaluation model, 120: data storage unit, 121: product information, 122: user feature, 123: reference weight

Claims

1. A first acquisition means for acquiring product information of a target product, the product information including one or more target product features that are one or more features related to the target product, on a product page of the target product sold on an electronic commerce (EC) site; A means for identifying a similar product to the target product among one or more other products that satisfy a predetermined condition and are sold on an EC site; A second acquisition means for acquiring product information of the similar product, the product information including one or more similar product features which are one or more features related to the similar product, on a product page of the similar product; A determination means for comparing the one or more target product features with the one or more similar product features and determining, among the one or more target product features, features that are not included in the one or more similar product features, as improvement points for the product page of the target product; A presentation means for presenting the improvement points to a user; having the one or more target product features include one or more images of the target product, and the one or more similar product features include one or more images of the similar products; When the number of images of the target product and the number of images of the similar products are different, the determining means determines the number of images as an improvement point for the product page of the target product.

23. An information processing apparatus comprising:

2. the identification means acquires product information of the one or more other products included in the product pages of the one or more other products, predicts a similarity between the product information of the one or more other products and the product information of the target product, and identifies the similar products based on a result of the prediction; 2. The information processing apparatus according to claim 1,

3. the identification means predicts the similarity based on at least one of the following features included in the product information of the one or more other products and the product information of the target product: a product title, a product image, a product price, and a product attribute; 3. The information processing apparatus according to claim 2.

4. The predetermined condition is that an index related to sales performance of the product is equal to or higher than a predetermined level.

2. The information processing apparatus according to claim 1,

5. the presenting means presents the improvement points to the user by displaying improvement point information indicating the improvement points on a user device owned by the user; 2. The information processing apparatus according to claim 1,

6. a third acquiring means for acquiring a plurality of characteristic elements related to the sale of the target product after the target product is sold on the EC site; an evaluation means for evaluating the quality of product management of the target product at the EC site based on the plurality of feature elements; 2. The information processing apparatus according to claim 1, further comprising:

7. An information processing method executed by an information processing device, A first acquisition step of acquiring product information of a target product, the product information including one or more target product features that are one or more features related to the target product, on a product page of the target product sold on an electronic commerce (EC) site; A step of identifying a similar product that is similar to the target product among one or more other products that satisfy a predetermined condition and are sold on an EC site; A second acquisition step of acquiring product information of the similar product, the product information including one or more similar product features which are one or more features related to the similar product, on the product page of the similar product; A determination step of comparing the one or more target product features with the one or more similar product features, and determining a feature of the one or more target product features that is not included in the one or more similar product features as an improvement point for the product page of the target product; a presentation step of presenting the improvement points to a user; having the one or more target product features include one or more images of the target product, and the one or more similar product features include one or more images of the similar products; In the determination step, when the number of images of the target product and the number of images of the similar products are different, the number of images is determined as an improvement point for the product page of the target product.

23. An information processing method comprising:

8. An information processing program for causing a computer to execute information processing, the program comprising: A first acquisition process for acquiring product information of a target product, the product information including one or more target product features that are one or more features related to the target product, on a product page of the target product sold on an electronic commerce (EC) site; A process of identifying a similar product that is similar to the target product among one or more other products that satisfy a predetermined condition and are sold on an EC site; A second acquisition process for acquiring product information of the similar product, the product information including one or more similar product features that are one or more features related to the similar product, on the product page of the similar product; A determination process of comparing the one or more target product features with the one or more similar product features, and determining a feature of the one or more target product features that is not included in the one or more similar product features as an improvement point in the product page of the target product; and a presentation process of presenting the improvement points to a user, the one or more target product features include one or more images of the target product, and the one or more similar product features include one or more images of the similar products; The determination process includes, when a number of images of the target product and a number of images of the similar products are different, determining the number of images as an improvement point in the product page of the target product. program.

Images