Information processing apparatus, information processing method, and information processing program

Abstract

To recognize a probability that a fraud will be performed on a product.SOLUTION: An information processing apparatus according to the present application includes: a collection unit which collects behavioral histories showing sequences of user's behaviors relating to products; an acquisition unit which acquires histories of fraudulent payments performed on the products; a generation unit which generates, on the basis of the behavioral histories, vectors corresponding to the products so that the more similar the sequences of user's behaviors relating to the products become, the more similar the vectors become; and an estimation unit which estimates a probability that a fraudulent payment will be performed on a prescribed product, on the basis of similarity between a vector corresponding to the prescribed product and a vector corresponding to a product for which a fraudulent payment has been performed.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and an information processing program.

Background Art

[0002] Conventionally, techniques for detecting illegal acts on a network are known. As an example of such a technique, a technique for estimating the presence or absence of illegal acts in the actions of Internet auction entities (for example, users, shops, etc.) is known.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, it cannot be said that the above-described technique can grasp the probability that fraud is committed against a product.

[0005] For example, the above-described technique only estimates the presence or absence of illegal acts in users, shops, etc., and there are cases where it cannot be said that the probability that fraud is committed against a product can be grasped.

[0006] The present application has been made in view of the above, and an object thereof is to grasp the probability that fraud is committed against a product.

Means for Solving the Problems

[0007] The information processing apparatus according to the present application includes a collection unit that collects an action history indicating the order of actions of a user regarding a product, an acquisition unit that acquires a history of unauthorized settlement for a product, and a generation unit that generates, based on the action history, a vector corresponding to each product such that the vectors are more similar as the order of actions of the user regarding the product is more similar. The information processing apparatus also includes an estimation unit that estimates the probability of unauthorized settlement for a predetermined product based on the similarity between the vector corresponding to the predetermined product and the vector corresponding to the product for which unauthorized settlement has been performed.

Effect of the Invention

[0008] According to one aspect of the embodiment, there is an effect that the probability of unauthorized actions on a product can be grasped.

Brief Description of the Drawings

[0009]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Mode for Carrying Out the Invention

[0010] Hereinafter, embodiments for implementing the information processing apparatus, information processing method, and information processing program according to the present application (hereinafter referred to as "embodiments") will be described in detail with reference to the drawings. Note that the information processing apparatus, information processing method, and information processing program according to the present application are not limited by this embodiment. Also, in the following embodiments, the same parts are denoted by the same reference numerals, and duplicate explanations are omitted.

[0011] 〔1. Embodiment〕 Using FIG. 1, the information processing realized by the information processing apparatus and the like of this embodiment will be described. FIG. 1 is a diagram showing an example of information processing according to the embodiment. In FIG. 1, it is assumed that the information processing according to the embodiment is realized by an information processing apparatus 10 which is an example of the information processing according to the present embodiment.

[0012] As shown in FIG. 1, the information processing system 1 according to the embodiment includes an information processing apparatus 10, a user terminal 100, and a provider terminal 200. The information processing apparatus 10, the user terminal 100, and the provider terminal 200 are connected to be communicable with each other by wire or wirelessly via a network N (for example, see FIG. 2). The network N is, for example, a WAN (Wide Area Network) such as the Internet. Note that the information processing system 1 shown in FIG. 1 may include a plurality of information processing apparatuses 10, a plurality of user terminals 100, and a plurality of provider terminals 200.

[0013] The information processing apparatus 10 shown in FIG. 1 is an information processing apparatus that performs information processing, and is realized by, for example, a server apparatus, a cloud system, or the like. For example, the information processing apparatus 10 provides an e-commerce service #1 to users. Here, the e-commerce service #1 may be, for example, a free market service in which a product offered by a user (provider) is purchased by another user, an auction service, etc., an EC site of a company site operated by an administrator of the information processing apparatus, or an electronic mall where a plurality of stores have stores.

[0014] Note that the information processing apparatus 10 may have a function as a web server that provides a website related to the e-commerce service #1. Further, the information processing apparatus 10 may be a device that distributes information to be displayed on an application related to the e-commerce service #1 installed on the user terminal 100 or the provider terminal 200 to the user terminal 100. Further, the information processing apparatus 10 may be a server that distributes the application data itself.

[0015] Further, the information processing apparatus 10 may function as a distribution device that distributes control information to the user terminal 100 or the provider terminal 200. Here, the control information is described, for example, by a script language such as JavaScript (registered trademark) or a style sheet language such as CSS (Cascading Style Sheets). Note that the application itself distributed from the information processing apparatus 10 may be regarded as control information.

[0016] The user terminal 100 shown in FIG. 1 is an information processing apparatus used by a user. The user terminal 100 is realized, for example, by a smartphone, a tablet terminal, a notebook PC (Personal Computer), a desktop PC, a mobile phone, a PDA (Personal Digital Assistant), or the like. Further, the user terminal 100 displays information distributed by the information processing apparatus 10 or the like on a web browser or an application. Note that, in the example shown in FIG. 1, the case where the user terminal 100 is a smartphone is shown.

[0017] The provider terminal 200 shown in FIG. 1 is an information processing apparatus used by a provider (for example, a user who has listed products on the e-commerce service #1, a company (manufacturer, etc.) that provides products, a store manager who has a store on the e-commerce service #1, etc.) to provide products to users via the e-commerce service #1. The provider terminal 200 is realized, for example, by a smartphone, a tablet terminal, a notebook PC, a desktop PC, a mobile phone, a PDA, or the like. Also, the user terminal 100 displays information distributed by the information processing apparatus 10 or the like on a web browser or an application. In the example shown in FIG. 1, the case where the provider terminal 200 is a notebook PC is shown.

[0018] Hereinafter, the information processing performed by the information processing apparatus 10 will be described with reference to FIG. 1. In the following description, the user terminals 100-1 to 100-N (N is an arbitrary natural number) will be described according to the users who use the user terminal 100. For example, the user terminal 100-1 is the user terminal 100 used by a user (user U1) identified by the user ID "UID#1". Also, hereinafter, when the user terminals 100-1 to 100-N are described without particular distinction, they will be described as the user terminal 100. Also, in the following description, the user terminal 100 may be regarded as the same as the user. That is, hereinafter, the user can also be read as the user terminal 100.

[0019] Also, in the following description, the provider terminals 200-1 to 200-N (N is an arbitrary natural number) will be described according to the providers who use the provider terminal 200. For example, the provider terminal 200-1 is the provider terminal 200 used by a provider (provider P1) identified by the provider ID "PID#1". Also, hereinafter, when the provider terminals 200-1 to 200-N are described without particular distinction, they will be described as the provider terminal 200. Also, in the following description, the provider terminal 200 may be regarded as the same as the provider. That is, hereinafter, the provider can also be read as the provider terminal 200.

[0020] First, the information processing device 10 collects the user's behavior history H1 regarding the products provided in the e-commerce service #1 from the user terminal 100 (step S1). For example, the information processing device 10 collects the behavior history H1 indicating the order of the user's actions regarding the products. To give a specific example, the information processing device 10 collects the behavior history H1 indicating the order in which the user viewed the products.

[0021] Subsequently, the information processing device 10 obtains, from the provider terminal 200, fraud settlement information indicating the history of products for which fraudulent settlements have been made in the e-commerce service #1 (step S2). For example, the information processing device 10 obtains fraud settlement information indicating the products for which a user has taken over another user's account in the e-commerce service #1 and made a settlement. Also, the information processing device 10 obtains fraud settlement information indicating the products for which a user has illegally used another user's credit card or the like and made a settlement. To give a specific example, the information processing device 10 obtains, as fraud settlement information, identification information for identifying the products for which fraudulent settlements have been made and product information of the products (for example, product name, price, category, quantity, product description, etc.).

[0022] Note that the information processing device 10 may obtain the fraud settlement information stored in its own storage unit from the storage unit. Also, the information processing device 10 may obtain fraud settlement information from the user terminal 100. Further, the information processing device 10 may obtain fraud settlement information from another server device that provides services related to settlements using a credit card or the like.

[0023] Subsequently, the information processing device 10 generates a vector (feature amount) corresponding to each product indicated by the behavior history based on the behavior history H1 collected in step S1 (step S3). For example, the information processing device 10 generates a vector corresponding to each product such that the more similar the other products viewed by the user before and after viewing the product corresponding to the vector, the more similar the vector.

[0024] For example, as shown in the action history H1, assume that user U1 views products in the order of product A, product B, product C, product D, and product E, and user U2 views products in the order of product A, product B, product P, product D, and product E. In such a case, as shown in the vector group V1, the information processing apparatus 10 generates vectors corresponding to each product such that the vector corresponding to product C and the vector corresponding to product P are similar.

[0025] Subsequently, the information processing apparatus 10 estimates the probability that an illegal settlement is made for a product (step S4). For example, the information processing apparatus 10 estimates the probability that an illegal settlement is made for a target product based on the similarity between the vector corresponding to the product indicated by the illegal settlement information and the vector corresponding to the target product. As a specific example, when an illegal settlement is made for product C, the information processing apparatus 10 estimates that product P has a higher probability of an illegal settlement than products A, B, D, and E.

[0026] Subsequently, the information processing apparatus 10 notifies the provider terminal 200 used by the provider of the product for which the probability of an illegal settlement is equal to or higher than a predetermined threshold of information regarding the illegal settlement (step S5). For example, when the probability that an illegal settlement is made for product P is equal to or higher than a predetermined threshold, the information processing apparatus 10 notifies the provider terminal 200 used by the provider of product P of an alert indicating that there is a high possibility of an illegal settlement being made for product P.

[0027] Note that the information processing apparatus 10 may estimate the probability that an illegal settlement is made for a product by using a model that has learned the relationship between the vector corresponding to the product and the information indicating whether an illegal settlement has been made for the product. For example, when a vector corresponding to a product for which an illegal settlement has been made is input, the information processing apparatus 10 performs learning of the model so as to output information (for example, a probability (score) of "1") indicating that an illegal settlement has been made for the product. Further, when a vector corresponding to a product for which no illegal settlement has been made is input, the information processing apparatus 10 performs learning of the model so as to output information (for example, a probability of "0") indicating that no illegal settlement has been made for the product. Then, the information processing apparatus 10 inputs the vector corresponding to the product to be estimated for probability into the model, and when the output probability is equal to or higher than a predetermined threshold, it notifies the provider terminal 200 used by the provider of the product that there is a high possibility that an illegal settlement has been made for the product.

[0028] Note that any known technique can be applied to the learning of the model, and a learning method appropriately selected according to the information used as learning data may be used. For example, the learning of the model may be performed using various conventional techniques related to machine learning (for example, techniques related to supervised machine learning such as SVM (Support Vector Machine)). Further, techniques of deep learning may be used for the learning of the model. For example, various deep learning techniques such as RNN (Recurrent Neural Network) and CNN (Convolutional Neural Network) may be used for the learning of the model.

[0029] As described above, the information processing apparatus 10 according to the embodiment generates the vector corresponding to each product so that the vectors are more similar as the order of the actions of the user related to the product is more similar, and estimates the probability that an illegal settlement is made for the product based on the similarity to the vector corresponding to the product for which an illegal settlement has been made. Thereby, the information processing apparatus 10 according to the embodiment can grasp the probability that an illegal act is performed on the product.

[0030] Conventionally, regarding the estimation of the probability that a product becomes a target of illegal settlement, keywords of products that are likely to be targets of illegal settlement (for example, popular products, highly convertible products, etc.) are listed, and matching with the list (rule-based determination) or scoring by a natural language processing model using the character string of the product name of products that are likely to be targets of illegal settlement has been used. However, since the keywords of products cover a wide range, rule-based determination incurs operation costs for list registration and may result in a delayed response. Also, the accuracy of scoring based on the product name may not be sufficient.

[0031] In contrast, the information processing apparatus 10 according to the embodiment can estimate the probability that a product becomes a target of illegal settlement based on the history of the user's actions regarding the product without requiring metadata such as the product name. Further, since the information processing apparatus 10 according to the embodiment can represent the similarity of products that do not appear in the metadata, it is possible to represent the similarity with products that have become targets of illegal settlement with higher accuracy, and the probability that a product becomes a target of illegal settlement can be estimated with higher accuracy.

[0032] 〔2. Other processing examples〕 Note that the above-described processing is merely an example, and the information processing apparatus 10 may perform various processes using various information. Regarding this point, examples are listed below.

[0033] 〔2-1. Regarding the generation of vectors based on purchase history〕 In the example of FIG. 1, the information processing apparatus 10 may collect an action history indicating the order in which the user purchased the products from the user terminal 100. Then, the information processing apparatus 10 generates a vector corresponding to each product such that the other products purchased by the user before and after purchasing the product corresponding to the vector are more similar to each other. For example, assume that user U1 purchases products in the order of product A, product B, product C, product D, and product E, and user U2 purchases products in the order of product A, product B, product P, product D, and product E. In such a case, the information processing apparatus 10 generates vectors corresponding to each product such that the vector corresponding to product C and the vector corresponding to product P are similar to each other.

[0034] [2-2. Generation of Vectors Based on Product Information] In the example of FIG. 1, in addition to the action history, the information processing apparatus 10 may further collect product information (e.g., product name, price, product category, quantity, product description, etc.) regarding the products viewed by the user. Then, the information processing apparatus 10 generates a vector corresponding to each product based on the action history indicating the order in which the user viewed the products and the product information regarding the product such that the vectors are more similar to each other as the product information is more similar to each other.

[0035] Note that, in addition to the action history, the information processing apparatus 10 may further collect product information regarding the products purchased by the user. Then, the information processing apparatus 10 generates a vector corresponding to each product based on the action history indicating the order in which the user purchased the products and the product information regarding the product such that the vectors are more similar to each other as the product information is more similar to each other.

[0036] [3. Configuration of Information Processing Apparatus] Next, the configuration of the information processing apparatus 10 will be described with reference to FIG. 2. FIG. 2 is a diagram showing a configuration example of the information processing apparatus 10 according to the embodiment. As shown in FIG. 2, the information processing apparatus 10 includes a communication unit 20, a storage unit 30, and a control unit 40.

[0037] (Regarding Communication Unit 20) The communication unit 20 is realized by, for example, a NIC (Network Interface Card) or the like. The communication unit 20 is connected to the network N by wire or wirelessly, and transmits and receives information to and from the user terminal 100, the provider terminal 200, and the like.

[0038] (Regarding the storage unit 30) The storage unit 30 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. As shown in FIG. 2, the storage unit 30 has an action history database 31 and an unauthorized settlement information database 32. Note that the storage unit 30 may further have a database that stores a model that has learned the relationship between a vector corresponding to a product and information indicating whether an unauthorized settlement has been made for the product.

[0039] (Regarding the action history database 31) The action history database 31 stores various types of information related to the history of the user's actions. Here, an example of the information stored in the action history database 31 will be described with reference to FIG. 3. FIG. 3 is a diagram showing an example of the action history database 31. In the example of FIG. 3, the action history database 31 has items such as "user ID", "browsing history", and "purchase history".

[0040] The "user ID" indicates identification information for identifying the user. The "browsing history" indicates information related to the user's browsing of products, and has items such as "product ID", "browsed product information", and "browsing date and time". The "product ID" indicates identification information for identifying the product browsed by the user. The "browsed product information" indicates information related to the product browsed by the user, and stores information such as the name of the product, the price of the product, and the category. The "browsing date and time" indicates the date and time when the user browsed the product.

[0041] The "purchase history" indicates information regarding the user's browsing of products, and has items such as "product ID", "purchased product information", "purchase date and time", etc. The "product ID" indicates identification information for identifying the product purchased by the user. The "purchased product information" indicates information regarding the product purchased by the user, and stores information such as the name of the product, the price of the product, the category, the quantity, the product description text, etc. The "purchase date and time" indicates the date and time when the user purchased the product.

[0042] That is, in FIG. 3, the product browsed by the user identified by the user ID "UID#1" is identified by the product ID "CID#1", the browsed product information of the browsed product is "browsed product information #1", the browsing date and time is "browsing date and time #1", and an example is shown where the product purchased by the user is identified by the product ID "CID#2", the purchased product information of the purchased product is "purchased product information #1", and the purchase date and time is "purchase date and time #1".

[0043] (Regarding the unauthorized settlement information database 32) The unauthorized settlement information database 32 indicates the history of products for which unauthorized settlements have been made. Here, using FIG. 4, an example of the information stored in the unauthorized settlement information database 32 will be described. FIG. 4 is a diagram showing an example of the unauthorized settlement information database 32. In the example of FIG. 4, the unauthorized settlement information database 32 has items such as "unauthorized settlement ID", "product ID", and "product information".

[0044] The "unauthorized settlement ID" indicates identification information for identifying an unauthorized settlement. The "product ID" indicates identification information for identifying the product for which an unauthorized settlement has been made. The "product information" indicates information regarding the product for which an unauthorized settlement has been made, and stores information such as the name of the product, the price of the product, the category, etc.

[0045] That is, in FIG. 4, an example is shown where the product for which an unauthorized settlement identified by the unauthorized settlement ID "SID#1" has been made is identified by the product ID "CID#1", and the product information of the said product is "product information #1".

[0046] (Regarding the control unit 40) The control unit 40 is a controller, which is realized, for example, by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), etc., when various programs stored in the storage device inside the information processing device 10 are executed with the RAM as a working area. Also, the control unit 40 is a controller and is realized, for example, by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). As shown in FIG. 2, the control unit 40 according to the embodiment has a collection unit 41, an acquisition unit 42, a generation unit 43, an estimation unit 44, and a provision unit 45, and realizes or executes the information processing functions and operations described below.

[0047] (Regarding the collection unit 41) The collection unit 41 collects an action history indicating the order of a user's actions regarding a product. For example, in the example of FIG. 1, the collection unit 41 collects the action history H1 of the user regarding the products provided in the e-commerce service #1 from the user terminal 100 and stores it in the action history database 31.

[0048] Also, the collection unit 41 may collect an action history indicating the order in which the user viewed the products. For example, in the example of FIG. 1, the collection unit 41 collects the action history H1 indicating the order in which the user viewed the products.

[0049] Also, the collection unit 41 may collect an action history indicating the order in which the user purchased the products. For example, in the example of FIG. 1, the collection unit 41 collects from the user terminal 100 the action history indicating the order in which the user purchased the products.

[0050] Also, the collection unit 41 may further collect product information regarding the products. For example, in the example of FIG. 1, in addition to the action history, the collection unit 41 further collects the product information regarding the products viewed by the user.

[0051] (Regarding the acquisition unit 42) The acquisition unit 42 acquires the history of illegal settlements made for the products. For example, in the example of FIG. 1, the acquisition unit 42 acquires, from the provider terminal 200, the illegal settlement information indicating the history of products for which illegal settlements have been made in the e-commerce service #1, and stores it in the illegal settlement information database 32.

[0052] (Regarding the generation unit 43) The generation unit 43 generates, based on the behavior history, vectors corresponding to each product such that the vectors are more similar as the order of the user's behavior regarding the products is more similar. For example, in the example of FIG. 1, the generation unit 43 refers to the behavior history database 31 and generates, based on the behavior history H1, vectors corresponding to each product indicated by the behavior history such that the vectors are more similar as the order of the user's behavior regarding the products is more similar.

[0053] Also, the generation unit 43 may generate, based on the behavior history, vectors corresponding to each product such that the vectors are more similar as the other products browsed by the user before and after browsing the products corresponding to the vectors are more similar. For example, in the example of FIG. 1, when the user U1 browses products in the order of product A, product B, product C, product D, product E, and the user U2 browses products in the order of product A, product B, product P, product D, product E, the generation unit 43 generates vectors corresponding to each product such that the vector corresponding to product C and the vector corresponding to product P are similar.

[0054] Note that the generation unit 43 may generate a vector corresponding to a product using a model that has learned the relationship of the order of the products viewed by the user from a large number of behavior histories. For example, the generation unit 43 inputs the product IDs of the products viewed before and after the product to be vectorized (the number of products before and after can be arbitrary) into the model, and uses a model (for example, the CBOW model of Word2Vec) in which the vector output from the model is learned as the correct data of the product to be vectorized to generate a vector corresponding to the product. Taking FIG. 1 as an example, for example, the generation unit 43 inputs the product IDs of products A, B, D, and E viewed before and after product C into the model and generates a vector corresponding to product C.

[0055] Also, for example, the generation unit 43 inputs the product ID of the target product into the model, and uses a model (for example, the SkipGram model of Word2Vec) in which the vector output from the model is learned as the correct data of the products viewed before and after the target product (the number of products before and after can be arbitrary) to generate a vector corresponding to the product. Taking FIG. 1 as an example, for example, the generation unit 43 inputs the product ID of product C into the model and generates vectors corresponding to products A, B, D, and E viewed before and after product C.

[0056] Also, the generation unit 43 may generate vectors corresponding to each product based on the behavior history so that the vectors become more similar as the other products purchased by the user before and after purchasing the product corresponding to the vector are more similar. For example, in the example of FIG. 1, when the user U1 purchases products in the order of products A, B, C, D, and E, and the user U2 purchases products in the order of products A, B, P, D, and E, the generation unit 43 generates vectors corresponding to each product so that the vector corresponding to product C and the vector corresponding to product P are similar.

[0057] Note that the generation unit 43 may generate a vector corresponding to a product by using a model that has learned the relationship of the order of the products purchased by the user from a large amount of behavior history. For example, the generation unit 43 inputs the product IDs of the products purchased before and after the product to be vectorized (the number of products before and after can be arbitrary) into the model, and uses a model (for example, the CBOW model of Word2Vec) in which the vector output from the model is used as the correct data of the product to be vectorized to generate a vector corresponding to the product. Taking FIG. 1 as an example, for example, the generation unit 43 inputs the product IDs of products A, B, D, and E purchased before and after product C into the model and generates a vector corresponding to product C.

[0058] Also, for example, the generation unit 43 inputs the product ID of the target product into the model, and uses a model (for example, the SkipGram model of Word2Vec) in which the vector output from the model is used as the correct data of the products purchased before and after the target product (the number of products before and after can be arbitrary) to generate a vector corresponding to the product. Taking FIG. 1 as an example, for example, the generation unit 43 inputs the product ID of product C into the model and generates vectors corresponding to products A, B, D, and E purchased before and after product C.

[0059] Further, the generation unit 43 may generate vectors corresponding to each product so that the more similar the product information is, the more similar the vectors are. For example, in the example of FIG. 1, the generation unit 43 generates vectors corresponding to each product based on the behavior history indicating the order in which the user viewed the products and the product information regarding the product, so that the more similar the product information (for example, the name of the product, the price of the product, the category, the quantity, the product description, etc.) is, the more similar the vectors are. Also, the generation unit 43 generates vectors corresponding to each product based on the behavior history indicating the order in which the user purchased the products and the product information regarding the product, so that the more similar the product information is, the more similar the vectors are.

[0060] Note that the generation unit 43 may generate a vector corresponding to a product by using a model that has learned the relationship of the order of the products viewed by the user from a large amount of behavior history. For example, the generation unit 43 inputs product information (for example, product name, product price, category, quantity, product description, etc.) of the products viewed before and after the product to be vectorized (the number of products before and after is arbitrary) into the model, and uses a model (for example, the CBOW model of Word2Vec) in which the vector output from the model is learned as the correct data of the product to be vectorized to generate a vector corresponding to the product. Taking FIG. 1 as an example, for example, the generation unit 43 inputs the product information of products A, B, D, and E viewed before and after product C into the model and generates a vector corresponding to product C.

[0061] Also, for example, the generation unit 43 inputs the product information of the target product into the model, and uses a model (for example, the SkipGram model of Word2Vec) in which the vector output from the model is learned as the correct data of the products viewed before and after the target product (the number of products before and after is arbitrary) to generate a vector corresponding to the product. Taking FIG. 1 as an example, for example, the generation unit 43 inputs the product information of product C into the model and generates vectors corresponding to products A, B, D, and E viewed before and after product C.

[0062] Note that the generation unit 43 may generate a vector corresponding to a product by using a model that has learned the relationship of the order of the products purchased by the user from a large amount of behavior history. For example, the generation unit 43 inputs product information (for example, product name, product price, category, quantity, product description, etc.) of the products purchased before and after the product to be vectorized (the number of products before and after is arbitrary) into the model, and uses a model (for example, the CBOW model of Word2Vec) in which the vector output from the model is learned as the correct data of the product to be vectorized to generate a vector corresponding to the product. Taking FIG. 1 as an example, for example, the generation unit 43 inputs the product information of products A, B, D, and E purchased before and after product C into the model and generates a vector corresponding to product C.

[0063] Also, for example, the generation unit 43 inputs the product information of the target product into the model, and uses a model (for example, the SkipGram model of Word2Vec) that has been trained with the vectors output from the model as the correct data of the products purchased before and after the target product (the number of products before and after can be arbitrary) to generate a vector corresponding to the product. Taking Figure 1 as an example, for example, the generation unit 43 inputs the product information of product C into the model and generates vectors corresponding to products A, B, D, and E purchased before and after product C.

[0064] Note that the storage unit 30 may store a model that has learned the relationship of the order of the products browsed by the user, and a model that has learned the relationship of the order of the products purchased by the user. Then, the generation unit 43 refers to the storage unit 30 and uses the model thus trained to generate a vector corresponding to the product.

[0065] (Regarding the estimation unit 44) The estimation unit 44 estimates the probability that an illegal settlement is made for a given product based on the similarity between the vector corresponding to the given product and the vector corresponding to the product for which an illegal settlement has been made. For example, in the example of Figure 1, the estimation unit 44 refers to the illegal settlement information database 32 and estimates the probability that an illegal settlement is made for the target product based on the similarity between the vector corresponding to the product indicated by the illegal settlement information and the vector corresponding to the target product.

[0066] Note that the estimation unit 44 may further refer to the storage unit 30 and use a model that has learned the relationship between the vector corresponding to the product and the information indicating whether an illegal settlement has been made for the product to estimate the probability that an illegal settlement is made for the product.

[0067] (Regarding the provision unit 45) The providing unit 45 notifies the provider of a product, for which the probability of an illegal settlement is equal to or higher than a predetermined threshold value, of information regarding the illegal settlement. For example, in the example of FIG. 1, when the probability of an illegal settlement for product P is equal to or higher than a predetermined threshold value, the providing unit 45 notifies the provider terminal 200 used by the provider of product P of an alert indicating that there is a high possibility of an illegal settlement for product P.

[0068] [4. Information Processing Flow] Using FIG. 5, the procedure of information processing of the information processing apparatus 10 according to the embodiment will be described. FIG. 5 is a flowchart showing an example of the procedure of information processing according to the embodiment.

[0069] As shown in FIG. 5, the information processing apparatus 10 collects an action history indicating the order of actions of the user regarding the product (step S101). Subsequently, the information processing apparatus 10 acquires a history of illegal settlements made for the product (step S102). Subsequently, based on the action history, the information processing apparatus 10 generates a vector corresponding to each product such that the vectors become more similar as the order of actions of the user regarding the product becomes more similar (step S103). Subsequently, based on the similarity between the vector corresponding to a predetermined product and the vector corresponding to the product for which an illegal settlement has been made, the information processing apparatus 10 estimates the probability of an illegal settlement being made for the predetermined product (step S104). Subsequently, the information processing apparatus 10 determines whether the probability of an illegal settlement being made for the predetermined product is equal to or higher than a predetermined threshold value (step S105). If the probability is equal to or higher than the predetermined threshold value (step S105; Yes), the information processing apparatus 10 notifies the provider of the predetermined product of information regarding the illegal settlement (step S106) and ends the process.

[0070] On the other hand, if the probability is less than the predetermined threshold value (step S105; No), the information processing apparatus 10 ends the process without notifying the provider of the predetermined product of information regarding the illegal settlement.

[0071] [5. Modification Example] The above-described embodiments are merely examples, and various modifications and applications are possible.

[0072] Regarding the processing mode Among the various processes described in the above embodiments, all or part of the processes described as being automatically performed can also be manually performed, and conversely, all or part of the processes described as being manually performed can be automatically performed by known methods. In addition, regarding the processing procedures, specific names, and information including various data and parameters shown in the above text and drawings, they can be arbitrarily changed unless otherwise specified. For example, the various information shown in each figure is not limited to the illustrated information.

[0073] Also, each component of each device shown in the drawings is a functional concept, and it is not necessarily physically configured as shown in the drawings. That is, the specific form of the distribution and integration of each device is not limited to that shown, and all or part of it can be functionally or physically distributed and integrated in any unit according to various loads and usage situations.

[0074] In addition, the above-described embodiments can be appropriately combined within a range that does not conflict with the processing content.

[0075] 〔6. Effects〕 As described above, the information processing apparatus 10 according to the embodiment includes a collection unit 41, an acquisition unit 42, a generation unit 43, an estimation unit 44, and a provision unit 45. The collection unit 41 collects an action history indicating the order of a user's actions regarding a product. The acquisition unit 42 acquires a history of unauthorized settlements made for a product. The generation unit 43 generates a vector corresponding to each product based on the action history such that the vectors are more similar as the order of the user's actions regarding the product is more similar. The estimation unit 44 estimates the probability that an unauthorized settlement is made for a predetermined product based on the similarity between the vector corresponding to the predetermined product and the vector corresponding to the product for which an unauthorized settlement has been made. The provision unit 45 notifies the provider of a product for which the probability of an unauthorized settlement is equal to or greater than a predetermined threshold of information regarding the unauthorized settlement.

[0076] As a result, the information processing apparatus 10 according to the embodiment generates vectors corresponding to each product so that the vectors become more similar as the order of the user's actions regarding the product is more similar, and based on the similarity with the vector corresponding to the product for which an illegal settlement has been made, it is possible to estimate the probability that an illegal settlement is made for the product, so that it is possible to grasp the probability that an illegal act is made for the product.

[0077] Further, in the information processing apparatus 10 according to the embodiment, for example, the collection unit 41 collects an action history indicating the order in which the user has browsed the products. Then, the generation unit 43 generates vectors corresponding to each product based on the action history so that the vectors become more similar as the other products browsed by the user before and after browsing the product corresponding to the vector are more similar. Also, the collection unit 41 collects an action history indicating the order in which the user has purchased the products. Then, the generation unit 43 generates vectors corresponding to each product based on the action history so that the vectors become more similar as the other products purchased by the user before and after purchasing the product corresponding to the vector are more similar.

[0078] As a result, the information processing apparatus 10 according to the embodiment can generate vectors corresponding to the products based on various actions of the user regarding the products, so that the probability that an illegal act is made for the products can be estimated with higher accuracy.

[0079] Further, in the information processing apparatus 10 according to the embodiment, for example, the collection unit 41 further collects product information regarding the products. Then, the generation unit 43 generates vectors corresponding to each product so that the vectors become more similar as the product information is more similar.

[0080] As a result, the information processing apparatus 10 according to the embodiment can generate vectors corresponding to the products using the product information regarding the products in addition to the actions of the user, so that the probability that an illegal act is made for the products can be estimated with higher accuracy.

[0081] [7. Hardware Configuration] Also, the information processing apparatus 10 according to each of the above-described embodiments is realized by, for example, a computer 1000 having a configuration as shown in FIG. 6. Hereinafter, the information processing apparatus 10 will be described as an example. FIG. 6 is a hardware configuration diagram showing an example of a computer that realizes the functions of the information processing apparatus 10. The computer 1000 includes a CPU 1100, a ROM 1200, a RAM 1300, an HDD 1400, a communication interface (I / F) 1500, an input / output interface (I / F) 1600, and a media interface (I / F) 1700.

[0082] The CPU 1100 operates based on programs stored in the ROM 1200 or the HDD 1400, and controls each part. The ROM 1200 stores a boot program executed by the CPU 1100 when the computer 1000 is started up, programs dependent on the hardware of the computer 1000, and the like.

[0083] The HDD 1400 stores programs executed by the CPU 1100, data used by such programs, and the like. The communication interface 1500 receives data from other devices via the communication network 500 (corresponding to the network N of the embodiment) and sends it to the CPU 1100, and also sends data generated by the CPU 1100 via the communication network 500 to other devices.

[0084] The CPU 1100 controls output devices such as a display and a printer, and input devices such as a keyboard and a mouse via the input / output interface 1600. The CPU 1100 acquires data from the input device via the input / output interface 1600. Also, the CPU 1100 outputs data generated via the input / output interface 1600 to the output device.

[0085] The media interface 1700 reads the program or data stored in the recording medium 1800 and provides it to the CPU 1100 via the RAM 1300. The CPU 1100 loads such a program from the recording medium 1800 onto the RAM 1300 via the media interface 1700 and executes the loaded program. The recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc), PD (Phase change rewritable Disk), etc., a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory, etc.

[0086] For example, when the computer 1000 functions as the information processing apparatus 10, the CPU 1100 of the computer 1000 realizes the functions of the control unit 40 by executing the program loaded on the RAM 1300. Also, each data in the storage device of the information processing apparatus 10 is stored in the HDD 1400. The CPU 1100 of the computer 1000 reads and executes these programs from the recording medium 1800, but as another example, these programs may be acquired from other devices via a predetermined communication network.

[0087] [8. Others] As described above, some of the embodiments of the present application have been described in detail with reference to the drawings, but these are examples, and the present invention can be implemented in other forms in which various modifications and improvements are made based on the knowledge of those skilled in the art, starting from the aspects described in the column of the disclosure of the invention.

[0088] Also, the above-described information processing apparatus 10 can be flexibly configured, such as by calling an external platform or the like via an API (Application Programming Interface) or network computing depending on the function.

[0089] Also, the "section" described in the claims can be read as "means", "circuit", etc. For example, the collection section can be read as a collection means or a collection circuit.

Explanation of Symbols

[0090] 10 Information Processing Device 20 Communication Unit 30 Storage Unit 31 Action History Database 32 Illegal Settlement Information Database 40 Control Unit 41 Collection Unit 42 Acquisition Unit 43 Generation Unit 44 Estimation Unit 45 Provision Unit 100 User Terminal 200 Provider Terminal

Claims

1. A collecting unit that collects an action history indicating the order of a user's actions related to a product; An acquiring unit that acquires a history of unauthorized settlement for a product; A generating unit that generates, based on the action history, a vector corresponding to each product such that the vectors are more similar as the order of the user's actions related to the product is more similar; An estimating unit that estimates the probability of unauthorized settlement for a given product based on the similarity between the vector corresponding to the given product and the vector corresponding to the product for which unauthorized settlement has been made An information processing apparatus characterized by comprising the above.

2. The collecting unit collects the action history indicating the order in which the user viewed the products, The generating unit generates, based on the action history, a vector corresponding to each product such that the vectors are more similar as the other products viewed by the user before and after viewing the product corresponding to the vector are more similar The information processing apparatus according to claim 1, characterized by the above.

3. The collecting unit collects the action history indicating the order in which the user purchased the products, The generating unit generates, based on the action history, a vector corresponding to each product such that the vectors are more similar as the other products purchased by the user before and after purchasing the product corresponding to the vector are more similar The information processing apparatus according to claim 1, characterized by the above.

4. The collecting unit further collects product information related to the product, The generating unit generates a vector corresponding to each product such that the vectors are more similar as the product information is more similar The information processing apparatus according to claim 1, characterized by the above.

5. A notification unit that notifies the provider of a product for which the probability of unauthorized settlement is equal to or greater than a predetermined threshold of information related to unauthorized settlement The information processing apparatus according to claim 1, further characterized by comprising the above.

6. An information processing method executed by a computer, comprising: A collecting step of collecting an action history indicating the order of a user's actions related to a product; An acquiring step of acquiring a history of unauthorized settlement for a product; A generating step of generating, based on the action history, a vector corresponding to each product such that the vectors are more similar as the order of the user's actions related to the product is more similar; An estimating step of estimating the probability of unauthorized settlement for a given product based on the similarity between the vector corresponding to the given product and the vector corresponding to the product for which unauthorized settlement has been made An information processing method characterized by including the above.

7. A collection procedure for collecting an action history indicating the order of a user's actions regarding a product, An acquisition procedure for acquiring a history of unauthorized settlements made for a product, A generation procedure for generating, based on the action history, a vector corresponding to each product such that the vectors are more similar as the order of the user's actions regarding the product is more similar, An estimation procedure for estimating the probability that an unauthorized settlement is made for a given product based on the similarity between the vector corresponding to the given product and the vector corresponding to the product for which an unauthorized settlement has been made An information processing program characterized by causing a computer to execute the above.

Images