JP2018045505A - Determination device, determination method, and determination program - Google Patents

Abstract

An object of the present invention is to improve the selection accuracy of a transaction target proposed to a user. A determination apparatus according to the present application uses a distributed expression based on information related to a transaction object to specify another transaction object that is similar to the predetermined transaction object, and the other part specified by the specification unit. And a determination unit that determines whether or not to present the predetermined transaction object to a user based on a purchase history of the transaction object. [Selection] Figure 1

Description

  The present invention relates to a determination device, a determination method, and a determination program.

  In recent years, information distribution via the Internet has been actively performed. As an example of such information distribution, there is known a technique for proposing a product or service (hereinafter, referred to as “transaction target”) that is an object of electronic commerce based on past sales results. For example, there is known a technique for targeting a product whose design and color are similar to a user's preference among products to be traded.

JP 2006-071609 A

  However, in the above-described conventional technology, there are cases where it is not possible to accurately select a transaction target proposed to the user.

  For example, there is a case where a transaction object newly registered as an object of electronic commerce does not register what kind of design or color is the object of transaction. In such a case, it is not possible to specify whether or not the transaction target is similar to the user's preference, and the selection accuracy of the transaction target proposed to the user is deteriorated.

  The present application has been made in view of the above, and an object thereof is to improve the selection accuracy of a transaction target proposed to a user.

  The determination apparatus according to the present application uses a distributed expression based on information about a transaction target, specifies a specific unit that identifies another transaction target similar to the predetermined transaction target, and purchases of the other transaction target specified by the specification unit And a determination unit that determines whether or not to present the predetermined transaction target to a user based on a history.

  According to one aspect of the embodiment, the accuracy of selecting a transaction target proposed to the user is improved.

FIG. 1 is a diagram illustrating an example of a determination process executed by the determination apparatus according to the embodiment. FIG. 2 is a diagram illustrating a configuration example of the determination apparatus according to the embodiment. FIG. 3 is a diagram illustrating an example of information registered in the transaction target database according to the embodiment. FIG. 4 is a diagram illustrating an example of information registered in the purchase history database according to the embodiment. FIG. 5 is a diagram illustrating an example of information registered in the classification result database according to the embodiment. FIG. 6 is a flowchart illustrating an example of a flow of determination processing executed by the determination device according to the embodiment. FIG. 7 is a flowchart illustrating an example of a process in which the determination apparatus according to the embodiment presents a transaction target. FIG. 8 is a diagram illustrating an example of a hardware configuration.

  Hereinafter, modes for carrying out a determination apparatus, a determination method, and a determination program according to the present application (hereinafter referred to as “embodiments”) will be described in detail with reference to the drawings. Note that the determination device, the determination method, and the determination program according to the present application are not limited by this embodiment. In the following embodiments, the same portions are denoted by the same reference numerals, and redundant description is omitted.

[Embodiment]
[1-1. Example of determination device)
First, an example of determination processing executed by the determination device will be described with reference to FIG. FIG. 1 is a diagram illustrating an example of a determination process executed by the determination apparatus according to the embodiment. In FIG. 1, the determination device 10 can communicate with an EC (Electronic Commerce) server 100 and a terminal device 200 (for example, FIG. 2) used by a user U01 via a predetermined network N such as the Internet. Note that the determination device 10 may be able to communicate with an arbitrary number of EC servers 100 and an arbitrary number of terminal devices 200.

  The EC server 100 is an information processing device that provides various services related to electronic commerce, and is realized by a server device, a cloud system, or the like. For example, the EC server 100 provides the user U01 with an online shopping mall service that provides an arbitrary product or service (hereinafter referred to as “transaction object”). The EC server 100 may be a server that provides services related to auctions. That is, the EC server 100 may provide an arbitrary service as long as it is a service related to electronic commerce.

  The terminal device 200 is a terminal device used by the user U01, and is realized by, for example, a portable mobile terminal device such as a smartphone or a tablet, a PC (Personal Computer) terminal, a server device, or the like. For example, the terminal device 200 has a screen capable of displaying arbitrary content, and can display information related to the content to the user U01 by displaying the content received from the determination device 10.

  The determination device 10 selects content provided by the terminal device 200 to the user U01, that is, content related to a transaction target (for example, an advertisement). For example, the determination apparatus 10 determines whether to present content to the user U01 for each transaction target. And when the determination apparatus 10 determines with presenting a transaction object with respect to the user U01, the content regarding the transaction object is delivered to the terminal device 200.

[1-2. About judgment processing)
Here, in order to improve the browsing opportunities and purchase opportunities of the transaction object, the transaction object that is highly likely to be purchased by the user U01 is estimated based on the purchase history of each transaction object, and the content related to the estimated transaction object is distributed. The method of doing is conceivable. For example, the determination apparatus 10 specifies attribute information of a user who is a content distribution target, such as demographic attributes such as the gender, age, and residential area of the user U01, and psychographic attributes such as values and preferences. Further, the determination apparatus 10 specifies, for each transaction target, a purchase tendency such as what kind of attribute the user has purchased and how much the user has purchased at what date and time based on the purchase history.

  And the determination apparatus 10 can consider the process of selecting the transaction object in which purchase tendency corresponds with the attribute information of the user who becomes a content distribution destination among the transaction objects, and distributing the content of the selected transaction object. For example, when the user who is the distribution destination of the content is a woman, the determination device 10 distributes the content to be traded having a purchase tendency that the woman often purchases.

  However, with such a method, it is not possible to understand what purchasing tendency a transaction object having no purchase history, such as a newly registered transaction object, has. For this reason, in the prior art, there is a possibility that the selection accuracy of the transaction target proposed to the user is deteriorated. For example, in the prior art, even if a newly registered transaction object is a transaction object popular with men, the transaction object may be proposed to women.

  Therefore, the determination apparatus 10 executes the following determination process. First, the determination apparatus 10 specifies another transaction object similar to a predetermined transaction object using a distributed expression based on information on the transaction object. And the determination apparatus 10 determines whether a predetermined | prescribed transaction object is shown to a user based on the purchase history of the other identified transaction object. For example, the determination apparatus 10 specifies another transaction object similar to the predetermined transaction object using a distributed expression based on at least one of a description or a photograph of the predetermined transaction object. In addition, for example, the determination apparatus 10 identifies another transaction target that is estimated to have a purchase tendency similar to a predetermined transaction target. And the determination apparatus 10 determines whether a predetermined | prescribed transaction object is shown to a user based on the purchase history of the other identified transaction object.

  Here, transaction objects having the same purchase tendency are predicted to have similar attributes such as a category to which the transaction object belongs, appearance, performance, price range and the like. For this reason, it is predicted that the purchase tendency of a transaction object having no purchase history is similar to the purchase tendency of other transaction objects having similar attributes. Therefore, the determination apparatus 10 selects another transaction target that is predicted to have a similar purchase tendency based on the attribute of the predetermined transaction target, and determines the predetermined transaction target based on the purchase trend of the selected other transaction target. Is determined to be presented to the user.

  As a more specific example, the determination apparatus 10 classifies the transaction objects having a purchase history into a plurality of groups in accordance with the purchase tendency indicated by the purchase history. Moreover, the determination apparatus 10 specifies the transaction object similar to the attribute of the transaction object in which the attribute was newly registered from each transaction object. And the determination apparatus 10 acquires the purchase history of the specified transaction object, and determines whether a predetermined transaction object is shown to a user according to the acquired purchase history. For example, the determination apparatus 10 identifies a transaction target whose attribute is similar to an attribute of a predetermined transaction target among the transaction targets classified into a plurality of groups according to a purchase tendency, and classifies the same transaction target as the identified transaction target. Get the purchase history of the transaction target.

[1-3. (About attributes of transactions)
Here, if the determination apparatus 10 is information other than the purchase history and is information related to a transaction target, any information can be adopted as an attribute of the transaction target. For example, the determination device 10 may use a transaction target type (that is, category) such as an information processing device, a food item, or a beverage as the transaction target attribute. Moreover, the determination apparatus 10 is good also considering the price of a transaction object, a manufacturer, the store etc. which sell a transaction object in an online shopping mall as an attribute of transaction object. Further, the determination apparatus 10 may use the appearance, material, color, size (size), weight, and the like of the transaction object as attribute information. Moreover, the determination apparatus 10 is good also considering the purchase time etc. which are the performance of transaction object, the date when purchasing was started, etc. as attribute information.

  Here, in an electronic commerce such as an online shopping mall or an auction, there may be a case where an explanatory text indicating an attribute such as a transaction target state or performance is registered. Such an explanation is considered to indicate what kind of attribute the transaction object has. In addition, depending on the contents of such an explanatory note, it is considered that the tendency of users to purchase changes such that men frequently purchase or women frequently purchase. Therefore, the determination device 10 may use the transaction target description as transaction target attribute information.

  Further, in electronic commerce, an image to be traded may be registered. Such an image is considered to include the characteristics of the transaction object such as the appearance and atmosphere of the color and shape of the transaction object. Moreover, the determination apparatus 10 is good also considering the image of transaction object as attribute information.

  Also, in electronic commerce, it is considered that the tendency of users to purchase changes depending on the seller who sells the transaction, the seller's credit, and the like. Therefore, the determination apparatus 10 may use the seller that performs the transaction for sale, information indicating the credit of the seller, evaluation or review for the seller, and the like as the attribute information for the transaction.

  In addition to the information described above, the determination device 10 may employ any information related to the transaction target as the transaction target attribute. For example, when the seller who sells the transaction target registers the transaction target in an online shopping mall, an auction, or the like, the determination apparatus 10 uses any information registered as information explaining the transaction target as the transaction target. It may be adopted as attribute information. As will be described later, such attribute information is converted into a distributed expression indicating the contents and characteristics of the attribute information, and is stored in association with the transaction target. And the determination apparatus 10 will select other transaction objects similar to a predetermined transaction object, such as a newly registered transaction object, using such a distributed expression.

[1-4. (User attributes)
Here, the determination apparatus 10 can employ any information related to the user who is the transaction target presentation destination as the attribute of the user. For example, the determination apparatus 10 may use a user's demographic attribute or psychographic attribute as user attribute information. Moreover, the determination apparatus 10 is good also considering the purchase history which shows the log | history of the transaction object which the user purchased in the past as a user's attribute information. In other words, the determination device 10 can employ arbitrary information as the attribute information of the user if it is useful when selecting a transaction target that is likely to be purchased by the user.

[1-5. (About purchasing trends)
Here, the determination apparatus 10 may classify the transaction target based on an arbitrary purchase tendency. More specifically, the determination apparatus 10 may perform classification such as what kind of user is purchasing in electronic commerce, that is, classification according to the attribute of the user. For example, the determination device 10 is frequently used as a transaction target that men frequently purchase, a transaction target that women frequently purchase, a transaction target that users in their 30s frequently purchase, and the like. Classification according to the attribute type of the user to be purchased may be performed. Moreover, the determination apparatus 10 may perform classification according to a plurality of attribute types, such as a transaction target that a teenage male frequently purchases.

  Moreover, the determination apparatus 10 may perform classification that combines not only user information but also other information. For example, the determination apparatus 10 may perform classification according to the date and time when the transaction target is purchased, such as a transaction target that men frequently purchase on Sundays. Moreover, the determination apparatus 10 may perform classification according to the timing at which a transaction target is purchased, such as a transaction target (for example, a printer and ink) purchased at the same time as a certain transaction target. Here, for example, the determination apparatus 10 classifies ink and paper into the same group as a transaction target purchased after the printer is purchased. However, the determination device 10 does not use the printer as a transaction target purchased after the ink or paper is purchased. Classification may be performed in consideration of the order in which transaction objects are purchased.

  In this way, the determination apparatus 10 may include not only the attributes of each transaction target but also situations in which the transaction target is purchased such as day of the week, date and time, weather, temperature, atmospheric pressure, etc. Conditions such as the target and the order in which the transaction target is purchased may be included in the purchase tendency. In other words, the determination device 10 may classify each transaction target based on a purchase tendency in consideration of arbitrary information as long as it is information indicating a trend related to the purchase behavior of the transaction target.

[1-6. (Processing using distributed representation)
Here, the determination apparatus 10 executes a determination process using a distributed expression indicating an attribute of a transaction object as a technique for realizing the above-described process. More specifically, the determination apparatus 10 converts the registered transaction target attributes into the distributed representation so that the distributed representations of the transaction targets with similar purchasing tendencies are similar. Then, the determination device 10 converts the attribute of a predetermined transaction object such as a newly registered transaction object into a distributed expression, and the distributed expression is similar to the distributed expression of the predetermined transaction object among the registered transaction objects. Select the transaction object to be traded. That is, the determination apparatus 10 selects a transaction target that is estimated to have a purchase tendency similar to a newly registered transaction target among registered transaction targets based on the attribute of each transaction target. And the determination apparatus 10 specifies what kind of attribute the newly registered transaction object is likely to be purchased by the user based on the purchase history of the selected transaction object, and based on the identification result, It is determined whether or not to present the transaction target registered in the item to the user.

  Here, the determination apparatus 10 calculates a distributed expression indicating an attribute of a transaction target using a model learned by a predetermined learning process. Such a model can be realized by, for example, DNN (Deep Neural Network), RNN (Recurrent Neural Network), or the like.

  For example, the determination device 10 selects a transaction target having a sufficient purchase history as a learning target from among registered transaction targets. Moreover, the determination apparatus 10 specifies each purchase tendency of learning object based on the purchase history of the selected transaction object. The determination apparatus 10 is a model for determining a distributed representation of a transaction target from the transaction target attribute selected as a learning target, and a similar distributed representation as a distributed representation of a transaction target having a predetermined purchase tendency. Learn the model that determines That is, the determination apparatus 10 learns a model that classifies transaction objects having similar purchasing trends into the same group.

  For example, the determination apparatus 10 uses the purchase trend of the transaction target selected as the learning target as information (for example, a teacher signal) indicating the classification target of the transaction target, and attribute information indicating the attribute of the transaction target as the input information. To do. And the determination apparatus 10 learns a model using arbitrary learning methods so that a model may output a similar distributed expression when a plurality of transaction target attribute information with similar or the same purchasing tendency is input. . For example, when adopting DNN as a model, the determination apparatus 10 is similar when a photograph or a description of a transaction object having a similar purchase tendency, such as a transaction object purchased by a user having the same attribute, is input. Model learning using a learning method such as back-propagation method so that a dissimilar distributed expression is output when a photograph or description of a transaction object with a similar purchasing tendency is output. I do.

  In addition, for example, when the determination apparatus 10 performs classification based on a purchase tendency in consideration of time series, such as classification of transaction objects that are frequently purchased after a certain transaction object is purchased, the RNN May be adopted as a model. For example, the determination apparatus 10 inputs the attribute information of the transaction target to be purchased first into the model, and subsequently inputs the attribute information of the transaction target to be purchased later into the model, so that the transaction target to be purchased earlier is the same. Alternatively, a model for obtaining a distributed expression of a transaction object to be purchased later is learned so that distributed expressions of similar transaction objects are similar. The model obtained by such learning is a transaction target that is often purchased before a predetermined transaction target, or a transaction target that is often purchased before a predetermined transaction target is different from the learning target. The distributed representations will not be similar. For this reason, the determination apparatus 10 learns a model that performs classification based on a purchase tendency considering a time series, such as when performing classification of transaction objects that are frequently purchased after a certain transaction object is purchased. Can do.

  In the distributed representation calculated by such a model, not only the clear characteristics of the transaction object such as the appearance and performance of the user etc., but also the characteristic that the transaction object having a common purchasing tendency is implicitly shared ( For example, it is predicted to indicate a concept indicating a transaction object having a common purchasing tendency. Therefore, the determination apparatus 10 selects a transaction target having a similar purchase tendency to a new transaction target using the model obtained by the learning process described above. For example, the determination apparatus 10 inputs attribute information such as a newly registered transaction target image or description to the model, and obtains a distributed expression. The distributed representation obtained in this way is considered to be similar to other transaction objects that are estimated (predicted) to have similar purchasing trends. Therefore, the determination device 10 selects a transaction object having a distributed representation similar to that of the newly registered transaction object from among the registered transaction objects, so that a transaction having a purchase tendency similar to that of the newly registered transaction object. The target is specified, and it is determined whether to present the newly registered transaction target to the user based on the purchase tendency of the specified transaction target.

  More specifically, when receiving a content distribution request from the user U01, the determination device 10 selects a transaction target having a purchase tendency similar to a transaction target having no purchase history, and selects the purchase trend of the selected transaction target. Based on the matching result with the attribute of the user U01, it is determined whether or not to present a transaction target having no purchase history to the user U01. If the determination device 10 determines to present a transaction target without a purchase history to the user U01, the determination device 10 generates a transaction target advertisement content without a purchase history, and distributes the generated advertisement content to the user U01. To do. As a result, the determination apparatus 10 can present a transaction target that is highly likely to be purchased by the user even when there is a transaction target for which no purchase history exists. The selection accuracy can be improved.

[1-7. About an example of judgment processing)
Next, an example of determination processing executed by the determination device 10 will be described with reference to FIG. In the following description, the determination apparatus 10 calculates a distributed representation of a transaction target in which a purchase history exists using the model learned by the learning process described above, and uses the calculated distributed representation to determine the transaction target. Are classified according to the purchasing tendency.

  For example, in the example illustrated in FIG. 1, the determination apparatus 10 has classified the transaction objects # 1-1 to # 1-3 into the group # 1 having the purchase tendency # 1 and has the purchase tendency # 2. It is assumed that transaction object # 2-1 to transaction object # 2-3 have been classified into group # 2. The distributed expressions # 1-1 to # 1-3 of the transaction object # 1-1 to the transaction object # 1-3 classified into the group # 1 are similar to each other and are classified into the group # 2. The distributed representation # 2-1 to the distributed representation # 2-3 of the trading object # 2-1 to the trading object # 2-3 are similar to each other. Note that the determination apparatus 10 stores a plurality of groups corresponding to purchase trends in addition to the group # 1 and the group # 2.

  For example, the determination apparatus 10 can obtain a transaction target such as the name, attribute information, and purchase history of the transaction target #A that is a newly registered transaction target (hereinafter referred to as “new transaction target”) from the EC server 100 or the like. Information is acquired (step S1). In such a case, the determination apparatus 10 selects a transaction target group that is predicted to have a purchase tendency similar to the transaction target #A, which is a new transaction target, from among the transaction targets classified for each purchase trend. (Step S2). That is, the determination apparatus 10 specifies a transaction target having similar attributes of the transaction target #A and the distributed expression.

  For example, the determination apparatus 10 acquires the distributed representation #A of the transaction target #A from the attribute information of the transaction target #A, using the model learned by the learning process described above. Then, the determination apparatus 10 searches for a transaction object whose distributed expression is similar to the distributed expression #A. That is, the determination apparatus 10 selects a transaction target having a purchase tendency similar to that of the transaction target #A based on the attribute information of the transaction target #A. For example, as indicated by the solid line arrow, the determination device 10 determines that the transaction object # 2-1 if any of the distributed expression # 2-1 to the distributed expression # 2-3 is similar to the distributed expression #A of the transaction object #A. Group # 2 is selected as a transaction target group that is predicted to have a purchasing tendency similar to A.

  In the above-described search, it is not necessary that all of the distributed expressions # 2-1 to # 2-3 are similar to the distributed expression #A. For example, if any of the distributed expressions # 2-1 to # 2-3 is similar to the distributed expression #A of the transaction object #A, the determination apparatus 10 determines that the transaction object #A and the purchase tendency are similar. Group # 2 may be selected as a transaction target group that is predicted to be similar. For example, the determination apparatus 10 may select a group having the largest number of distributed expressions similar to the distributed expression #A among the groups. Further, in a search for similar distributed expressions, any search method using a Hamming distance or a cosine distance can be employed.

  Subsequently, the determination apparatus 10 estimates that the purchase history of the transaction object #A, which is a new transaction object, is similar to the purchase tendency based on the purchase history of the transaction object included in the selected group (step S3). For example, the determination apparatus 10 has a tendency that the purchase tendency # 2 of the group # 2 is frequently purchased by users having attributes such as the age group “30s”, the sex “female”, and the region “Hokkaido”. In this case, it is estimated that the purchase tendency of transaction object #A is purchase tendency # 2.

  Here, the determination apparatus 10 receives a request for requesting presentation of a transaction target from the user U01 (step S4). For example, the terminal device 200 used by the user U01 makes a request for advertisement distribution as a transaction target when searching in an online shopping mall or auction, or when browsing web content. In such a case, the determination apparatus 10 selects a transaction target whose purchase tendency is similar to the attribute of the user U01 (step S5). For example, the determination apparatus 10 compares the user information # 1 indicating the attribute of the user U01 with the estimated purchase trend # 2 of the transaction target #A. Then, when the user information # 1 and the purchase tendency # 2 match, the determination apparatus 10, for example, has the user attribute indicated by the purchase tendency # 2 and the user U01 attribute indicated by the user information # 1. Are similar or coincident with each other, it is determined that the transaction target #A is presented to the user U01. Note that the determination device 10 may determine that the transaction target # 2-1 to transaction target # 2-3 classified into the group # 2 are presented to the user U01.

  And the determination apparatus 10 produces | generates the content which introduces transaction object #A, and distributes the produced | generated content with respect to the user's U01 terminal device 200 (step S6). As a result, the determination apparatus 10 can present to the user U01 who is highly likely to purchase even if the transaction target #A has no purchase history.

[2. (Configuration of judgment device)
Hereinafter, an example of a functional configuration of the determination apparatus 10 that realizes the above-described determination process will be described. FIG. 2 is a diagram illustrating a configuration example of the determination apparatus according to the embodiment. As illustrated in FIG. 2, the determination device 10 includes a communication unit 20, a storage unit 30, and a control unit 40.

  The communication unit 20 is realized by, for example, a NIC (Network Interface Card). The communication unit 20 is connected to the network N in a wired or wireless manner, and transmits / receives information to / from the EC server 100 and the terminal device 200.

  The storage unit 30 is realized by, for example, a semiconductor memory device such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. Further, the storage unit 30 stores a transaction target database 31, a purchase history database 32, a classification result database 33, and a model database 34 (hereinafter may be collectively referred to as “databases 31 to 34”).

  The transaction target database 31 stores transaction target attribute information. For example, FIG. 3 is a diagram illustrating an example of information registered in the transaction target database according to the embodiment. As shown in FIG. 3, information having items such as “transaction object ID (Identifier)”, “attribute information”, and “distributed expression” is registered in the transaction object database 31. The “attribute information” includes information such as “category”, “price”, “manufacturer”, “image”, “description”, “purchase date”, etc., that is, each item of attribute information. In addition, it is assumed that other arbitrary information indicating the transaction target attribute is registered in the transaction target database 31.

  Here, “transaction target ID” is an identifier of a transaction target. The “attribute information” is attribute information of a transaction object indicated by the associated transaction object ID, and is information indicating a category, a price, a manufacturer, an image, an explanation, a purchase time, and the like. The “distributed expression” is a distributed expression calculated by the model from the associated attribute information.

  For example, in the example illustrated in FIG. 3, the transaction target ID “ID # 1”, the category “category # 1”, the price “price # 1”, the manufacturer “manufacturer # 1”, the image “image # 1”, and the description “description” Information such as “text # 1”, purchase time “purchase time # 1”, and distributed expression “distributed expression # 1” is registered in association with each other. In such information, the category to which the transaction object indicated by the transaction object ID “ID # 1” belongs is the category “category # 1”, the price is the price “price # 1”, and the manufacturer is the manufacturer “manufacturer # 1”. The image is the image “image # 1”, the explanatory text is the explanatory text “explanatory text # 1”, and the purchase time is the purchase time “purchase time # 1”. Such information indicates that the distributed representation of the attribute information associated with the transaction target ID “ID # 1” is the distributed representation “distributed representation # 1”.

  In the example illustrated in FIG. 3, “ID # 1”, “Category # 1”, “Price # 1”, “Manufacturer # 1”, “Image # 1”, “Explanation # 1”, “Purchase Time #” Although conceptual values such as “1”, “distributed expression # 1” and the like are described, actually, text, numerical values, image data, and the like indicating each item are registered.

  Returning to FIG. 2, the description will be continued. The purchase history database 32 stores information indicating the purchase history of each transaction target. For example, FIG. 4 is a diagram illustrating an example of information registered in the purchase history database according to the embodiment. As shown in FIG. 4, “transaction target ID” and “purchase history” are registered in the purchase history database 32 in association with each other. Here, the “purchasing history” is the purchase history of the transaction target indicated by the associated transaction target ID, the identifier of the user who purchased the transaction target, the attribute of the user, the date and time when the purchase was performed, and the quantity Etc. are stored.

  For example, in the example illustrated in FIG. 4, the transaction target ID “ID # 1” and the purchase history “N / A” are registered in association with each other, and the transaction target ID “ID # 2” and the purchase history “purchase history #” are registered. 2 ”is registered in association with each other. Such information indicates that there is no purchase history in the transaction target indicated by the transaction target ID “ID # 1”, and the purchase history of the transaction target indicated by the transaction target ID “ID # 2” is the purchase history “purchase history # 2”. It shows that it is. In the example shown in FIG. 4, conceptual values such as “purchase history # 2” are described, but in practice, detailed data indicating the purchase history is registered.

  Returning to FIG. 2, the description will be continued. In the classification result database 33, a classification result obtained by classifying each transaction object based on a purchase tendency is registered. For example, FIG. 5 is a diagram illustrating an example of information registered in the classification result database according to the embodiment. As shown in FIG. 5, “classification ID”, “transaction target ID”, and “purchasing tendency” are registered in the classification result database 33 in association with each other. Here, the “classification ID” is an identifier indicating a group including transaction objects having a similar or similar purchasing tendency. The “purchasing tendency” is information indicating a purchasing tendency based on the purchase history of the transaction target indicated by the associated transaction target ID.

  For example, in the example illustrated in FIG. 5, the classification ID “group # 1”, the transaction target IDs “ID # 1, ID # 2”, and the purchase tendency “purchasing tendency # 1” are registered in association with each other. In such information, the purchase trend of the transaction target indicated by the transaction target ID “ID # 1” and the transaction target ID “ID # 2” is the purchase trend “purchase trend # 1”, and the classification ID “group # 1” is Indicates that it is classified into the group indicated.

  Returning to FIG. 2, the description will be continued. In the model database 34, data of a model learned by the learning process is registered. For example, in the model database 34, information on DNN nodes used as models, connection coefficients in connection paths between nodes, and the like are registered.

  The control unit 40 is a controller. For example, various programs stored in a storage device inside the determination apparatus 10 are stored in a RAM or the like by a processor such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). This is realized by being executed as a work area. The control unit 40 is a controller, and may be realized by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

  As illustrated in FIG. 2, the control unit 40 includes a collection unit 41, a calculation unit 42, a classification unit 43, a specification unit 44, a determination unit 45, and a learning unit 46. The collection unit 41 collects transaction target attribute information and purchase history. For example, the collection unit 41 collects transaction target information and purchase history including transaction target attribute information from the EC server 100 at predetermined time intervals, and sets the attribute information included in the collected transaction target information to the transaction target database 31. And the purchase history is registered in the purchase history database 32.

  The calculation unit 42 uses the model learned by the learning unit 46 to calculate a distributed expression indicating the attribute information of each transaction target. That is, the calculation unit 42 is a model that determines a distributed expression indicating an attribute of a transaction object from the attributes of the transaction object, and is a model in which learning is performed so that the distributed expressions of transaction objects with similar purchasing tendencies are similar. Is used to calculate the distributed representation of each transaction object.

  For example, the calculation unit 42 reads a model from the model database 34. And the calculation part 42 calculates the dispersion | distribution expression which shows attribute information for every transaction object registered into the transaction object database 31 using the read model. That is, the calculation unit 42 calculates a distributed expression based on the attributes of each transaction object so that the distributed expressions of transaction objects having the same purchase tendency are similar. Then, the calculation unit 42 registers the calculated distributed expression in the transaction target database 31.

  The classification unit 43 uses the distributed expression to classify the transaction objects into sets corresponding to purchase trends. For example, the classification unit 43 refers to the purchase history database 32 and identifies the purchase tendency of each transaction target. Then, the classification unit 43 classifies transaction objects having similar purchasing trends into the same group, and registers the classification results and the purchasing trends in the classification result database 33 in association with each other. Here, the classification | category part 43 notifies the transaction object ID of new transaction object to the specific | specification part 44, when the transaction object which does not have a purchase history, ie, new transaction object exists.

  The identification unit 44 identifies another transaction object similar to the predetermined transaction object using a distributed expression based on information on the transaction object. For example, the specifying unit 44 specifies another transaction target similar to the predetermined transaction target by using a distributed expression based on at least one of a description or a photograph of the predetermined transaction target. That is, the specifying unit 44 specifies another transaction target that is estimated to have a similar purchase tendency to the predetermined transaction target.

  In addition, for example, the identification unit 44 classifies the transaction target classified into a plurality of sets according to a purchase tendency as another transaction target in the same set as the transaction target similar to the attribute of the predetermined transaction target. Identify the transaction target. Further, for example, the specifying unit 44 specifies a transaction target having a distributed expression indicating a transaction target attribute similar to a predetermined transaction target as another transaction target.

  For example, the specifying unit 44 refers to the transaction target database 31, specifies a distributed representation of a new transaction target, and searches for another distributed representation similar to the specified distributed representation. That is, the specifying unit 44 searches for a transaction object whose distributed representation is similar to a new transaction object. By such processing, the specifying unit 44 can specify a transaction target that is predicted to have a similar new purchase target and purchase history. Then, the specifying unit 44 specifies from the classification result database 33 a group to which a transaction target whose distributed representation is similar to the new transaction target belongs, and classifies the new transaction target into the specified group.

  The determination unit 45 determines whether or not to present a predetermined transaction target to the user based on the purchase history of another transaction target specified by the specifying unit 44. For example, when the determination unit 45 receives a request from the terminal device 200, the determination unit 45 determines whether or not each purchase tendency registered in the classification result database 33 matches the user information of the user U01 who uses the terminal device 200. To do. And when it determines with the determination part 45 matching with the user information of the user U01 about a certain purchase tendency, the transaction object ID matched with the purchase tendency will be extracted, and the transaction which the extracted transaction object ID shows It is determined that the target is presented to the user.

  Here, in the classification result database 33, a new transaction target having no purchase history is registered in the same group as a transaction target having a similar purchase tendency by the process executed by the specifying unit 44. In addition, purchase trends based on purchase histories of transaction targets other than new transaction targets among the transaction targets classified into groups are registered in association with each group. Therefore, the determination unit 45 determines whether each purchase tendency registered in the classification result database 33 matches the attribute of the user U01, thereby purchasing other transaction targets similar to the new transaction target. Based on the history, it can be accurately determined whether or not a new transaction target should be presented to the user U01.

  Moreover, the determination part 45 produces | generates the content which advertises the transaction object which the extracted transaction object ID shows. Then, the determination unit 45 distributes the generated content to the terminal device 200. Note that the determination unit 45 may select a transaction target as a content distribution target from the transaction target indicated by the extracted transaction target ID based on any listing logic, and distribute only the selected transaction target content. .

  Further, the determination unit 45 may select the transaction target to be presented in consideration of the purchase history of the user U01 when the purchase order of the transaction target and the classification considering the simultaneous purchase are performed. Good. For example, the classification unit 43 classifies transaction targets that tend to be purchased at the same time as the first transaction target into a predetermined group (hereinafter referred to as “simultaneous purchase group”). Further, when the attribute of the new transaction target is similar to the transaction target of the simultaneous purchase group, the specifying unit 44 classifies the new transaction target into the simultaneous purchase group.

  In such a case, the determination unit 45 inquires of the EC server 100 about the transaction target put in the shopping cart of the user U01, and whether the first transaction target is put in the shopping cart, that is, the first It is determined whether or not the transaction object is to be purchased. And when the determination part 45 determines with the 1st transaction object being a purchase plan, it determines with showing the user the transaction object classified into the simultaneous purchase group, and the transaction object classified into the simultaneous purchase group Is presented to the user U01. By executing such processing, the determination apparatus 10 is a transaction target that is highly likely to be purchased at the same time, such as ink or paper, which becomes a new product, for a user who wants to purchase a printer. A new transaction object having no purchase history can be presented.

  In addition, for example, the classification unit 43 classifies transaction targets that tend to be purchased after the first transaction target into a predetermined group (hereinafter referred to as “sequential purchase group”). Further, when the attribute of the new transaction target is similar to the transaction target of the sequential purchase group, the specifying unit 44 classifies the new transaction target into the sequential purchase group.

  In such a case, the determination unit 45 inquires of the purchase history of the user U01 to the EC server 100 or the like, and determines whether the user U01 has already purchased the first transaction target. And when the determination part 45 determines with the 1st transaction object having been purchased, it determines with showing the transaction object classified into the sequential purchase group to a user, and the transaction object classified into the sequential purchase group Is presented to the user U01. By executing such processing, the determination apparatus 10 can present a transaction target such as ink or paper to be a new product to a user who has purchased a printer. Further, the determination device 10 can prevent uncertain presentation such as presenting a printer of a new product to a user who has purchased ink, paper, or the like.

  The learning unit 46 executes a learning process for learning a model. For example, the learning unit 46 learns a model for determining the distributed representation of the transaction target from the attributes of the transaction target so that the distributed representation of the transaction target with similar tendencies indicated by the purchase history is similar. Further, for example, the learning unit 46 uses the tendency indicated by the purchase history of the transaction object as information indicating the classification destination of the transaction object, and causes the model to learn the characteristics of the attribute of the transaction object. Further, for example, the learning unit 46 sets the attribute of the user who purchased the transaction target as information indicating the classification target of the transaction target, that is, a purchase tendency.

  For example, the learning unit 46 refers to the purchase history database 32 and selects a transaction target having a sufficient purchase history as a learning target. Further, like the classification unit 43, the learning unit 46 classifies learning objects with similar purchasing trends into the same group. Note that the learning unit 46 may specify the learning target to be classified into the same group with reference to the classification result database 33.

  Then, the learning unit 46 learns a model for calculating the distributed expression from the attributes of each learning object so that the distributed expressions of the learning objects classified into the same group are similar. At this time, the learning unit 46 has the similarity of the distributed representation of the learning target classified into the group having similar purchasing tendency becomes the first threshold value or more and the second threshold value or less, and the learning target classified into the same group. You may learn a model so that the similarity of a distributed expression may become larger than a 2nd threshold value. Further, the learning unit 46 may perform learning of the model so that the similarity of the distributed representation of the learning target classified into the dissimilar groups is equal to or less than a predetermined threshold.

  That is, the learning unit 46 may perform arbitrary learning as long as it is a model that can calculate a similar distributed expression from the attributes of transaction objects with similar purchasing tendencies. As a result, the learning unit 46 can generate a model that can classify a new transaction object having no purchase history into a group having a similar purchase tendency, based on the similarity of attributes in consideration of the purchase tendency.

[3. Example of judgment process flow)
Subsequently, a flow of determination processing executed by the determination device 10 will be described with reference to FIG. FIG. 6 is a flowchart illustrating an example of a flow of determination processing executed by the determination device according to the embodiment. For example, the determination apparatus 10 collects new transaction target information together with information on each transaction target (step S101). In such a case, the determination apparatus 10 uses the model for calculating the distributed representation from the transaction target attributes so that the distributed representations of the transaction targets with similar purchasing tendencies are similar to each other. (Step S102). And the determination apparatus 10 specifies the other transaction object with which distributed expression is similar (step S103), and complete | finishes a process.

  Next, an example of processing in which the determination apparatus 10 presents a transaction target will be described with reference to FIG. FIG. 7 is a flowchart illustrating an example of a process in which the determination apparatus according to the embodiment presents a transaction target. In addition, in the example shown in FIG. 7, it has described about the flow of the process performed about a new transaction object. For example, when receiving a request for a transaction target from a user, the determination device 10 determines whether to present the new transaction target to the user based on the purchase history of another transaction target selected for the new transaction target ( Step S104). That is, the determination apparatus 10 determines whether or not to present a new transaction target to the user based on a purchase tendency indicated by a purchase history of another transaction target classified into the same group as the new transaction target.

  If the determination device 10 determines to present (step S105: Yes), the information about the new transaction is distributed to the user (step S106), and the process ends. Moreover, when the determination apparatus 10 determines not to present (step S105: No), it distributes other transaction target information and ends the process.

[4. (Modification)
In the above, an example of the determination process by the determination apparatus 10 has been described. However, the embodiment is not limited to this. Hereinafter, the variation of the determination process which the determination apparatus 10 performs is demonstrated.

[4-1. Device configuration〕
The determination device 10 may be realized by a server device, a cloud system, or the like that can be identified with the EC server 100, and may operate as a back-end server of the EC server 100. Further, the function exhibited by the determination device 10 may be included in the EC server 100.

  Moreover, the determination apparatus 10 may implement | achieve the determination process and learning process which were mentioned above, when a some apparatus cooperates. For example, the determination apparatus 10 may be realized by a determination server having a collection unit 41, a calculation unit 42, a classification unit 43, a specification unit 44, and a determination unit 45, and a learning server having a learning unit 46. Moreover, each database 31-34 stored in the memory | storage part 30 may be stored not in the determination apparatus 10, but in an external storage server etc., for example.

[4-2. (Delivery timing)
The determination device 10 may distribute the content related to the transaction target to the terminal device 200 at an arbitrary timing. For example, the determination apparatus 10 may distribute content related to a transaction target as advertisement content displayed when browsing arbitrary web content such as a portal site or a web page.

  Moreover, the determination apparatus 10 may distribute the content regarding a transaction object, when displaying the search result of a transaction object. When executing such processing, the determination apparatus 10 may perform content classification in consideration of a search query input by the user. For example, the determination apparatus 10 learns a model so as to classify what kind of search query is purchased as a purchase tendency, and according to the search query input by the user U01, What is necessary is just to select the group of transaction object shown to user U01.

[4-3. About the model)
Here, the distribution apparatus 10 is a model that determines the scattered expression from the attributes of the transaction target, and is a model that has been learned so that the distribution expression of the transaction target having a predetermined tendency is similar to the tendency indicated by the purchase history. Was used to calculate the distributed representation of the new transaction object and other transaction objects, and based on the calculation results, identified other transaction objects with similar purchase trends to the new cancellation object. However, the embodiment is not limited to this.

  For example, the determination apparatus 10 determines the value of the distributed expression from the information about the transaction target or the information about the user so that the distributed expression of the information about the transaction target is similar to the distributed expression of the information about the user who purchased the transaction target. Learn the model to be determined. Then, the determination device 10 generates a distributed representation from information regarding each transaction target including the new transaction target, and performs clustering of similar distributed representations, thereby selecting another transaction target having a similar purchase tendency to the new transaction target. May be. Further, the determination device 10 compares the distributed representation of the new transaction target with the distributed representation of the user, and determines that the information on the new transaction target is provided to the user when each distributed representation is similar. Also good.

  Further, the determination device 10 calculates a distributed representation of the user using the model, and identifies a transaction target having a distributed representation similar to the calculated distributed representation, thereby determining a transaction target that is likely to be purchased by the user. You may specify. As described above, the determination apparatus 10 learns a model so as to generalize the model to both the transaction target and the user, and realize clustering according to the sales tendency of the transaction target. As a result, the determination apparatus 10 can improve the accuracy of the transaction target proposed to the user.

[4-4. Others]
In addition, among the processes described in the above embodiment, all or part of the processes described as being automatically performed can be performed manually, or the processes described as being performed manually can be performed. All or a part can be automatically performed by a known method. In addition, the processing procedures, specific names, and information including various data and parameters shown in the document and drawings can be arbitrarily changed unless otherwise specified. For example, the various types of information illustrated in each drawing is not limited to the illustrated information.

  Further, each component of each illustrated apparatus is functionally conceptual, and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. Can be integrated and configured.

  In addition, the above-described embodiments can be appropriately combined within a range in which processing contents are not contradictory.

[5. program〕
Further, the determination apparatus 10 according to the above-described embodiment is realized by a computer 1000 having a configuration as illustrated in FIG. 8, for example. FIG. 8 is a diagram illustrating an example of a hardware configuration. The computer 1000 is connected to an output device 1010 and an input device 1020, and an arithmetic device 1030, a primary storage device 1040, a secondary storage device 1050, an output IF (Interface) 1060, an input IF 1070, and a network IF 1080 are connected via a bus 1090. Have

  The arithmetic device 1030 operates based on a program stored in the primary storage device 1040 and the secondary storage device 1050, a program read from the input device 1020, and the like, and executes various processes. The primary storage device 1040 is a memory device such as a RAM that temporarily stores data used by the arithmetic device 1030 for various arithmetic operations. The secondary storage device 1050 is a storage device in which data used for various calculations by the calculation device 1030 and various databases are registered, and is realized by a ROM (Read Only Memory), HDD, flash memory, or the like.

  The output IF 1060 is an interface for transmitting information to be output to an output device 1010 that outputs various types of information such as a monitor and a printer. For example, USB (Universal Serial Bus), DVI (Digital Visual Interface), This is realized by a standard connector such as HDMI (registered trademark) (High Definition Multimedia Interface). The input IF 1070 is an interface for receiving information from various input devices 1020 such as a mouse, a keyboard, and a scanner, and is realized by, for example, a USB.

  The input device 1020 includes, for example, an optical recording medium such as a CD (Compact Disc), a DVD (Digital Versatile Disc), and a PD (Phase change rewritable disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), and a tape. It may be a device that reads information from a medium, a magnetic recording medium, a semiconductor memory, or the like. The input device 1020 may be an external storage medium such as a USB memory.

  The network IF 1080 receives data from other devices via the network N and sends the data to the arithmetic device 1030, and transmits data generated by the arithmetic device 1030 to other devices via the network N.

  The arithmetic device 1030 controls the output device 1010 and the input device 1020 via the output IF 1060 and the input IF 1070. For example, the arithmetic device 1030 loads a program from the input device 1020 or the secondary storage device 1050 onto the primary storage device 1040, and executes the loaded program.

  For example, when the computer 1000 functions as the determination device 10, the arithmetic device 1030 of the computer 1000 implements the function of the control unit 40 by executing a program loaded on the primary storage device 1040.

[6. effect〕
As described above, the determination apparatus 10 specifies another transaction target similar to the predetermined transaction target using the distributed expression based on the information regarding the transaction target. And the determination apparatus 10 determines whether a predetermined | prescribed transaction object is shown to a user based on the purchase history of the other identified transaction object. As a result, the determination device 10 can present to the user U01 who is highly likely to purchase even if the transaction target has no purchase history. For this reason, the determination apparatus 10 can select the transaction object proposed to the user with high accuracy.

  Moreover, the determination apparatus 10 specifies another transaction object similar to the predetermined transaction object by using a distributed expression based on at least one of or both of a description or a photograph of the predetermined transaction object. As a result, the determination apparatus 10 can accurately select a transaction target to be proposed to the user even if the transaction target has no purchase history.

  Moreover, the determination apparatus 10 specifies the other transaction object estimated that a predetermined transaction object and a purchase tendency are similar. More specifically, the determination apparatus 10 determines, as other transaction objects, among the transaction objects classified into a plurality of groups according to purchase tendency, the same set as the transaction object whose attribute is similar to the attribute of the predetermined transaction object Identify trading objects classified as. As a result, the determination device 10 can realize a proposal based on a purchase tendency even when the transaction target has no purchase history.

  Moreover, the determination apparatus 10 specifies the transaction object whose distributed expression which shows the attribute of a transaction object is similar to a predetermined transaction object as another transaction object. For example, the determination apparatus 10 is a model that determines a distributed expression indicating an attribute of a transaction object from the attributes of the transaction object, so that the distribution expression of the transaction object having a predetermined tendency is similar to the tendency indicated by the purchase history. A distributed representation of a predetermined transaction object and other transaction objects is calculated using the learned model. And the determination apparatus 10 specifies the other transaction object similar to a predetermined transaction object using the calculated dispersion | distribution expression. Specifically, the determination apparatus 10 identifies a transaction object that is similar in distribution representation to a predetermined transaction object.

  Further, for example, the determination apparatus 10 learns a model that determines the distributed representation of a transaction target from the attributes of the transaction target so that the distributed representations of the transaction target with similar tendencies indicated by the purchase history are similar. And the determination apparatus 10 calculates the dispersion | distribution expression of transaction object using a model. And the determination apparatus 10 uses the tendency which the purchase history of a transaction object shows as the information which shows the classification | category destination of the transaction object, and makes the model learn the characteristic which the attribute of the transaction object has.

  As described above, the determination apparatus 10 is a distributed expression indicating the attributes of a transaction object, and uses a distributed expression having a similarity according to a purchase tendency, and other transaction objects whose purchase tendency is similar to the predetermined transaction object. Select. For this reason, the determination apparatus 10 presents the transaction target to the user based on the purchase trend of another transaction target that is estimated to have a similar purchase trend even if the predetermined transaction target has no purchase history. Therefore, it is possible to improve the accuracy of the transaction target proposed to the user.

  Moreover, the determination apparatus 10 uses the attribute of the user who purchased the transaction target as information indicating the classification destination of the transaction target. Moreover, the determination apparatus 10 classifies each transaction target into a set corresponding to the purchase tendency of the transaction target using the distributed expression, and other than the set in which the transaction target similar in attribute to the predetermined transaction target is classified. Identify the target of the transaction. For this reason, the determination apparatus 10 can improve the accuracy of the transaction target proposed to the user.

  In addition, in the determination device 10, when a predetermined transaction object is classified into a group in which transaction objects to be purchased at the same time as the first transaction object are classified, the user plans to purchase the first transaction object. If it is determined that the purchase is scheduled, it is determined that a predetermined transaction target is presented to the user. For this reason, the determination apparatus 10 can select the transaction object proposed to the user in consideration of the simultaneous purchase of the transaction object.

  In addition, when the predetermined transaction target is classified into a group in which the transaction target purchased after the first transaction target is classified, the determination device 10 has purchased the first transaction target. If it is determined that it has been purchased, it is determined that a predetermined transaction target is presented to the user. For this reason, the determination apparatus 10 can select the transaction object proposed to the user in consideration of the order in which the transaction objects are purchased.

  In addition, the determination apparatus 10 determines the value of the distributed expression from the information about the transaction target or the information about the user so that the distributed expression of the information about the transaction target is similar to the distributed expression of the information about the user who purchased the transaction target. A distributed representation of a predetermined transaction target or a distributed representation of a predetermined user is calculated using a model to be determined. And the determination apparatus 10 specifies the other transaction object or other user whose calculated distributed expression is similar to a predetermined transaction object or a predetermined user. For this reason, the determination apparatus 10 can distribute, for example, information on a transaction target having no purchase history to an appropriate user. Moreover, the determination apparatus 10 can distribute appropriate transaction target information to a user who does not have a purchase history.

  As described above, some of the embodiments of the present application have been described in detail with reference to the drawings. However, these are merely examples, and various modifications, including the aspects described in the disclosure section of the invention, based on the knowledge of those skilled in the art, It is possible to implement the present invention in other forms with improvements.

  Moreover, the above-mentioned “section (module, unit)” can be read as “means”, “circuit”, and the like. For example, the classification unit can be read as classification means or a classification circuit.

DESCRIPTION OF SYMBOLS 10 Determination apparatus 20 Communication part 30 Storage part 31 Transaction object database 32 Purchasing history database 33 Classification result database 34 Model database 40 Control part 41 Collection part 42 Calculation part 43 Classification part 44 Identification part 45 Determination part 46 Learning part 100 EC server 200 Terminal apparatus

Claims (14)

Using a distributed representation based on information about the transaction object, a specific unit that identifies another transaction object similar to the predetermined transaction object,
And a determination unit that determines whether or not to present the predetermined transaction target to a user based on a purchase history of another transaction target specified by the specifying unit. The specifying unit specifies another transaction object similar to the predetermined transaction object by using a distributed expression based on at least one or both of the description or the photograph of the predetermined transaction object. The determination apparatus according to claim 1. The determination device according to claim 1, wherein the specifying unit specifies another transaction object that is estimated to have a purchasing tendency similar to the predetermined transaction object. A model for determining a distributed expression indicating an attribute of a transaction object from an attribute of the transaction object, and a model that is learned so that the distributed expression of the transaction object having a predetermined tendency is similar to the tendency indicated by the purchase history. Using a calculation unit that calculates a distributed representation of the predetermined transaction object and the other transaction object,
The said specific | specification part specifies the other transaction object similar to the said predetermined transaction object using the dispersion | distribution expression calculated by the said calculation part. The determination apparatus described. The determination unit according to claim 4, wherein the specifying unit specifies a transaction object having a distributed representation similar to the predetermined transaction object as the other transaction object. A learning unit that learns a model for determining a distributed representation of a transaction object from an attribute of the transaction object so that the distribution expression of the transaction object having a predetermined tendency is similar to the tendency indicated by the purchase history;
The determination device according to claim 5, wherein the calculation unit calculates a distributed representation of the transaction object using a model learned by the learning unit. The determination according to claim 6, wherein the learning unit uses the tendency indicated by the purchase history of the transaction target as information indicating the classification destination of the transaction target, and causes the model to learn the characteristics of the attribute of the transaction target. apparatus. The determination device according to claim 6, wherein the learning unit uses, as information indicating a classification destination of the transaction target, an attribute of a user who purchased the transaction target. A classifying unit that classifies each transaction object into a set according to a purchase tendency using the distributed expression;
The said specific | specification part specifies the said other transaction object from the group by which the transaction object with an attribute similar to the said predetermined transaction object was classified. The one of Claims 1-8 characterized by the above-mentioned. Judgment device. When the predetermined transaction object is classified into a set in which the transaction object to be purchased at the same time as the first transaction object is classified, the determination unit determines that the user plans to purchase the first transaction object. The determination apparatus according to any one of claims 1 to 9, wherein if it is determined whether or not the purchase is scheduled, the predetermined transaction object is determined to be presented to a user. . When the predetermined transaction object is classified into a group in which transaction objects purchased after the first transaction object are classified, the user has already purchased the first transaction object. The determination according to any one of claims 1 to 10, wherein if it is determined whether or not it has been purchased, it is determined to present the predetermined transaction target to a user. apparatus. A model for determining the value of the distributed representation from the information about the transaction target or the information about the user so that the distributed representation of the information about the transaction target is similar to the distributed representation of the information about the user who purchased the transaction target. A calculation unit for calculating a distributed representation of a predetermined transaction object or a distributed representation of a predetermined user;
The determination apparatus, wherein the distributed expression calculated by the calculation unit includes a specific unit that identifies the predetermined transaction target or another transaction target or other user similar to the predetermined user. A determination method executed by a determination device,
A specific process for identifying other transaction objects similar to a given transaction object using a distributed representation based on information about the transaction object;
And a determination step of determining whether or not to present the predetermined transaction object to a user based on a purchase history of the other transaction object specified by the specifying step. On the computer,
A specific procedure for identifying other transaction objects similar to a given transaction object using a distributed representation based on information about the transaction object;
And a determination procedure for determining whether or not to present the predetermined transaction object to a user based on a purchase history of another transaction object identified by the identification procedure.

Images